Compare commits

...

1468 Commits
v0.6.0 ... v0.9

Author SHA1 Message Date
Michael Vines
68ec814bf3 Rename buildkite-snap to buildkite-secondary 2018-11-05 08:48:58 -08:00
Michael Vines
d7f283a00a Add version for path dependencies 2018-10-24 20:29:20 -07:00
Pankaj Garg
609889cc58 Wildcard early OOM deb package revision () () 2018-10-22 10:59:22 -07:00
Michael Vines
83fa02ee44 Add ERC20-like Token program 2018-10-09 11:44:46 -07:00
Michael Vines
d8fcb22000 Purge BudgetTransaction from banking_stage 2018-10-09 11:44:46 -07:00
Michael Vines
6b1c90a8b5 Demote log messages 2018-10-09 11:44:46 -07:00
Michael Vines
67d72e709f Demote 'not enough peers in crdt table' log message 2018-10-02 21:55:28 -07:00
Michael Vines
f0d2870e0f Run a fullnode+drone automatically when the container starts up 2018-10-02 18:09:56 -07:00
Michael Vines
02c47b48da Publish minimal Solana docker images to dockerhub 2018-10-02 16:28:54 -07:00
Michael Vines
d30a39cd38 Remove SNAP_ prefix 2018-10-02 16:28:54 -07:00
Michael Vines
9e57d0467e Return all instances 2018-10-01 07:51:12 -07:00
Michael Vines
fb57d13c03 Correctly deserialize large userdata 2018-09-29 19:26:26 -07:00
Michael Vines
41f6e27bba Ignore keep alive for completed games 2018-09-29 19:26:26 -07:00
Michael Vines
aff5649b39 Add tic-tac-toe dashboard program 2018-09-28 19:13:09 -07:00
Michael Vines
9fd6ffe83f s/grid/board/g 2018-09-28 19:13:09 -07:00
Michael Vines
ff805361a9 Simplify game setup messaging 2018-09-28 19:13:09 -07:00
Michael Vines
30d1b0b4bf Add KeepAlive message so players can detect abandoned games 2018-09-28 19:13:09 -07:00
Tyera Eulberg
5233cf1ca6 Update for new solana-jsonrpc 2018-09-28 18:10:15 -07:00
Tyera Eulberg
d2754fd702 Create new wallet on each run of wallet-sanity 2018-09-28 07:40:17 -07:00
Michael Vines
ce9a0ae215 Specify zone 2018-09-28 07:35:33 -07:00
Michael Vines
067adcdfa8 Include -z when deleting network 2018-09-27 21:27:38 -07:00
Greg Fitzgerald
675ff64094 Fail CI on clippy warnings 2018-09-27 16:21:12 -06:00
Greg Fitzgerald
423e7ebc3f Pacify clippy 2018-09-27 16:21:12 -06:00
Greg Fitzgerald
f9fe6a0f72 Move clippy to Rust stable 2018-09-27 16:21:12 -06:00
Greg Fitzgerald
8d007bd7f7 Upgrade rustc and add clippy to stable 2018-09-27 16:21:12 -06:00
Stephen Akridge
6cdbdfbbcb Enable bench and fix upload-perf 2018-09-27 14:16:56 -07:00
Pankaj Garg
35e6343d61 Update testnet-deploy script to configure GPUs for leader node () 2018-09-27 13:42:24 -07:00
Pankaj Garg
7fb7839c8f Configure GPU type/count from command line in GCE scripts ()
* Configure GPU type/count from command line in GCE scripts

* Change CLI to input full leader machine type information with GPU
2018-09-27 11:55:56 -07:00
Tyera Eulberg
dbc1ffc75e Use jsonrpc fork 2018-09-27 12:50:38 -06:00
Michael Vines
1fdbe893c5 Improve game setup experience: X now shares game key and accepts O 2018-09-27 10:44:13 -07:00
Pankaj Garg
55a542bff0 Fix erasure and cuda related compilation errors 2018-09-27 10:42:37 -06:00
Pankaj Garg
e10574c64d Remove recycler and it's usage
- The memory usage due to recycler was high, and incrementing with
  time.
2018-09-27 10:42:37 -06:00
Anatoly Yakovenko
2e00be262e Remove data from BankError.
This reduces how much memory is written to last_id_sigs table on very TX, and has a 40% impact on
`cargo +nightly watch -x 'bench bench_banking_stage'`
2018-09-27 09:07:56 -06:00
Greg Fitzgerald
4172bde081 Only send a vote once a second 2018-09-27 09:06:41 -06:00
jackcmay
9c47e022dc break dependency of programs on solana core ()
* break dependency of programs on Solana core
2018-09-27 07:49:26 -07:00
Greg Fitzgerald
874addc51a Move KeyedAccount into Account
Now programs don't need to depend on dynamic_program and its
dependencies.
2018-09-26 20:40:40 -06:00
Greg Fitzgerald
b7ae5b712a Move Pubkey into its own module 2018-09-26 20:40:40 -06:00
Greg Fitzgerald
c6d7cd2d33 Move Account into its own module
Also use default Default generator, since system program ID is
[0; 32]. Bank should probably be the one to set this anyway.
2018-09-26 20:40:40 -06:00
Rob Walker
386a96b7e0 capture multinode logs by default () 2018-09-26 19:30:40 -07:00
Michael Vines
b238c57179 Add trace! when an error is mapped to GenericFailure 2018-09-26 19:30:20 -07:00
Michael Vines
1821e72812 Add getSignatureStatus 2018-09-26 19:00:34 -07:00
Rob Walker
a23c230603 fix reverse loop in write_stage, simplify banking_stage, add tooling to help find this () 2018-09-26 18:37:24 -07:00
Greg Fitzgerald
4e01fd5458 Update test to show when we should collect tx fees
See  for details. The `from` account should be cloned
before execute_transaction(), and that's the only one that should
be stored if there's an error executing the program.
2018-09-26 19:30:27 -06:00
Greg Fitzgerald
e416cf7adf Let clients know when transactions failed 2018-09-26 19:30:27 -06:00
Anatoly Yakovenko
25edb9e447 fix benches 2018-09-26 19:29:46 -06:00
Anatoly Yakovenko
93c4f6c9b8 Synchronize PoH, bank last_id queue and ledger entry channel.
PoH, bank's last_id queue and the Entry channel need to have a synchronized order of ids.
2018-09-26 16:19:03 -07:00
Anatoly Yakovenko
718031ec35 Ignore the test_leader_to_validator_transition until it can handle PoH entries 2018-09-26 16:59:57 -06:00
Stephen Akridge
d546614936 Handle deserialize failure with error 2018-09-26 15:17:07 -07:00
Greg Fitzgerald
ac8d738045 Don't call unwrap() in StorageProgram::process_tx 2018-09-26 15:17:07 -07:00
Greg Fitzgerald
ca962371b8 Fix build
Two PRs crossed in flight.
2018-09-26 14:40:48 -06:00
Rob Walker
e6f8922e35 fix issue () 2018-09-26 13:31:39 -07:00
Greg Fitzgerald
7292ece7ad Free up term instruction for new multi-instruction feature 2018-09-26 14:17:15 -06:00
Greg Fitzgerald
df3b78c18c Move BudgetTransaction into its own module 2018-09-26 14:17:15 -06:00
Greg Fitzgerald
c83dcea87d Move SystemTransaction into its own module 2018-09-26 14:17:15 -06:00
Greg Fitzgerald
be20c99758 Promote the one true transaction constructor 2018-09-26 14:17:15 -06:00
Greg Fitzgerald
694add9919 Move budget-specific and system-specific tx constructors into traits
These functions pull in budget-specific and system-specific
dependencies that aren't needed by the runtime.
2018-09-26 14:17:15 -06:00
Michael Vines
afc764752c Permit testnets without a GPU 2018-09-26 10:37:41 -07:00
Tyera Eulberg
113c8b5880 Rollback jsonrpc SendTransaction pool for signature; ignore flaky tests 2018-09-26 10:25:29 -07:00
Stephen Akridge
a5b28349ed Add max entry height to download for replicator 2018-09-26 09:57:22 -07:00
Greg Fitzgerald
bb7ecc7cd9 Migrate to solana-labs fork of jsonrpc
This changes aims to be a no-op. Future changes to rev should be
along the new solana-0.1 branch.
2018-09-26 10:08:37 -06:00
Tyera Eulberg
14bc160674 Clean up test and add signature return to rpc send tx 2018-09-25 16:38:51 -07:00
Tyera Eulberg
d438c22618 Update RFC 2018-09-25 16:38:51 -07:00
Tyera Eulberg
bcbae0a64f Fix witness functionality 2018-09-25 16:38:51 -07:00
Tyera Eulberg
f636408647 Fix timestamp and cancel functionality
- Also serialize and send helper fn
2018-09-25 16:38:51 -07:00
Tyera Eulberg
3ffc7aa5bc Add helper fn to get last id 2018-09-25 16:38:51 -07:00
Tyera Eulberg
7b7e8c0d3f Clippy 2018-09-25 16:38:51 -07:00
Tyera Eulberg
11ea9e7c4b Add cancelable handling 2018-09-25 16:38:51 -07:00
Tyera Eulberg
2b82121325 Fix wallet-sanity to reflect new wallet arg syntax 2018-09-25 16:38:51 -07:00
Tyera Eulberg
5038e5ccd7 Preliminary Wallet-Budget functionality 2018-09-25 16:38:51 -07:00
Tyera Eulberg
e943ed8caf Expand parse_command and add tests 2018-09-25 16:38:51 -07:00
Tyera Eulberg
c196952afd Flesh out Wallet CLI & add placeholder WalletCommands 2018-09-25 16:38:51 -07:00
carllin
e7383a7e66 Validator to leader ()
* Add check in window_service to exit in checks for leader rotation, and propagate that service exit up to fullnode

* Added logic to shutdown Tvu once ReplicateStage finishes

* Added test for successfully shutting down validator and starting up leader

* Add test for leader validator interaction

* fix streamer to check for exit signal before checking socket again to prevent busy leaders from never returning

* PR comments - Rewrite make_consecutive_blobs() function, revert genesis function change
2018-09-25 15:41:29 -07:00
Rob Walker
8a7545197f move tick generation back to banking_stage, add unit tests ()
* move tick generation back to banking_stage, add unit tests

fixes 

* remove channel() stuff for synchronous comm; use a mutex
2018-09-25 15:01:51 -07:00
Greg Fitzgerald
680072e5e2 No need to special case vote failures 2018-09-25 13:43:35 -06:00
Greg Fitzgerald
4ca377a655 Delete dead code 2018-09-25 13:43:35 -06:00
Tyera Eulberg
751dd7eebb Move vote into ReplicateStage after process_entries 2018-09-25 13:43:35 -06:00
Michael Vines
8f0e0c4440 Add tic-tac-toe program 2018-09-25 12:07:41 -07:00
Michael Vines
50cf73500e Remove rfc 004 2018-09-25 12:07:41 -07:00
Tyera Eulberg
db310a044c Add Budget::And element, and supporting functions () 2018-09-25 12:38:13 -06:00
Rob Walker
88a609ade5 groom write_stage 2018-09-25 00:18:35 -07:00
Rob Walker
304d63623f give replication some time to happen
fixes 
2018-09-24 23:57:09 -07:00
Rob Walker
407b2682e8 remove dead code 2018-09-24 23:12:09 -07:00
Pankaj Garg
0f4fd8367d Add counters for channel pressure and time spent in TPU pipeline ()
* Add counters for channel pressure and time spent in TPU pipeline

* Fixed failing tests

* Fix rust format issue
2018-09-24 17:13:49 -07:00
Greg Fitzgerald
747ba6a8d3 Boot BudgetState::last_error 2018-09-24 17:14:23 -06:00
Greg Fitzgerald
bb99fd40de Update transaction status in the bank
This will allow jsonrpc to query the system to find out if a
recent transaction failed.
2018-09-24 17:14:23 -06:00
Greg Fitzgerald
e972d6639d Return errors from BudgetProgram::process_transaction 2018-09-24 17:14:23 -06:00
Greg Fitzgerald
22e77c9485 Add a way of getting transaction errors out of the bank 2018-09-24 17:14:23 -06:00
Pankaj Garg
bc88473030 Increase wmem for kernel network memory usage ()
- Validators were running out of kernel buffer while retransmitting
  blobs
2018-09-24 13:02:56 -07:00
Greg Fitzgerald
95677a81c5 Pacify clippy 2018-09-24 13:36:31 -06:00
Greg Fitzgerald
ea37d29d3a Pass Bank::process_transactions() a reference to the txs instead of moving them 2018-09-24 13:36:31 -06:00
sakridge
e030673c9d Do a recv on join to prevent channel destruction ()
before window thread join
2018-09-24 11:50:37 -07:00
sakridge
3e76efe97e Fix bench compilation () 2018-09-24 10:40:42 -07:00
Stephen Akridge
f5a30615c1 Ignore replicator startup for now 2018-09-24 09:43:58 -06:00
Michael Vines
e5e325154b Add --shell argument 2018-09-24 08:05:47 -07:00
Rob Walker
9e3d2956d8 remove last recycle? 2018-09-24 08:09:41 -06:00
jackcmay
26b1466ef6 Initial integration of dynamic contracts and native module loading ()
* Integration of native dynamic programs
2018-09-23 22:13:44 -07:00
Jack May
a1f01fb8f8 revert is_some to not is_none, causes test failure 2018-09-23 17:09:18 -06:00
Jack May
b2be0e2e5e fix clippy warning 2018-09-23 17:09:18 -06:00
Jack May
1a45587c08 fix clippy warnings 2018-09-23 17:09:18 -06:00
sakridge
3199f174a3 Add option to pass boot disk type to gce create () 2018-09-22 16:43:47 -07:00
Rob Walker
a51c2f193e fix Rob and Carl crossing wires 2018-09-21 21:37:25 -07:00
Rob Walker
be31da3dce lastidnotfound step 2: ()
lastidnotfound step 2:
  * move "record stage", aka poh_service into banking stage
  * remove Entry.has_more, is incompatible with leader rotation
  * rewrite entry_next_hash in terms of Poh
  * simplify and unify transaction hashing (no embedded nulls)
  * register_last_entry from banking stage, fixes  (w00t!)
  * new PoH doesn't generate empty ledger entries, so some fixes necessary in 
         multinode tests that rely on that (e.g. giving validators airdrops)
  * make window repair less patient, if we've been waiting for an answer, 
          don't be shy about most recent blobs
   * delete recorder and record stage
   * make more verbost  thin_client error reporting
   * more tracing in window (sigh)
2018-09-21 21:01:13 -07:00
sakridge
54b407b4ca Wait on blob fetch before window, Seems to fix instability ()
also cleanup ledger.
2018-09-21 18:56:20 -07:00
Tyera Eulberg
e87cac06da Request/reqwest improvements
- Use json macro to simplify request builds
- Add proxy option for reqwest to use TLS
- Add rpc port options for configured nodes
2018-09-21 18:06:20 -06:00
Tyera Eulberg
ad4fef4f09 Doc for rpc_port configuration 2018-09-21 18:06:20 -06:00
Tyera Eulberg
e3b3701e13 Add RPC port option to fullnode 2018-09-21 18:06:20 -06:00
Tyera Eulberg
9228fe11c9 Port Wallet to jsonrpc and fix tests 2018-09-21 18:06:20 -06:00
carllin
5ab38afa51 Changed the window_service in Replicator to send entries instead of blobs () 2018-09-21 16:50:58 -07:00
Rob Walker
e49b8f0ce7 Update poh_service.rs 2018-09-21 16:03:54 -07:00
carllin
c50ac96f75 Moved deserialization of blobs to entries from replicate_stage to window_service () 2018-09-21 16:01:24 -07:00
sakridge
a9355c33b2 Placeholder storage contract and replicator client ()
* Add hooks for executing the storage contract

* Add store_ledger stage
  Similar to replicate_stage but no voting/banking stuff, just convert
  blobs to entries and write the ledger out

* Add storage_addr to tests and add new NodeInfo constructor
  to reduce duplication...
2018-09-21 15:32:15 -07:00
Rob Walker
3dcee9f79e Update poh_service.rs 2018-09-21 08:01:24 -07:00
Michael Vines
2614189157 cargo fmt 2018-09-20 19:46:20 -07:00
Michael Vines
beeb09646a suppress warning: unused variable: recycler 2018-09-20 19:46:20 -07:00
Michael Vines
67f1fbab5f Treat rustc warnings as errors in CI 2018-09-20 19:46:20 -07:00
Tyera Eulberg
c0e7e43e96 fixup! s/contract/program 2018-09-20 19:33:54 -07:00
Anatoly Yakovenko
9bfead2e01 s/contract/program 2018-09-20 19:33:54 -07:00
Greg Fitzgerald
6073cd57fa Boot Recycler::recycle() 2018-09-20 17:08:51 -06:00
Michael Vines
5174be5fe7 Rename getAccount to getAccountInfo 2018-09-20 15:18:56 -07:00
Rob Walker
62a18d4c02 step one of lastidnotfound: record_stage->record_service, trim recorder to hashes ()
step one of lastidnotfound

* record_stage->record_service, trim recorder to hashes
* doc updates, hash multiple without alloc()

cc 
2018-09-20 15:02:24 -07:00
Michael Vines
a6c15684c9 Avoid panicking invalid instructions 2018-09-20 14:08:39 -07:00
Michael Vines
5691bf557c Handle bad account userdata better 2018-09-20 14:08:39 -07:00
Michael Vines
8f01f7cf21 Trace syscalls for more helpful logs 2018-09-20 14:08:39 -07:00
Michael Vines
bb8c94ad2c Add getAccount JSON RPC request 2018-09-20 13:58:15 -07:00
Greg Fitzgerald
d98e35e095 Delete no longer used PaymentPlan trait 2018-09-20 14:22:45 -06:00
Greg Fitzgerald
3163fbad0e Remove 'Plan' indirection since it's implied by BUDGET_CONTRACT_ID 2018-09-20 14:22:45 -06:00
Greg Fitzgerald
0172422961 Require a self-assigned account ID 2018-09-20 14:16:14 -06:00
Rob Walker
8ccfb26923 tests for my IP picker 2018-09-20 09:21:09 -07:00
Rob Walker
12a474b6ee sort local interfaces before selecting one 2018-09-20 09:21:09 -07:00
Greg Fitzgerald
270fd6d61c Fix compiler warnings 2018-09-20 09:47:36 -06:00
carllin
7b9c7d4150 Cleaned up find_leader_rotation function. Added testing for WriteStage find_leader_rotation_index() function () 2018-09-19 18:16:00 -07:00
Carl
55126f5fb6 Marked Tvu functionality in Fullnode as unused for now 2018-09-19 16:05:31 -07:00
Anatoly Yakovenko
431692d9d0 Use a Drop trait to keep track of lifetimes for recycled objects.
* Move recycler instances to the point of allocation
* sinks no longer need to call `recycle`
* Remove the recycler arguments from all the apis that no longer need them
2018-09-19 16:59:42 -06:00
Michael Vines
6732a9078d Clarify AfterTimestamp wire format 2018-09-19 13:28:35 -07:00
sakridge
2981076a14 Add solana-upload-perf to parse json from bench and upload to influx () 2018-09-19 13:16:55 -07:00
Tyera Eulberg
5740ea3807 RFC 006: Wallet CLI 2018-09-19 12:10:53 -06:00
Carl
cd2d50e06c Changed transition to restart Rpu rather than modify bank to prevent lock contention 2018-09-19 10:48:05 -06:00
Carl
8c8a4ba705 debugging commit 2018-09-19 10:48:05 -06:00
Carl
b10de40506 Made LEADER_ROTATION_INTERVAL settable so that integration tests don't time out 2018-09-19 10:48:05 -06:00
Carl
2030dfa435 Implement PR comments, tidy up 2018-09-19 10:48:05 -06:00
Carl
bfe64f5f6e Added integration test for transitioning leader to validator to see that tpu pipeline can exit and restart a tvu. Fixed Tpu and broadcast stage so that exiting later stages in the pipeline also causes earlier stages to exit. 2018-09-19 10:48:05 -06:00
Carl
6d27751365 give fullnode ownership of state needed to dynamically start up a tpu or tvu for role transition 2018-09-19 10:48:05 -06:00
Carl
1fb1c0a681 added jointypes to the stages in the tpu involved in leader rotation 2018-09-19 10:48:05 -06:00
Carl
062f654fe0 formatted code 2018-09-19 10:48:05 -06:00
Carl
d3cb161c36 Added broadcast stage test for leader rotation exit 2018-09-19 10:48:05 -06:00
Carl
98b47d2540 Added check in broadcast stage to exit after transmitting last blob before leader rotation. Also added tests 2018-09-19 10:48:05 -06:00
Carl
f28ba3937b Added check in write stage to exit when scheduled entry_height for leader rotation is detected 2018-09-19 10:48:05 -06:00
Carl
91cf14e641 Rewrote service trait join() method to allow thread join handles to return values other than () 2018-09-19 10:48:05 -06:00
dependabot[bot]
7601a8001c Update reqwest requirement from 0.8.6 to 0.9.0
Updates the requirements on [reqwest](https://github.com/seanmonstar/reqwest) to permit the latest version.
- [Release notes](https://github.com/seanmonstar/reqwest/releases)
- [Changelog](https://github.com/seanmonstar/reqwest/blob/master/CHANGELOG.md)
- [Commits](https://github.com/seanmonstar/reqwest/commits/v0.9.0)

Signed-off-by: dependabot[bot] <support@dependabot.com>
2018-09-19 10:47:02 -06:00
Pankaj Garg
0ee6c5bf9d Read multiple entries in write stage ()
- Also use rayon to parallelize to_blobs() to maximize CPU usage
2018-09-18 21:45:49 -07:00
Michael Vines
6dee632d67 Remove Signature from ApplySignature 2018-09-18 20:00:42 -07:00
Michael Vines
51e5de4d97 Log specific send_transaction error messages 2018-09-18 16:17:08 -07:00
Michael Vines
1f08b22c8e Tweak log messages 2018-09-18 16:17:08 -07:00
Michael Vines
83ae5bcee2 Detect binary changes in serialized contract userdata 2018-09-18 16:17:08 -07:00
Michael Vines
339a570b26 Update comment 2018-09-18 16:17:08 -07:00
Pankaj Garg
5310b6e5a2 Move entry->blob creation out of write stage ()
- The write stage will output vector of entries
- Broadcast stage will create blobs out of the entries
- Helps reduce MIPS requirements for write stage
2018-09-18 13:49:10 -07:00
Pankaj Garg
7d14f44a7c Move register_entry_id() call out of write stage ()
* Move register_entry_id() call out of write stage

- Write stage is MIPS intensive and has become a bottleneck for
  TPU pipeline
- This will reduce the MIPS requirements for the stage

* Fix rust format issues
2018-09-18 11:42:25 -07:00
Rob Walker
c830eeeae4 Update RELEASE.md 2018-09-18 10:31:26 -07:00
Rob Walker
157fcf1de5 initial RELEASE.md ()
initial RELEASE.md and RELEASE_TEMPLATE.md
2018-09-18 10:23:15 -07:00
Michael Vines
e050160ce5 Use tagged perf-libs to enable controlled updates 2018-09-18 09:21:44 -07:00
Tyera Eulberg
f273351789 Add missing port number 2018-09-18 09:36:54 -06:00
Michael Vines
aebf7f88e5 Various spelling fixes 2018-09-17 19:37:59 -07:00
Michael Vines
aac1571670 mint now uses the SystemContract instead of Budget 2018-09-17 18:02:40 -07:00
Anatoly Yakovenko
8bae75a8a6 system contract tests 2018-09-17 14:34:55 -07:00
Tyera Eulberg
c2f7ca9d8f Change process_command return type and improve test 2018-09-17 13:45:47 -07:00
anatoly yakovenko
6ec0e42220 budget as separate contract and system call contract ()
* budget and system contracts and verification

* contract check_id methods
* system call contract
* verify contract execution rules
* move system into its own file
* allocate before transfer for budget
* store error in budget context
* budget contract and tests without bank
* moved budget of of bank
2018-09-17 13:36:31 -07:00
sakridge
072b244575 Add perf counters for record/write stages () 2018-09-17 11:07:04 -07:00
Tyera Eulberg
7ac9d6c604 Create keygen helper function for use in Wallet CLI, print keypair statement 2018-09-17 11:53:33 -06:00
Tyera Eulberg
0125163190 Remove wallet.sh, update entrypoint syntax for wallet network argument 2018-09-17 11:53:33 -06:00
Tyera Eulberg
a06f4b1d44 Update wallet to trigger keygen if no keypair provided and no keypair found in default location 2018-09-17 11:53:33 -06:00
Tyera Eulberg
10daa015c4 Simplify timeout arg 2018-09-17 11:53:33 -06:00
Tyera Eulberg
0babee39a4 Update wallet to take network arg 2018-09-17 11:53:33 -06:00
Michael Vines
7c08b397eb Update testnet documentation 2018-09-17 09:26:25 -07:00
Michael Vines
155ee8792f Add GPU support to ec2-provider 2018-09-17 09:26:25 -07:00
Michael Vines
f89f121d2b Add AWS EC2 support 2018-09-17 09:26:25 -07:00
Tyera Eulberg
27986d7abb Standardize CLI help text 2018-09-16 15:17:10 -06:00
Tyera Eulberg
8b7edc6d64 Alphabetize 2018-09-16 15:17:10 -06:00
Michael Vines
7dfab867fe Mark --outfile parameter as required 2018-09-16 10:49:02 -07:00
Anatoly Yakovenko
fd36954477 clippy 2018-09-15 05:12:53 -06:00
Jay Kickliter
fd51599fa8 Replace replace(..., None) with take()
This is strictly for simplicity, since Option::take() is imlemented with replace().
2018-09-15 05:12:09 -06:00
Michael Vines
3ca80c676c Disable large-network until it's fixed 2018-09-14 20:13:17 -07:00
Pankaj Garg
be7cce1fd2 Tweak GCE scripts for higher node count ()
* Tweak GCE scripts for higher node count

- Some validators were unable to rsync config from leader when
  the node count was high (e.g. 25). Looks like the leader node was
  getting more rsync requests in parallel than it count handle.
- This change staggers the validators bootup, and rsync time

* Address review comments
2018-09-14 17:17:08 -07:00
Pankaj Garg
e142aafca9 Use multiple sockets for receiving blobs on validators ()
* Use multiple sockets for receiving blobs on validators

- The blobs that are broadcasted by leader or retransmitted by peer
  validators are received on replicate_port
- Using reuse_addr/reuse_port, multiple sockets can be opened for
  the same port
- This allows the kernel to queue data to user space app on multiple
  socket queues, preventing over-running one queue
- This helps with reducing packets dropped due to queue over-runs

Fixes 

* Fixed failing tests
2018-09-14 16:56:06 -07:00
Michael Vines
4196cf43e8 cargo fmt 2018-09-14 16:37:49 -07:00
Michael Vines
a344eb7dd0 Upgrade rust stable to 1.29 2018-09-14 16:37:49 -07:00
Pankaj Garg
d12537bdb7 Include UDP sent statistics in net stats () 2018-09-14 13:32:13 -07:00
Tyera Eulberg
bcb3b3c21f Add integration tests to wallet module 2018-09-14 08:21:33 -06:00
Tyera Eulberg
d8c9a1aae9 Add method to run local drone for tests 2018-09-14 08:21:33 -06:00
Tyera Eulberg
9ca2f5b3f7 Move all handling except network/gossip from /bin to wallet module 2018-09-14 08:21:33 -06:00
Rob Walker
9e24775051 update README with v0.8 and update demo scripts to match 2018-09-13 18:37:37 -07:00
sakridge
4dc30ea104 Add recycler stats () 2018-09-13 14:49:48 -07:00
Pankaj Garg
90df6237c6 Implements recvmmsg() for UDP packets ()
* Implemented recvmmsg() for UDP packets

- This change implements binding between libc API for recvmmsg()
- The function can receive multiple packets using one system call

Fixes 

* Added unit tests for recvmmsg()

* Added recv_mmsg() wrapper for non Linux OS

* Address review comments for recvmmsg()

* Remove unnecessary imports

* Moved target specific dependencies to the function
2018-09-13 14:41:28 -07:00
Rob Walker
80caa8fdce add back some defaults for client.sh 2018-09-13 14:05:53 -07:00
carllin
8706774ea7 Rewrote service trait join() method to allow thread join handles to return values other than () () 2018-09-13 14:00:17 -07:00
Pankaj Garg
1d7e87d430 Increase number of sockets for transaction processing 2018-09-13 14:22:07 -06:00
sakridge
1a4cd763f8 Fix missing recycle in recv_from ()
In the error case that i>0 (we have blobs to send)
we break out of the loop and do not push the allocated r
to the v array. We should recycle this blob, otherwise it
will be dropped.
2018-09-13 08:29:18 -07:00
Michael Vines
ee74b367ce Add docker install script 2018-09-12 17:09:37 -07:00
Michael Vines
f06113500d bench-tps/net sanity: add ability to check for unexpected extra nodes 2018-09-12 15:38:57 -07:00
Rob Walker
9ab5692acf fix "leak" in Blob::recv_from ()
* fix "leak" in Blob::recv_from

fixes 
2018-09-12 14:45:43 -07:00
Rob Walker
e7a910b664 v0.9 2018-09-12 10:27:33 -07:00
Rob Walker
b52230097e groom Fullnode's new_with_bank() to match new() more 2018-09-12 09:24:42 -07:00
Rob Walker
a8fdb8a5a7 use a single BlobRecycler per fullnode 2018-09-11 16:56:54 -07:00
sakridge
297f859631 Change '>=' back to '>' to fix recycling of blobs/packets ()
Recycler will have a strong ref to the item so it will be at
least 1, >= will always prevent recycling.
2018-09-11 16:52:45 -07:00
Pankaj Garg
5d19b799af Fix snap configuration for netstat daemon ()
- Also increased the frequency at which the stats are sent
- Fixed file permissions for snapcraft.yaml
2018-09-11 14:49:05 -07:00
Michael Vines
af3eb5a16c .sh 2018-09-11 11:29:49 -07:00
Rob Walker
b313b7f6f9 Revert "move rpc_server to drop() semantics instead of having its own thread"
This reverts commit 40aa0654fa.
2018-09-10 22:48:33 -07:00
Rob Walker
016ee36808 remove -x 2018-09-10 21:40:14 -07:00
Rob Walker
c3fc98c48f use gossip to find the leader for every airdrop request 2018-09-10 21:29:45 -07:00
Rob Walker
40aa0654fa move rpc_server to drop() semantics instead of having its own thread 2018-09-10 20:25:53 -07:00
Michael Vines
bace2880d0 Correct spelling 2018-09-10 19:58:21 -07:00
sakridge
9d80eefb81 Log the number of accounts each 250k txes () 2018-09-10 17:40:00 -07:00
Pankaj Garg
1c17c6dd2b Report UDP network statistics ()
* Report UDP network statistics

Fixes 

* Address review comments

* Address additional review comments

* Fix shellcheck errors
2018-09-10 15:52:08 -07:00
Michael Vines
2be0dbddbb Correct spelling 2018-09-10 13:48:43 -07:00
Rob Walker
a91b785ba5 move fullnode trace generation into crdt 2018-09-10 13:47:57 -07:00
Tyera Eulberg
0ef05de889 Add sleep to prevent spinning thread 2018-09-10 12:50:28 -07:00
Greg Fitzgerald
a093d5c809 Fix erasure build 2018-09-10 11:40:26 -06:00
Greg Fitzgerald
fc64e1853c Initialize Window, not SharedWindow
Wrap with Arc<RwLock>> when/if needed, no earlier.
2018-09-10 11:40:26 -06:00
Greg Fitzgerald
7f669094de Split window into two modules 2018-09-10 11:40:26 -06:00
Greg Fitzgerald
5025d89c88 Inline window method implementations 2018-09-10 11:40:26 -06:00
Greg Fitzgerald
2b44c4504a Use WindowUtil for more idiomatic code 2018-09-10 11:40:26 -06:00
Greg Fitzgerald
d2c9beb843 Add a trait to pretend Window is an object 2018-09-10 11:40:26 -06:00
Michael Vines
9e6d3bf532 Correct spelling 2018-09-10 09:29:01 -07:00
anatoly yakovenko
a89b611e9e comments () 2018-09-09 07:07:38 -07:00
Michael Vines
ebcac3c2d1 Use a common solana user on all testnet instances 2018-09-08 22:34:26 -07:00
Michael Vines
7029e4395c Fix OOM reporting 2018-09-08 18:57:31 -07:00
Michael Vines
5afcdcbbe6 More log grooming 2018-09-08 14:16:34 -07:00
Michael Vines
3840b4b516 Groom log output 2018-09-08 14:10:18 -07:00
Michael Vines
7aeb6d642b Display log file 2018-09-08 13:59:45 -07:00
Michael Vines
1d6c4aacae Retry rsync a couple times before failing 2018-09-08 13:59:45 -07:00
Michael Vines
9f5c86e60c Install earlyoom at gce instance startup 2018-09-08 13:59:45 -07:00
Michael Vines
9f413fd656 Establish net/scripts/... for better scoping 2018-09-08 13:59:45 -07:00
Rob Walker
97c3125a78 improve localnet-sanity's robustness ()
* fix poll_gossip_for_leader() loop to actually wait
         for 30 seconds
    * reduce reuseaddr use to only when necessary,
         try to avoid already bound sockets
    * move nat.rs to netutil.rs
    * add gossip tracing to thin_client and bench-tps
2018-09-09 04:50:43 +09:00
Michael Vines
a77aca75b2 Add NO_VALIDATOR_SANITY back 2018-09-07 22:37:05 -07:00
Rob Walker
96bfd9478b make all the nodes have a pretty seq id () 2018-09-08 14:18:18 +09:00
Michael Vines
e8206cb2d4 Echo the network address before entering a quiet polling loop 2018-09-07 21:20:00 -07:00
Michael Vines
c3af0d9d25 Improve client.log 2018-09-07 21:20:00 -07:00
Michael Vines
932c994dc9 Use new bench-tps command-line args 2018-09-07 21:20:00 -07:00
anatoly yakovenko
c34d911eaf Migrate Budget DSL to use the Account state ()
* Migrate Budget DSL to use the Account state instead of global bank data structures.

* Serialize Instruction into Transaction::userdata.
* Store the pending set in the Account::userdata
* Enforce the token balance rules on contract execution. This becomes the entry point for generic contracts.
* This pr will have a performance impact on the bank. The next set of changes will fix this by locking each account during multi threaded execution of all the contracts.
* With this change a contract transaction needs to store its state under an address. That address could be the destination of the tokens, or any random address. For the latter, an extra step would be needed to claim the tokens which isn't implemented by budget_dsl at the moment.
* test tracking issue 1157
2018-09-07 20:18:36 -07:00
Michael Vines
ddd1871840 Install libssl1.1 for solanalabs/rust docker image compat 2018-09-07 19:57:41 -07:00
Michael Vines
db825788fa Document how to get ssh access into CD testnets 2018-09-07 19:41:13 -07:00
Michael Vines
b1b03ec13b Refine docker image tagging to avoid breaking stabilization branches on updates 2018-09-07 18:42:25 -07:00
Michael Vines
73a8441add /var/snap is not writable by most users 2018-09-07 17:41:20 -07:00
Rob Walker
bf29590f41 WSL needs ReuseAddr in addition to ReusePort (which it doesn't honor) () 2018-09-08 07:28:22 +09:00
Rob Walker
51b27779c9 client changes for TODOs and looping ()
* remove client.sh from snap
* default to ephemeral instead of ~/.config key
* rework CLI for bench-tps
* remote multinode-demo stuff from remote-client.sh
* remove multinode-demo from remote-sanity and localnet-sanity
2018-09-08 07:07:10 +09:00
Tyera Eulberg
5169c8d08f Add method to return hash of bank state 2018-09-07 15:38:53 -06:00
Michael Vines
0d945e6a92 Groom testnet-sanity logging 2018-09-07 12:45:48 -07:00
Michael Vines
1090254ba5 Add datapoints for leader/validator start 2018-09-07 12:45:48 -07:00
Michael Vines
e51445d857 🙃 2018-09-07 12:24:34 -07:00
Michael Vines
4b47abd3bf Fix --num-nodes argument parsing 2018-09-07 12:20:42 -07:00
Greg Fitzgerald
71a617b4dc Fix erasure build 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
a722802c95 Window write lock to read lock 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
e9f44b6661 window -> window_service 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
9693de1867 Reposition parameters 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
f7ea95aed1 Hoist lock, reposition parameters 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
f07ce59be8 Toggle parameters 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
da423b6cf0 Hoist read lock 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
d5f60b68e4 Hoist window write lock 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
78b3a8f7f9 Hoist repair_window() branches
This probably would have been done if repair_window() was unit-tested.
2018-09-07 13:18:19 -06:00
Greg Fitzgerald
d77699c126 Do the easy check first
All functions above operate on immutable values, so this shouldn't
change functionality, but no repair_window() tests to be certain.hI
2018-09-07 13:18:19 -06:00
Greg Fitzgerald
09ba0dae15 Remove redundant clone() 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
a5c7575207 Rewrite find_next_missing, call it clear_slots 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
50f040530b Remove redundant cast 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
7f99c90539 Simplify using early return and Result::ok() 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
d8564b725c Don't reference window to get each slot 2018-09-07 13:18:19 -06:00
Greg Fitzgerald
e4de25442a Hoist write lock
It needed to be passed the lock before, because it contained a
branch where one side didn't require locking. Now that that
defensive programming was hoisted, we can hoist the write lock
as well, leaving a simpler function for unit testing.
2018-09-07 13:18:19 -06:00
Greg Fitzgerald
3b2ea8fd40 Hoist untested branch in window
If there were unit tests for this function, the author would have
written it this way to make their own life easier.
2018-09-07 13:18:19 -06:00
Michael Vines
9a1832ed61 Bump ping timeout 2018-09-07 12:01:43 -07:00
Michael Vines
9e45f1f5e2 Doc fixup 2018-09-07 12:01:43 -07:00
Michael Vines
ee682d5bc3 Move wallet-sanity.sh out of multinode-demo/ 2018-09-07 12:01:43 -07:00
Michael Vines
05decc863f Make set -x more buildkite friendly 2018-09-07 12:01:43 -07:00
Michael Vines
506a81e8cc Assume -y 2018-09-07 12:01:43 -07:00
Michael Vines
dcb30a8489 Delete leader node first 2018-09-07 12:01:43 -07:00
Michael Vines
a2631e89f6 Use consistent style 2018-09-07 12:01:43 -07:00
Michael Vines
ab208ddb77 Clean up arg handling 2018-09-07 12:01:43 -07:00
Michael Vines
09a48d773a Run bench-tps in a tmux 2018-09-07 12:01:43 -07:00
Michael Vines
88298bf321 Add -n option 2018-09-07 12:01:43 -07:00
Michael Vines
d252f7f687 Revert "Default to 10 validators"
This reverts commit ed5fbaef06.
2018-09-07 12:01:43 -07:00
Michael Vines
533ebc17f2 Install multilog automatically on a CI machine 2018-09-07 11:56:23 -07:00
Michael Vines
f4947236dc Keep cargo-target-cache size under 6GB-ish 2018-09-07 11:45:27 -07:00
Michael Vines
e088833b81 s/create/start/ 2018-09-06 21:07:11 -07:00
Michael Vines
53e16f68d9 Improve error handling 2018-09-06 20:57:05 -07:00
Michael Vines
ed5fbaef06 Default to 10 validators 2018-09-06 20:46:49 -07:00
Michael Vines
b1bacf12a6 Add some log sections 2018-09-06 20:38:11 -07:00
Michael Vines
66ff602659 Rewrite ci/testnet-{deploy,sanity}.sh in terms of net/ primitives 2018-09-06 19:54:39 -07:00
Michael Vines
e175c9dea9 Remove ip address hardcode. Fixes 2018-09-06 19:54:39 -07:00
Michael Vines
5a57d9b5d9 de-y 2018-09-06 19:54:39 -07:00
Michael Vines
03e87e4169 Add more metrics 2018-09-06 19:54:39 -07:00
Michael Vines
abfff66d53 Retry ssh a couple times before giving up 2018-09-06 19:54:39 -07:00
Michael Vines
31dee553d5 Split start/version reporting 2018-09-06 19:54:39 -07:00
Michael Vines
9ca6a2d25b Configure boot disk size 2018-09-06 19:54:39 -07:00
Michael Vines
a3178c3bc7 Remove unused name tag 2018-09-06 19:54:39 -07:00
Michael Vines
aa07bdfbaa Optionally suppress delete confirmation 2018-09-06 19:54:39 -07:00
Michael Vines
eaef9be710 Clarify -f 2018-09-06 19:54:39 -07:00
Michael Vines
cae345b416 Allow - in prefix 2018-09-06 19:54:39 -07:00
Michael Vines
acb1171422 Add -e option 2018-09-06 19:54:39 -07:00
Greg Fitzgerald
52d8f293b6 Add links to citations
And fix hyphens in quote.
2018-09-06 20:41:05 -06:00
Greg Fitzgerald
636eb8d058 Add Leslie Lamport quote to README 2018-09-06 20:41:05 -06:00
Greg Fitzgerald
0fa27f65bb Use the default Pubkey formatter instead of debug_id() 2018-09-06 16:31:47 -06:00
sakridge
8f94e3f7ae Buffer tokens when switching directions to prevent errors ()
Even if transactions are dropped, accounts will have buffer
of tokens. Should reduce or eliminate AccountNotFound errors seen in the
leader while bench-tps is running.
2018-09-06 14:20:01 -07:00
Pankaj Garg
05460eec0d Open multiple sockets for transaction UDP port ()
* Reuse UDP port and open multiple sockets for transaction address

* Fixed failing crdt tests

* Add tests for reusing UDP ports

* Address review comments

* Updated bench-streamer to use multiple receive sockets

* Fix minimum number of recv sockets for bench-streamer

* Address review comments

Fixes 

* Moved bind_to function to nat.rs
2018-09-06 14:13:40 -07:00
Michael Vines
072d0b67e4 Send deploy metrics to the testnet-specific database 2018-09-06 08:30:03 -07:00
Rob Walker
fdc48d521c use USER instead of whoami ()
* use USER instead of whoami

make gcloud_FigureRemoteUsername robust against unsolicited output
   (that I get on login ;) )

validate --prefix argument

* Update gcloud.sh
2018-09-07 00:18:05 +09:00
Michael Vines
6560b0e2cc s/whoami/id -un/ 2018-09-05 14:26:21 -07:00
Michael Vines
ec38dba209 GCE leader nodes can now be provisioned with a static IP address 2018-09-05 14:26:21 -07:00
sakridge
d9e4bce6ad Add drop stats to bench-tps ()
See how many transactions made it through
2018-09-05 11:58:41 -07:00
sakridge
1fd4343621 Add total count to stat () 2018-09-05 09:28:18 -07:00
Michael Vines
8d87627a49 t 2018-09-05 09:09:50 -07:00
Michael Vines
aacf27fb76 Add convienience link to current Snap log files 2018-09-05 09:02:02 -07:00
Michael Vines
a51536d107 Add log tail hint 2018-09-05 09:02:02 -07:00
Michael Vines
1c874fbc1b Make This is little more hacky 2018-09-05 09:02:02 -07:00
Michael Vines
0362169671 Better scope leader and validator setup 2018-09-05 09:02:02 -07:00
Michael Vines
e2e569cb43 Set rsync url for local deployments 2018-09-05 09:02:02 -07:00
Michael Vines
8c51b47e85 Preserve existing ssh config 2018-09-05 09:02:02 -07:00
Michael Vines
017eb10e76 Add file header doc 2018-09-05 09:02:02 -07:00
Michael Vines
f50aeb0e58 Always add perf-libs to LD_LIBRARY_PATH 2018-09-05 09:02:02 -07:00
Michael Vines
48c19d3100 Enable cargo features to be specified 2018-09-05 09:02:02 -07:00
Michael Vines
aaf0a23134 Add Tips section 2018-09-05 09:02:02 -07:00
Michael Vines
89db85dbf9 Work around concurrent |gcloud compute ssh| terminal issue 2018-09-05 09:02:02 -07:00
Michael Vines
e677cda027 Private IP networks now work, and are the default 2018-09-05 09:02:02 -07:00
Michael Vines
db9219ccc8 Improve error monitoring 2018-09-05 09:02:02 -07:00
Michael Vines
06fd945f85 Set node config correctly 2018-09-05 09:02:02 -07:00
Michael Vines
6ad4a81123 s/_/-/g in filenames 2018-09-05 09:02:02 -07:00
Michael Vines
bcaa0fdcb1 net/ can now deploy Snaps 2018-09-05 09:02:02 -07:00
Michael Vines
2cb1375217 Run gcloud_PrepInstancesForSsh in parallel 2018-09-05 09:02:02 -07:00
Michael Vines
9365a47d42 Employ a startup script 2018-09-05 09:02:02 -07:00
Michael Vines
6ffe205447 Add -g option 2018-09-05 09:02:02 -07:00
Michael Vines
ec3e62dd58 Add net/ sanity 2018-09-05 09:02:02 -07:00
Michael Vines
fa07c49cc9 net/ can now deploy Snaps 2018-09-05 09:02:02 -07:00
Michael Vines
449d7042f0 Configure metrics correctly 2018-09-05 09:02:02 -07:00
Michael Vines
7e2b65374d gce instance types are now configurable 2018-09-05 09:02:02 -07:00
Michael Vines
8e39465700 Drop .sh extension to hide from shellcheck 2018-09-05 09:02:02 -07:00
Michael Vines
43b4207101 Run oom-monitor in net/ testnets 2018-09-05 09:02:02 -07:00
Michael Vines
ff991b87da Add support for deploying from non-Linux machines 2018-09-05 09:02:02 -07:00
Michael Vines
c81c19234f Improve incremental speed of docker cargo builds outside of CI 2018-09-05 09:02:02 -07:00
Michael Vines
399caf343c Morph gce_multinode-based scripts into net/ 2018-09-05 09:02:02 -07:00
sakridge
ffb72136c8 Remove account from balances table after error seen ()
If balance goes to 0, then bank removes the account
from it's account table and returns no account error. Thin client
should also update the account to this state or it will
still have the cached balance from the last successful get_balance().
2018-09-04 21:33:19 -07:00
jackcmay
1a615bde2b Update README.md ()
* Update README.md

* Fix spelling

* Improved punctuation
2018-09-04 20:41:11 -07:00
Greg Fitzgerald
cf2626a1c5 Update instructions to upgrade nightly docker image 2018-09-04 20:56:40 -06:00
Greg Fitzgerald
68c72d6f34 Fix nightly build 2018-09-04 20:56:40 -06:00
Greg Fitzgerald
65f78905cd Install cargo-cov on latest nightly 2018-09-04 20:56:40 -06:00
Pankaj Garg
70a8ae4612 Fixed private IP variable in gcloud script () 2018-09-04 16:24:19 -07:00
sakridge
d82ec2634c Fix is_leader boolean ()
A node is the leader if the address is none
2018-09-04 13:38:24 -07:00
Michael Vines
b4a7a18334 Update README.md 2018-09-04 13:29:00 -07:00
Rob Walker
c44c5f0b09 take into account size of an Entry () 2018-09-05 05:07:58 +09:00
Rob Walker
226d3b9471 Trace recycle() calls ()
* trace recycle() calls fixes 
2018-09-05 05:07:02 +09:00
Stephen Akridge
2752bde683 Print to indicate what drone is doing while waiting for gossip 2018-09-04 13:45:08 -06:00
Greg Fitzgerald
b8816d722c Fix Block::to_blobs() benchmark
16% speedup, w00t!

name                                control  ns/iter  variable  ns/iter  diff ns/iter   diff %  speedup
bench_block_to_blobs_to_block       29,897            25,807                   -4,090  -13.68%   x 1.16
2018-09-04 07:50:23 -10:00
Greg Fitzgerald
2aa72cc72e Return a Vec from to_blobs() instead of using a mut parameter 2018-09-04 07:50:23 -10:00
Greg Fitzgerald
8cc030ef84 Use Vec instead of VecDeque for SharedBlobs 2018-09-04 07:50:23 -10:00
Stephen Akridge
9a9f89293a Better error handling messages for airdrops 2018-09-04 06:46:43 -10:00
anatoly yakovenko
501deeef56 accounts should never be negative () 2018-09-04 06:43:18 -10:00
Greg Fitzgerald
05f921d544 Don't call println in the test suite 2018-09-04 06:01:32 -10:00
Greg Fitzgerald
ab7a2960b1 Don't use product name in solana library 2018-09-04 06:01:32 -10:00
Greg Fitzgerald
4e2deaa33b Less mut 2018-09-04 06:01:32 -10:00
Greg Fitzgerald
d5ef18337c Remove redundant return value
And don't log the same error twice.
2018-09-04 06:01:32 -10:00
Greg Fitzgerald
d18ea501b7 Minimize unsafe code 2018-09-04 06:01:32 -10:00
Greg Fitzgerald
c9a1ac9b8c Don't propogate errors we'll never handle 2018-09-04 06:01:32 -10:00
Greg Fitzgerald
c2a4cb544e Borrow, don't clone entries 2018-09-04 06:01:32 -10:00
Greg Fitzgerald
3ab12076e8 Convert voting functions to methods
More idiomatic Rust.
2018-09-04 05:53:58 -10:00
Tyera Eulberg
6a383c45fc Update sendTransaction example to reflect new array size 2018-09-04 05:44:10 -10:00
Tyera Eulberg
7cc27e7bd1 Doc requestAirdrop rpc method 2018-09-04 05:44:10 -10:00
Tyera Eulberg
0464087327 Add api definitions 2018-09-04 05:44:10 -10:00
Tyera Eulberg
c193c7de12 Add JSON-RPC API Documentation 2018-09-04 05:44:10 -10:00
Rob Walker
61abee204f don't check for snap mode in common.sh, is only relevant to snap daemons ()
snap mode is for daemons, remove it from client (i.e. common.sh)

supply leader info to client via snap
2018-09-04 14:31:54 +09:00
Rob Walker
a99dbb2a0c set -x in client.sh 2018-09-04 11:55:04 +09:00
Rob Walker
e834c76b40 --count => --num-nodes 2018-09-04 07:07:25 +09:00
Rob Walker
7b3c7f148b supply leader and leader_address 2018-09-02 02:27:05 +09:00
anatoly yakovenko
fb4b33b81b make the repair_backoff test more robust ()
* more the repair_backoff test more robust

* fix names and magic numbers
2018-08-31 12:40:56 -10:00
Rob Walker
25d7dc7b96 fixups 2018-09-01 04:38:18 +09:00
Rob Walker
d1f1cbe88f leader-address=>leader-ip 2018-09-01 04:38:18 +09:00
Rob Walker
a4e7b6e90c more fixups for client.sh changes 2018-09-01 03:33:21 +09:00
Rob Walker
fbc7c9c431 fix client_start to deal with new client.sh 2018-09-01 03:23:05 +09:00
Rob Walker
8b248dcf09 specify port 2018-09-01 02:56:24 +09:00
Rob Walker
4938aad939 fixups 2018-09-01 02:21:46 +09:00
Rob Walker
7e882dfe62 inform all snaps where the network is 2018-09-01 02:21:46 +09:00
Rob Walker
5c8cb96f88 rebase fixup 2018-08-31 23:21:07 +09:00
Rob Walker
9d1eb4f9ea remove 'localhost' leader (redundant, un-dig-friendly) 2018-08-31 23:21:07 +09:00
Rob Walker
210a4d0640 fixup 2018-08-31 23:21:07 +09:00
Rob Walker
176e806d94 rework of netwrk rendezvous
* rename NodeInfo field of Node from "data" to "info"
      (touches a lot of files)

  * update client to use gossip to find leader, a la drone

  * rework multinode scripts
      * move more stuff into rust
      * added usage to all
      * no more rsync unless you're a validator (TODO: whack that, too)
  * fullnode doesn't bail if drone isn't up yet, just keeps trying
  * drone doesn't bail if network isn't up yet, just keeps trying
2018-08-31 23:21:07 +09:00
Rob Walker
eb4e5a7bd0 fixups 2018-08-31 23:21:07 +09:00
Rob Walker
ba27596076 fixups 2018-08-31 23:21:07 +09:00
Rob Walker
63e44dcc35 continue rendezvous refactor for gossip and repair
* remove trailing whitespace in ci/audit.sh

  * code review fixups
     * rename GOSSIP_PORT_RANGE => SOLANA_PORT_RANGE
     * remove out-of-date TODO in localnet-sanity.sh

  * remove features=test and code that was using it (localhost prohibitions in
      crdt) added TODO in crdt.rs, maybe we should boot localhost in production
      networks?

  * boot tvu_window from NodeInfo: instead, send repair requests from the repair
      socket (to gossip on peer) and answer repair requests via the sockaddr
      from the repair request

  * remove various unused pub functions

  * banish SocketAddr parse().unwrap() to a macro that can also accept simpler stuff
2018-08-31 23:21:07 +09:00
Rob Walker
c0ba676658 fixup 2018-08-31 23:21:07 +09:00
Rob Walker
1af4cee63b fix
* move gossip/NCP off assuming anything about its address
  * use a single socket to send and receive gossip
  * remove --addr/-a from CLIs
  * rearrange networking utility code
  * use Arc<UdpSocket> to share the Sync-safe UdpSocket among threads
  * rename TestNode to Node

TODO:

  * re-enable 127.0.0.1 as a valid address in crdt
  * change repair request/response to a similar, single socket
  * pick cloned sockets or Arc<UdpSocket> for all these (rpu uses tryclone())
  * update contact_info with network truthiness instead of what the node
      says?
2018-08-31 23:21:07 +09:00
Rob Walker
cb52a335bd re-enable localnet-sanity 2018-08-31 23:21:07 +09:00
Tyera Eulberg
e308a4279e Update RPC requestAirdrop endpoint to return airdrop tx signature 2018-08-28 18:27:41 -06:00
Tyera Eulberg
513a934ff6 Update request_airdrop utility function to pass along airdrop tx signature 2018-08-28 18:27:41 -06:00
Tyera Eulberg
77d820c842 Update drone module to return airdrop tx signature 2018-08-28 18:27:41 -06:00
Tyera Eulberg
30cbe7c6a9 Update jsonrpc crate version 2018-08-28 18:27:24 -06:00
Michael Vines
18ef643dc7 Keep locals local 2018-08-28 08:11:44 -07:00
Michael Vines
73a0bf8d30 Avoid unbounded /var/tmp growth 2018-08-28 08:11:44 -07:00
Michael Vines
9d53208d68 Use gcloud_DeleteInstances 2018-08-28 08:11:44 -07:00
Michael Vines
d26f135159 Find metrics-write-datapoint.sh again 2018-08-27 22:41:58 -07:00
Michael Vines
c8e3ce26a9 Start of scripts/gcloud.sh 2018-08-27 22:35:14 -07:00
Rob Walker
f88970a964 source oom-score-adj.sh from validator.sh 2018-08-28 10:01:41 +09:00
Rob Walker
51d911e3f4 Update testnet-sanity.sh 2018-08-27 15:44:10 -07:00
Michael Vines
bd5c6158ae Move some common scripts from multinode-demo/ to scripts/ 2018-08-27 13:52:38 -07:00
Michael Vines
cd0db7842c Remove unused _config.yml 2018-08-27 13:52:38 -07:00
Michael Vines
31d1087103 Documentation 2018-08-27 13:52:38 -07:00
Rob Walker
0efd64df6f no need for sudo, move ledger copy out of SNAP_DATA 2018-08-28 05:42:05 +09:00
Rob Walker
28bdf346f6 clean up after ledger sanity 2018-08-28 05:42:05 +09:00
anatoly yakovenko
48762834d9 Randomize repair requests ()
* randomize packet repair requests

* exponential random repair requests

* use gen_range to get a uniform distribution
2018-08-27 07:05:48 -07:00
Anatoly Yakovenko
8d0d429acd update 2018-08-26 23:34:25 -07:00
Anatoly Yakovenko
e5408368f7 fmt 2018-08-26 22:35:26 -07:00
Anatoly Yakovenko
61492fd27e exit if no leader 2018-08-26 22:35:26 -07:00
Anatoly Yakovenko
bbce08a67b bench needs to discover leader as well 2018-08-26 22:35:26 -07:00
Anatoly Yakovenko
a002148098 retry transfer and poll 2018-08-26 16:10:46 -07:00
Greg Fitzgerald
90ae662e4d Fix packet header offset
And update transaction offsets to use the same approach as packet.rs.
Maybe this should be serialized_size(), but thanks to this
GenericArray update, those values are the same.
2018-08-26 14:27:19 -06:00
Greg Fitzgerald
60d8f5489f Update transaction layout offsets
24 less bytes in minimal transactions. 10% TPS boost?
2018-08-26 14:27:19 -06:00
dependabot[bot]
59dd8b650d Update generic-array requirement from 0.11.1 to 0.12.0
Updates the requirements on [generic-array](https://github.com/fizyk20/generic-array) to permit the latest version.
- [Release notes](https://github.com/fizyk20/generic-array/releases)
- [Changelog](https://github.com/fizyk20/generic-array/blob/master/CHANGELOG.md)
- [Commits](https://github.com/fizyk20/generic-array/commits)

Signed-off-by: dependabot[bot] <support@dependabot.com>
2018-08-26 14:27:19 -06:00
anatoly yakovenko
738247ad44 advertise valid gossip address in drone and wallet ()
* advertize valid gossip address in drone and wallet

get rid of asserts

check for valid ip address

check for valid address

ip address

* tests

* cleanup

* cleanup

* print error

* bump

* disable tests

* disable nightly
2018-08-26 11:36:27 -07:00
anatoly yakovenko
5b0bb7e607 Skip invalid nodes for finality ()
* skip invalid nodes for finality

* check valid last_ids only

* fixup!

* fixup!
2018-08-25 23:12:41 -07:00
anatoly yakovenko
f7c0d30167 Disallow localhost in deployment ()
* disallow localhost in deployment

* tests

* fmt

* integration tests do not have a flag to check

* fmt
2018-08-25 21:09:18 -07:00
Anatoly Yakovenko
8e98c7c9d6 fix purge test 2018-08-25 19:56:09 -07:00
Pankaj Garg
50661e7b8d Added poll_balance_with_timeout method ()
* Added poll_balance_with_timeout method

- updated bench-tps, fullnode and wallet to use this method instead
  of repeatedly calling poll_get_balance()

* Address review comments

- Revert some changes to use wrapper poll_get_balance()

* Reverting bench-tps to use poll_get_balance

- The original code is checking if the balance has been updated,
  instead of just retrieving the balance. The logic is different
  than poll_balance_with_timeout()

* Reverting wallet to use poll_get_balance

- The break condition in the loop is different than poll_balance_with_timeout().
  It's checking if the balance has been updated.
2018-08-25 18:24:25 -07:00
Pankaj Garg
ad159e0906 Fix crash in fullnode when poll_get_balance() returns error () 2018-08-25 15:25:13 -07:00
Pankaj Garg
d3fac8a06f Dynamically bind to available UDP ports in Fullnode ()
* Dynamically bind to available UDP ports in Fullnode

* Added tests for dynamic port binding

- Also removed hard coding of port range from CRDT
2018-08-25 10:24:16 -07:00
sakridge
c641ba1006 Up network buffers to 64MB max ()
500ms of data at 1Gbps = 125GB/2 = 64MB
Seems to help tx rate in GCP network tests.
2018-08-24 18:17:48 -07:00
sakridge
de379ed915 Fix sig verify counters to be unique and tweak perf counters ()
print events and add current events to old value to report
2018-08-24 16:05:32 -07:00
Michael Vines
d4554c6b78 RFC Branches, Channels, and Tags 2018-08-23 21:28:05 -07:00
sakridge
6fc21a4223 Don't hang in transaction_count ()
Situation is there can be that there can be bad entries in
the bench-tps CRDT table until they get purged later. Threads however
are created for those bad entries and then will hang on trying
to get the transaction_count from those bad addresses and never end.
2018-08-23 20:57:13 -07:00
sakridge
71319978df Up drone request amount ()
Multiple clients will request 500k each so up this to support them.
2018-08-23 15:30:35 -07:00
sakridge
6147e54686 Cap repair requests timeout () 2018-08-23 15:30:21 -07:00
Greg Fitzgerald
0c8eec2563 Cleanup Fullnode construction
leader_id was already set by Fullnode constructor. And cleanup the
rest of that code while in the neighborhood.

Thanks @CriesofCarrots!
2018-08-23 13:42:54 -07:00
Greg Fitzgerald
4ab58f069a Add back JsonRpcService changes 2018-08-23 13:42:54 -07:00
Greg Fitzgerald
85f96d926a Pacify clippy 2018-08-23 13:42:54 -07:00
Greg Fitzgerald
816de4f8ec Hoist shared code between leaders and validators 2018-08-23 13:42:54 -07:00
Greg Fitzgerald
42229a1105 Hoist thread_hdls 2018-08-23 13:42:54 -07:00
Greg Fitzgerald
d8820053af Inline create_leader_threads and create_validator_threads 2018-08-23 13:42:54 -07:00
Greg Fitzgerald
731f8512c6 Hoist Arc<Bank> 2018-08-23 13:42:54 -07:00
Greg Fitzgerald
a133784706 Rename mode-specific constructors and return only thread handles 2018-08-23 13:42:54 -07:00
Greg Fitzgerald
be58fdf1bb Less constructors 2018-08-23 13:42:54 -07:00
Greg Fitzgerald
57daeb35d2 Drop all references to new_leader and new_validator 2018-08-23 13:42:54 -07:00
Greg Fitzgerald
9c5e69bf3d Don't offer two ways to specify a leader 2018-08-23 13:42:54 -07:00
Greg Fitzgerald
cfac127e4c Extract lower-level constructor
Passing in the bank is useful for unit-tests since Fullnode doesn't
store it in a member variable.
2018-08-23 13:42:54 -07:00
Greg Fitzgerald
fda4523cbf Fix broken doc 2018-08-23 13:42:54 -07:00
sakridge
cabe80b129 Increment counter by number of packets received ()
So that we can see the total packets/s
2018-08-23 12:32:50 -07:00
Tyera Eulberg
d4c41219f9 Improve gossip use for drone and wallet
- Add utility function
  - Add thread sleep
  - Enable configurable timeout for gossip poll
2018-08-23 13:08:59 -06:00
Tyera Eulberg
4fdd9fbfca Wallet: use gossip to identify leader's port config 2018-08-23 13:08:59 -06:00
Tyera Eulberg
bdf5ac9c1a Drone: use gossip to identify leader's port config 2018-08-23 13:08:59 -06:00
sakridge
f1785c76a4 Rework counter increment outside apply_debits loop ()
Reduces prints/atomics work inside the process_transactions loop
2018-08-23 09:42:59 -07:00
Tyera Eulberg
2de8fe9c5f Pass bank to rpc as reference 2018-08-23 09:06:17 -06:00
Tyera Eulberg
d910ed68a3 Use balance to verify requestAirdrop success 2018-08-23 09:06:17 -06:00
Tyera Eulberg
f7f7ecd4c6 Add json-rpc requestAirdrop endpoint 2018-08-23 09:06:17 -06:00
Tyera Eulberg
a9c3a28a3b Add json-rpc sendTransaction endpoint 2018-08-23 09:06:17 -06:00
Greg Fitzgerald
96787ff4ac Use builtin sum 2018-08-22 16:24:19 -06:00
Stephen Akridge
c3ed4d28de Change average TPS to max average tps seen for any node and...
add script to collect perf stats
2018-08-22 14:55:04 -07:00
Pankaj Garg
f1e35c3bc6 GCE script change to use GCE private network for multinode tests ()
- Also the user can specify the zone where the nodes should be created
2018-08-22 13:21:33 -07:00
Greg Fitzgerald
db3fb3a27c Boot criterion ()
* Revert benchmarks back to libtest

Criterion has too many dependencies, it's execution as slower, and
we didn't see the kind of precision we had hoped for to use it to
block CI builds.

* Ignore benchmarks that take more than a few milliseconds per iteration

* Revert "Ignore benchmarks that take more than a few milliseconds per iteration"

This reverts commit b87cdf6ef4.

* Don't run benchmarks in CI

They are already built in the nightly build. Executing them in CI
doesn't add much value until the results are precise enough to act
on.
2018-08-22 08:57:07 -06:00
Rob Walker
8282442956 fixes 2018-08-22 17:47:59 +09:00
Tyera Eulberg
a355d9f46c Add error catch for rpc server builder 2018-08-21 14:04:52 -06:00
Tyera Eulberg
be4824c955 Add custom panic hook for RPC port bind 2018-08-21 14:04:52 -06:00
Tyera Eulberg
86c1d97c13 Fix validator rpc addr to match leader 2018-08-20 22:35:06 -07:00
Rob Walker
0b48aea937 echo commands, use PID (good form) 2018-08-21 11:41:00 +09:00
Rob Walker
cdec0cead2 files have to appear in the snap 2018-08-21 11:41:00 +09:00
Rob Walker
831709ce7e fixups 2018-08-21 10:36:03 +09:00
Rob Walker
b7b8a31532 make a copy of the ledger for sanity check
we can't verify a live ledger, unfortunately, fixes 
2018-08-21 10:36:03 +09:00
Michael Vines
15406545d8 Document how to adjust the number of clients or validators on the testnet 2018-08-20 18:35:01 -07:00
Rob Walker
5aced8224f Revert "make a copy of the ledger for sanity check"
This reverts commit af20a43b77.
2018-08-21 10:34:52 +09:00
Rob Walker
af20a43b77 make a copy of the ledger for sanity check
we can't verify a live ledger, unfortunately, fixes 
2018-08-21 09:45:52 +09:00
Greg Fitzgerald
39c3280860 Don't block on large network test 2018-08-20 16:48:37 -06:00
Greg Fitzgerald
2d35345c50 Boot unused creates 2018-08-20 16:48:37 -06:00
Michael Vines
a02910be32 Remove pubkey from getBalance response 2018-08-20 15:02:48 -07:00
Pankaj Garg
b9ec97a30b Add counter for bank transaction errors () 2018-08-20 14:56:01 -07:00
Anatoly Yakovenko
2e89999d88 # This is a combination of 4 commits.
# This is the 1st commit message:

Fix tesetment readme

# This is the commit message :

updte

# This is the commit message :

typo

# This is the commit message :

cleanup
2018-08-20 13:49:56 -07:00
Pankaj Garg
24b0031925 Reduce number of nodes in multinode test () 2018-08-20 13:40:42 -07:00
Michael Vines
9eeaf2d502 Bind RPC port on all interfaces 2018-08-20 12:45:50 -07:00
Michael Vines
c9e6fb36c3 Avoid unncessary cargo rebuilds in non-perf configuration 2018-08-20 12:03:44 -07:00
Michael Vines
8de317113c clippy: remove identity conversion 2018-08-20 10:55:55 -07:00
Michael Vines
a1ec549630 Pin nightly rust for more controlled updating 2018-08-20 10:55:55 -07:00
Michael Vines
ecddff98f5 Add --nopull argument 2018-08-20 10:55:55 -07:00
Michael Vines
10066d67bf Add llvm deb repository 2018-08-19 09:01:36 -07:00
Michael Vines
a07f7435c6 \ 2018-08-19 08:49:29 -07:00
Michael Vines
d3523ebbe5 Nightly image now derives from stable image 2018-08-19 08:47:59 -07:00
Nikhil Saraf
133ddb11ff typo in README 2018-08-18 18:24:42 -07:00
Tyera Eulberg
1bf15ae907 Temporarily disable cargo audit CI failure 2018-08-18 12:29:49 -06:00
Tyera Eulberg
f73f3941cd Revert ill-advised jsonrpc marker, and handle jsonrpc server close 2018-08-18 12:29:49 -06:00
Tyera Eulberg
d69d79612b Simplify Rpc request processing 2018-08-18 12:29:49 -06:00
Tyera Eulberg
64ea5126e0 Fix early return for invalid parameter 2018-08-18 12:29:49 -06:00
Tyera Eulberg
9df3aa50d5 Remove unnecessary solana_ prefixes 2018-08-18 12:29:49 -06:00
Tyera Eulberg
cab75b7829 Handle potential panics 2018-08-18 12:29:49 -06:00
Tyera Eulberg
d9fac86015 Use jsonrpc git repo, allowing removal of Default bound for Metadata 2018-08-18 12:29:49 -06:00
Tyera Eulberg
1eb8724a89 Disable Rpc module for other tests to prevent port conflicts 2018-08-18 12:29:49 -06:00
Tyera Eulberg
c6662a4512 Implement Rpc in Fullnode 2018-08-18 12:29:49 -06:00
Tyera Eulberg
d3c09b4e96 Update jsonrpc dependency syntax 2018-08-18 12:29:49 -06:00
Tyera Eulberg
124f6e83d2 Rpc get last id endpoint 2018-08-18 12:29:49 -06:00
Tyera Eulberg
569ff73b39 Rpc tests 2018-08-18 12:29:49 -06:00
Tyera Eulberg
fc1dbddd93 Implement json-rpc functionality 2018-08-18 12:29:49 -06:00
Rob Walker
3ae867bdd6 fixups 2018-08-18 02:22:52 -07:00
Rob Walker
bc5f29150b fix erasure, remove Entry "pad"
* fixes 
 * Entry pad is no longer required since erasure coding aligns data length
2018-08-18 02:22:52 -07:00
Rob Walker
46016b8c7e crashes generate_coding() 2018-08-18 02:22:52 -07:00
Rob Walker
5dbecd6b6b add logging, more conservative reset 2018-08-18 02:22:52 -07:00
Michael Vines
877920e61b Compute snap channel using ci/channel-info.sh 2018-08-17 23:15:48 -07:00
Michael Vines
3d1e908dad Add script to fetch latest channel info 2018-08-17 23:15:48 -07:00
Michael Vines
6880c2bef0 Exclude ci/semver_bash/; don't want to diverge from upstream 2018-08-17 23:15:48 -07:00
Michael Vines
78872ffb4b Vendor https://github.com/cloudflare/semver_bash/tree/c1133faf0e 2018-08-17 23:15:48 -07:00
Michael Vines
229d825fe0 Fix master-perf basename 2018-08-17 21:59:36 -07:00
Michael Vines
edc5fc098e Make SNAP_CHANNEL more visible in build log 2018-08-17 21:39:54 -07:00
Anatoly Yakovenko
bbe815468d Add instructions on how to run the demo against testnet.solana.com and watch it on the dashboard 2018-08-17 21:26:06 -07:00
Michael Vines
82e7725a42 Invert logic 2018-08-17 21:16:35 -07:00
Michael Vines
dc61cf1c8d Keep v0.7 snap off the edge channel 2018-08-17 21:12:10 -07:00
Michael Vines
aba63e2c6c Log expansion directive must be on its own line 2018-08-17 20:58:14 -07:00
Michael Vines
c2ddd056e2 Add option to skip ledger verification 2018-08-17 20:41:30 -07:00
Michael Vines
c9508e84f2 0.8.0 2018-08-17 17:56:35 -07:00
Pankaj Garg
f6f0900506 Large network test to not poll validator for sigs ()
- The finality is already reached. The test will check the signature
  in validators once, instead of polling. This will help speed up the test.
2018-08-17 14:38:19 -07:00
Rob Walker
7aeef27b99 not quite banishing build.rs, but better 2018-08-16 22:33:31 -07:00
Michael Vines
98d0ef6df5 Add some wget retries 2018-08-16 20:22:49 -07:00
Tyera Eulberg
208a7f16cb Fix bench-tps nokey error 2018-08-16 19:38:26 -06:00
Rob Walker
16cf31c3a3 fix 2018-08-16 15:52:30 -07:00
Rob Walker
2b48daaeba accept multiple expected outputs 2018-08-16 14:44:51 -07:00
Anatoly Yakovenko
79d24ee227 fixed test according to @rob-solana 2018-08-16 14:44:51 -07:00
Anatoly Yakovenko
a284030ecc Account type with state
comments

fixups!

fixups!

fixups for a real Result<> from get_balance()

on 2nd thought, be more rigorous

Merge branch 'rob-solana-accounts_with_state' into accounts_with_state

update

review comments

comments

get rid of option
2018-08-16 14:44:51 -07:00
Rob Walker
fc0d7f5982 updated nightly versions 2018-08-16 13:17:29 -07:00
Rob Walker
f697632edb update clippy install instructions, from here:
https://github.com/rust-lang-nursery/rust-clippy

fixes  ?
2018-08-16 13:17:29 -07:00
Rob Walker
73797c789b back to 4 TX threads 2018-08-16 12:02:11 -07:00
Rob Walker
036fcced31 test -t nproc 2018-08-16 12:02:11 -07:00
Rob Walker
1d3157fb80 fixups 2018-08-16 12:02:11 -07:00
Rob Walker
0b11c2e119 restart testnet clients in case airdrop fails 2018-08-16 12:02:11 -07:00
Stephen Akridge
96af892d95 Add docs about the testnet 2018-08-16 07:39:17 -07:00
Pankaj Garg
c2983f824e Refactored large network test to use finality to assert success () 2018-08-15 20:05:43 -07:00
anatoly yakovenko
88d6fea999 Revert "Accounts with state ()"
This reverts commit c23fa289c3.
2018-08-15 19:44:39 -07:00
anatoly yakovenko
c23fa289c3 Accounts with state ()
* Account type with state

* fixed test according to @rob-solana
2018-08-15 14:32:11 -07:00
Pankaj Garg
db35f220f7 Run multinode test for enough iterations for a small node count test () 2018-08-15 10:44:14 -07:00
Pankaj Garg
982afa87a6 Retransmit blobs from leader from window ()
- Some nodes don't have leader information while leader is broadcasting
  blobs to those nodes. Such blobs are not retransmitted. This change
  rertansmits the blobs once the leader's identity is know.
2018-08-14 21:51:37 -07:00
Rob Walker
dccae18b53 cfg=erasure fixes, use return value of align!() 2018-08-14 12:14:59 -07:00
Rob Walker
53e86f2fa2 use align! 2018-08-14 12:14:59 -07:00
Michael Vines
757dfd36a3 Report errors better in build log 2018-08-14 11:44:26 -07:00
Rob Walker
708add0e64 fixups 2018-08-14 10:16:34 -07:00
Rob Walker
d8991ae2ca fix UPnP backout, fixes 2018-08-14 10:16:34 -07:00
Rob Walker
5f6cbe0cf8 fixups 2018-08-13 21:07:26 -07:00
Rob Walker
f167b0c2c5 fixups 2018-08-13 21:07:26 -07:00
Rob Walker
f784500fbb fixups
fixes 
2018-08-13 21:07:26 -07:00
Rob Walker
83df47323a initialize recycled data 2018-08-13 21:07:26 -07:00
Greg Fitzgerald
c75d4abb0b Tuck away PoH duration 2018-08-13 20:17:16 -06:00
Greg Fitzgerald
5216a723b1 Pacify clippy 2018-08-13 20:17:16 -06:00
Greg Fitzgerald
b801ca477d Declare fullnode a word 2018-08-13 20:17:16 -06:00
Greg Fitzgerald
c830c604f4 Make BroadcastStage an actual stage
TODO: Why isn't BroadcastStage/RetransmitStage managed by the NCP?
2018-08-13 20:17:16 -06:00
Greg Fitzgerald
0e66606c7f Rename broadcaster to broadcast_stage
And move retransmitter code into retransmit_stage.

TODO: Add a BroadcastStage service
2018-08-13 20:17:16 -06:00
Greg Fitzgerald
8707abe091 Fix erasure build 2018-08-13 20:17:16 -06:00
Greg Fitzgerald
dc2a840985 Move FullNode::new_window into window module 2018-08-13 20:17:16 -06:00
Greg Fitzgerald
2727067b94 Move winow into its own module 2018-08-13 20:17:16 -06:00
Greg Fitzgerald
6a8a494f5d Rename WindowStage to RetransmitStage
The window is used for both broadcasting from leader to validator
and retransmitting between validators.
2018-08-13 20:17:16 -06:00
Greg Fitzgerald
a09d2e252a Move window dependencies out of streamer
No tests!?
2018-08-13 20:17:16 -06:00
Greg Fitzgerald
3e9c463ff1 Offer only 1 way to create a fullnode with an empty window 2018-08-13 20:17:16 -06:00
Tyera Eulberg
46d50f5bde Remove p2p crate (and uPnP support) 2018-08-13 18:22:58 -07:00
Rob Walker
e8da903c6c move tmp_ledger back to target dir 2018-08-13 16:52:53 -07:00
Rob Walker
ab10b7676a use stable cache 2018-08-13 16:23:30 -07:00
Rob Walker
fa44a71d3e move bench to a seprate, parallel step 2018-08-13 16:23:30 -07:00
Rob Walker
c86e9e8568 pad max_data_size to jerasure's alignment requirements 2018-08-13 16:10:51 -07:00
Rob Walker
9e22e23ce6 increase stable timeout until tomorrow 2018-08-13 15:45:50 -07:00
Rob Walker
835f29a178 off by 2 2018-08-13 15:12:12 -07:00
Michael Vines
9688f8fb64 Update IP address 2018-08-13 12:32:09 -07:00
Michael Vines
df5cde74b0 Back out pre-0.7.1 workaround 2018-08-13 12:13:00 -07:00
Michael Vines
231d5e5968 0.7.1 2018-08-13 12:12:27 -07:00
Michael Vines
c2ba72fe1f Fix up validator sanity 2018-08-13 10:23:35 -07:00
Michael Vines
d93786c86a Revert "turn off validator sanity while I work on it"
This reverts commit d4304eea28.
2018-08-13 10:23:35 -07:00
Pankaj Garg
bf15cad36b Add get_finality request and use it from multinode test () 2018-08-13 08:55:13 -07:00
anatoly yakovenko
288ed7a8ea Vote should be valid ()
* test that fails

* fix for test

* rename
2018-08-12 18:19:54 -07:00
anatoly yakovenko
f07c038266 Fix bank coalescing ()
* fix bank coalescing

* comments

* fix bench

* fix bench

* backout banking stage coalescing

* 120 nodes

* 100
2018-08-12 10:04:21 -07:00
Rob Walker
8eed120c38 add missing backslash 2018-08-10 23:45:08 -07:00
Rob Walker
5dbcb43abd more enhancements 2018-08-10 19:53:58 -07:00
Rob Walker
dd1eefaf62 change verify-internal to precheck
update to new ledger API
2018-08-10 19:53:58 -07:00
Rob Walker
35de159d00 better error messages 2018-08-10 19:53:58 -07:00
Rob Walker
546a1e90d5 clippy fixups 2018-08-10 19:53:58 -07:00
Rob Walker
b033e1d904 enhance ledger-tool
* add json, which does the thing with json, move print to Rust's {:?}
  * add --head NUM, to limit how much work gets done for print, json, verify
  * add verify-internal, which very carefully checks ledger format without
      trying first to "recover" it
  * exit with errors on mis-usage
2018-08-10 19:53:58 -07:00
Rob Walker
96d6985895 rework read_ledger, LedgerWriter, and LedgerWindow for recover()
fixes 
2018-08-10 18:07:23 -07:00
Michael Vines
58f220a3b7 Add tic-tac-toe program flow concept 2018-08-10 17:39:54 -07:00
Stephen Akridge
a206f2570d Add hostname to metrics on panic 2018-08-10 17:08:40 -07:00
anatoly yakovenko
2318ffc704 Use a different counter for validator account not found errors. ()
* Use a different counter for validator account not found errors.  This is a usefull signal of something going wrong with the ledger
2018-08-10 15:18:44 -07:00
Rob Walker
d4304eea28 turn off validator sanity while I work on it 2018-08-10 14:56:46 -07:00
Rob Walker
06af9de753 fixups 2018-08-10 11:41:31 -07:00
Rob Walker
7f71e1e09f fixups 2018-08-10 11:41:31 -07:00
Rob Walker
bb7eccd542 check validator startup in testnet-sanity.sh 2018-08-10 11:41:31 -07:00
Rob Walker
b04c71acd9 check issue 910 in testnet-sanity 2018-08-10 11:41:31 -07:00
Rob Walker
bbf9ea89c5 add some flushing to ledger 2018-08-10 11:41:31 -07:00
Rob Walker
846ad61941 use ~/.solana instead of PWD to keep cargo happy, don't rsync --append 2018-08-10 11:41:31 -07:00
Rob Walker
8b41c415b7 add equal sign 2018-08-10 08:05:48 -07:00
Elisa Rossi
197ba8b395 Fixed punctuation 2018-08-09 16:39:04 -07:00
Tyera Eulberg
8d2a61a0c9 Alphabetize bins 2018-08-09 16:23:05 -06:00
Tyera Eulberg
7512317243 Alphabetize dependencies 2018-08-09 16:23:05 -06:00
Greg Fitzgerald
bca2294655 cargo fmt 2018-08-09 13:41:37 -06:00
Greg Fitzgerald
abd55e4159 Add coding guidelines document 2018-08-09 13:41:37 -06:00
Greg Fitzgerald
4a980568ac Rename sig variables to signature
We'll avoid introducing three-letter terms to free up the namespace
for three-letter acronyms.

But recognize the term "sigverify", a verb, to verify a digital
signature.
2018-08-09 13:41:37 -06:00
Greg Fitzgerald
9d436fc5f8 Rename pk variables to pubkey 2018-08-09 13:41:37 -06:00
Greg Fitzgerald
ad331e6d56 Rename PublicKey type to Pubkey
Recognize pubkey as a noun meaning the public key of a keypair.
2018-08-09 13:41:37 -06:00
Greg Fitzgerald
d7e4e57548 Rename public_key variables to pubkey 2018-08-09 13:41:37 -06:00
Greg Fitzgerald
b2067d2721 Rename kp variables to keypair 2018-08-09 13:41:37 -06:00
Greg Fitzgerald
c2bbe4344e Rename KeyPair to Keypair 2018-08-09 13:41:37 -06:00
Greg Fitzgerald
8567253833 Ignore flaky test 2018-08-09 10:15:10 -06:00
Greg Fitzgerald
ca7d4c42dd Rename cur_hashes to num_hashes 2018-08-09 10:15:10 -06:00
Michael Vines
8ca514a5ca Remove unnecessary : 2018-08-08 22:45:39 -07:00
Michael Vines
b605552079 Record network version in testnet-deploy start datapoint 2018-08-08 22:41:02 -07:00
Michael Vines
74f5538bd3 Verify the ledger as a part of sanity 2018-08-08 16:10:54 -07:00
Michael Vines
ff57c7b7df Include ledger tool 2018-08-08 15:12:40 -07:00
Rob Walker
ce8a4fa831 allow received to outpace window, we're already constraining repair
correctly identify sender in ledger_window repair responses, enabling re-transmission
2018-08-08 15:10:44 -07:00
Stephen Akridge
8331aab26a Enable Crdt debug messages to debug validators 2018-08-08 14:22:20 -07:00
Pankaj Garg
a6857dbaaa Updated node count to 230. Increased wmem on CI large 2018-08-08 13:13:18 -07:00
Michael Vines
054298d957 Retry snap install 3 times, sometimes the snap server 503s 2018-08-08 08:56:05 -07:00
Michael Vines
cca240c279 Add SOLANA_NET_NAME, rename SOLANA_NET_URL to SOLANA_NET_ENTRYPOINT 2018-08-08 08:49:30 -07:00
Michael Vines
89f17ceecf Route setup-args 2018-08-08 08:32:23 -07:00
Michael Vines
fe97857c62 Add 'setup-args' snap configuration parameter, to override -p 2018-08-08 08:10:56 -07:00
pgarg66
75854cc234 Update dynamic network test with more nodes ()
- Check for correct OS params in test-large-network.sh
2018-08-08 06:52:57 -07:00
Rob Walker
9783d47fd1 write a "unit" test for WindowLedger (it was working ;)
clear flags on fresh blobs, lest they sometimes impersonate coding blobs...

fix bug: advance *received whether the blob_index is in the window or not,
  failure to do so results in a stalled repair request pipeline
2018-08-08 04:28:09 -07:00
Stephen Akridge
38be61bd22 Check for log level before doing perf counter work
Perf counters, especially when running the dynamic test can cause
functions like crdt::apply_updates to be really slow (>500ms).
2018-08-08 00:16:53 -07:00
Rob Walker
c64e2acf8b set destination address when for ledger window repair responses 2018-08-07 23:31:01 -07:00
Stephen Akridge
a200cedb4b Lower UDP data size to 64k - 128 bytes
rust API gives errors for packets larger than ~65500 and
wikipedia says 65507 is the max size, lowering this avoids the errors.
2018-08-07 18:39:36 -07:00
Michael Vines
5fec0ac82f Validators now rsync the ledger smarter
- Don't re-rsync parts of the ledger that are already present
- Disable compression
2018-08-07 17:38:26 -07:00
Rob Walker
999534248b fixups 2018-08-07 17:27:53 -07:00
Rob Walker
fbc754ea25 plug in LedgerWindow
fixes 
2018-08-07 17:27:53 -07:00
Michael Vines
ecea41a0ab Install EarlyOOM on testnet nodes 2018-08-07 16:58:46 -07:00
pgarg66
1b6d472cb2 Fixed counters for coalescing and broadcast index () 2018-08-07 16:46:48 -07:00
Michael Vines
f0446c7e88 Package curl in Snap for metrics_write_datapoint.sh 2018-08-07 22:41:26 +00:00
Rob Walker
2a0025bb57 get buffered IO back for ledger 2018-08-07 15:34:15 -07:00
pgarg66
64d6d3015a Counters for broadcasted blob idx and coalesced packets () 2018-08-07 14:54:26 -07:00
Stephen Akridge
90550c5b58 Switch to slice arguments and remove clippy exceptions 2018-08-07 14:43:44 -07:00
Michael Vines
53cd2cdd9f Only monitor for OOM kills when a leader, validator or drone is enabled 2018-08-07 14:20:52 -07:00
Stephen Akridge
1ac5d300a4 Rearrange start hash for process_ledger and add a unit test 2018-08-07 14:10:36 -07:00
Michael Vines
642c25bd3b Update path 2018-08-07 13:40:49 -07:00
Michael Vines
df808dedd1 Add simple OOM Killer monitor 2018-08-07 13:35:01 -07:00
Michael Vines
02f9cb415b Ignore failure to write oom_score_adj 2018-08-07 13:35:01 -07:00
Michael Vines
e3cf1e6598 Bundle metrics_write_datapoint.sh in Snap 2018-08-07 13:35:01 -07:00
Michael Vines
7681211c02 Pacify shellcheck 2018-08-07 13:35:01 -07:00
Michael Vines
0ee935dd72 Adjust fullnode/drone oom_score_adj to goad the kernel into killing it first 2018-08-07 10:42:53 -07:00
pgarg66
16772d3d51 Coalesce multiple blobs if received close together ()
- This helps reduce unnecessary growth of window if small blobs
  are received in small space of time
2018-08-07 10:29:57 -07:00
Michael Vines
1c38e40dee Validate ledger once all the tests complete 2018-08-07 10:00:52 -07:00
Pankaj Garg
ceb5a76609 Refactor validator windowing
- a unit test for windowing functions
- issue 
2018-08-07 08:17:32 -07:00
Stephen Akridge
db2392a691 Use last_id from the entries stream instead of last_id from bank
bank will only register ids when has_more is not set because those are
the only ids it has advertised, so it will not register all ids,
however the entry stream will contain unbroken last_id chain, so we
need to track that to get the correct start hash.
2018-08-07 08:14:06 -07:00
Kazuyoshi Kato
9c1b6288a4 Use ? instead of unwrap()
This change addresses , while there are still some unwrap() though.
2018-08-07 08:10:22 -07:00
Michael Vines
575179be8e y 2018-08-06 23:55:00 -07:00
Michael Vines
5b6ffaecc0 s/r$/f/ 2018-08-06 23:36:09 -07:00
Kazuyoshi Kato
efc72b9572 Support -V/--version on all CLI apps
All CLI apps that use clap (in other words, except for bench-streamer)
can use crate_version! to take the version from Cargo.toml.

This change addresses .
2018-08-06 22:03:58 -07:00
Michael Vines
5dc7177540 Remove manually created help text, use clap's text instead. 2018-08-06 19:15:52 -07:00
Michael Vines
78a4b1287d Initialize logger 2018-08-06 19:04:04 -07:00
Michael Vines
c5001869f1 Add verify subcommand 2018-08-06 19:04:04 -07:00
Stephen Akridge
7c31f217d5 Add voting metric even when there are not enough validators 2018-08-06 15:58:10 -07:00
Rob Walker
1152457691 avoid normal validator port a little better for sanity 2018-08-06 15:06:16 -07:00
Rob Walker
3beb38ac8a /tmp/farf no good on multi-user machine 2018-08-06 14:53:40 -07:00
Michael Vines
8cbaa19d2e Report the address that failed to bind 2018-08-06 11:21:25 -07:00
Rob Walker
63d2b2eb42 adjust bank notion of entry_count to aid debugging 2018-08-06 11:20:52 -07:00
Michael Vines
e02da9a15a Clean up tx_count usage 2018-08-06 11:00:25 -07:00
Michael Vines
ae111a131c Condense stdout 2018-08-06 11:00:25 -07:00
Michael Vines
4402e1128f Cleanup 2018-08-06 11:00:25 -07:00
Michael Vines
f55bb6d95c Send/confirm a loopback payment after each batch of transactions 2018-08-06 11:00:25 -07:00
Michael Vines
91741e20fa Add rustc/cargo version check 2018-08-06 09:32:08 -07:00
Rob Walker
0514f5e573 sync() apparently imposes a serious performance penalty 2018-08-06 08:51:41 -07:00
Rob Walker
637d403415 move bank.process_entries() to firsties 2018-08-06 08:51:41 -07:00
Rob Walker
9fabd34156 remove trace! calls, re-arrange replicate_requests to have the bank earlier 2018-08-06 08:51:41 -07:00
Rob Walker
039ed01abf on 2nd thought: do not copy_ledger() for this test 2018-08-06 08:51:41 -07:00
Rob Walker
ead0eb2754 move copy_ledger() back into ledger.rs
Don't recover() for copy(), as copy() is already tolerant of things
    recover() guards against.  Note: recover() is problematic if the ledger is
    "live", i.e. is currently being written to.
2018-08-06 08:51:41 -07:00
Rob Walker
c3db2df7eb tweak random access ledger
* add recover_ledger() to deal with expected common ledger corruptions
  * add verify_ledger() for future use cases (ledger-tool)
  * increase ledger testing
  * allow replicate stage to run without a ledger
  * ledger-tool to output valid json
2018-08-06 08:51:41 -07:00
Rob Walker
ee6c15d2db start on ledger recovery with a description of what that might mean 2018-08-06 08:51:41 -07:00
Rob Walker
715a3d50fe Revert "Revert "clippy fixup""
This reverts commit d173e6ef87.
2018-08-06 08:51:41 -07:00
Rob Walker
692b125391 Revert "Revert "fixups""
This reverts commit e2c68d8775.
2018-08-06 08:51:41 -07:00
Rob Walker
5193819d8e Revert "Revert "plug in new ledger""
This reverts commit 57e928d1d0.
2018-08-06 08:51:41 -07:00
Stephen Akridge
210b9d346f Add voting metrics and -h/--help to get usage for client.sh script 2018-08-05 14:21:49 -07:00
Rob Walker
4c4b0f551e clippy fixups 2018-08-05 13:30:45 -07:00
Rob Walker
6800ff1882 solana-ledger-tool initial commit
does nothing but convert from random-access ledger to json
2018-08-05 13:30:45 -07:00
Michael Vines
399a3852b1 Add sigverify_stage-total_verify_time datapoint 2018-08-04 21:45:58 -07:00
Michael Vines
e7d3069f58 macOS: Adjust maxdgram to allow for large UDP packets 2018-08-04 21:42:59 -07:00
Rob Walker
40ea3e3e61 tweak multinode-demo to work better in snap, validator-x be more stand-alone 2018-08-04 01:04:06 -07:00
Rob Walker
dc9a11bae0 remove rsync size limit for validator's ledger 2018-08-03 23:31:25 -07:00
Rob Walker
906d18a709 move VOTE to trace, info too verbose 2018-08-03 23:04:54 -07:00
Michael Vines
a13058b6c4 Look for 3 nodes (1 leader, 2 validators) 2018-08-03 20:30:29 -07:00
Rob Walker
98ee4b4672 fix up some nits in multinode-demo 2018-08-03 20:19:41 -07:00
pgarg66
7fd7310b96 Prevent a node from overrunning it's receive window ()
- The node drops blobs that will cause it to overrun window
- The node does not ask to repair a blob that overruns the window
2018-08-03 20:15:14 -07:00
Michael Vines
28fa43d2a9 Use env_logger@0.5.12 2018-08-03 20:08:30 -07:00
Michael Vines
1a9e6ffdd7 Try multiple times to confirm a non-zero balance 2018-08-03 19:57:38 -07:00
Rob Walker
c998199954 fixups, add validator-x to sanity 2018-08-03 15:34:11 -07:00
Rob Walker
19792192a7 support any number of self-setup validators on a single host 2018-08-03 15:34:11 -07:00
Rob Walker
4aab413154 recycle the skipped, outside-window blob, fixes 2018-08-03 15:02:55 -07:00
Michael Vines
15a6179b97 Stop installing rustfmt-preview, it's already present 2018-08-03 14:27:11 -07:00
Michael Vines
83b308983f Include rustfmt-preview 2018-08-03 14:11:42 -07:00
Rob Walker
f2b1a04bca cargo fmt fixups 2018-08-03 11:59:25 -07:00
Michael Vines
3e36e6dcf8 Upgrade to rust 1.28 2018-08-03 11:30:40 -07:00
Michael Vines
6feb6a27be Run localnet-sanity in test-stable-perf 2018-08-03 10:46:48 -07:00
Michael Vines
c5ceb15e02 Skip network tuning on CI machines 2018-08-03 10:46:48 -07:00
Rob Walker
57e928d1d0 Revert "plug in new ledger"
This reverts commit 46d9ba5ca0.
2018-08-03 10:24:51 -07:00
Rob Walker
e2c68d8775 Revert "fixups"
This reverts commit b72e91f681.
2018-08-03 10:24:51 -07:00
Rob Walker
d173e6ef87 Revert "clippy fixup"
This reverts commit 384b486b29.
2018-08-03 10:24:51 -07:00
Michael Vines
c230360f4c Wait until recycled machines are reachable before provisioning them 2018-08-02 22:13:17 -07:00
Rob Walker
384b486b29 clippy fixup 2018-08-02 21:50:47 -07:00
Rob Walker
b72e91f681 fixups 2018-08-02 21:50:47 -07:00
Rob Walker
46d9ba5ca0 plug in new ledger 2018-08-02 21:50:47 -07:00
Michael Vines
a9240a42bf Delete unreachable validators to cause a fresh one to be spawned 2018-08-02 20:45:29 -07:00
Michael Vines
a7204d5353 Use a local user to avoid GCP login quota limits 2018-08-02 19:43:35 -07:00
pgarg66
f570ef1c66 Defer repair request for blobs that may still be in avalanche transit () 2018-08-02 19:12:57 -07:00
Stephen Akridge
ee0195d588 Try to measure finality from time seen to when 2/3 of validator..
..set has voted. Add a timestamp to last_ids and use that to
see how long from when 2/3s validator set has voted on them.
2018-08-02 13:21:29 -07:00
Tyera Eulberg
448b8b1c17 Add Hash wrapper and supporting traits 2018-08-01 17:00:51 -07:00
Tyera Eulberg
4d77fa900b Add Signature wrapper and supporting traits 2018-08-01 17:00:51 -07:00
Stephen Akridge
7ccd771ccc Only send sigverify to GPU if batch size is >64
Seems to be a decent crossover point for Xeon E5-2620 v4 8c,16t vs. nvidia 1080ti
2018-08-01 16:38:15 -07:00
Greg Fitzgerald
e9f8b5b9db Fix bench 2018-08-01 16:24:47 -07:00
Greg Fitzgerald
2366c1ebaf Enable cargo audit in CI
Fixes 
2018-08-01 16:24:47 -07:00
Greg Fitzgerald
c5de237276 Upgrade ring and untrusted 2018-08-01 16:24:47 -07:00
Greg Fitzgerald
aa9bc57b4d Implement GenKeys without SecureRandom 2018-08-01 16:24:47 -07:00
Greg Fitzgerald
11df477b20 Make GenKey functions mut
We hide the mutability to implement SecureRandom, but that's going
away.
2018-08-01 16:24:47 -07:00
Greg Fitzgerald
7141750668 new_key -> gen_keypair 2018-08-01 16:24:47 -07:00
Greg Fitzgerald
68675bd1ab Less pub 2018-08-01 16:24:47 -07:00
Greg Fitzgerald
19b3cacd60 Generate a fixed-size array instead of a vector 2018-08-01 16:24:47 -07:00
Tyera Eulberg
bcfaf5d994 Rebase ledger change 2018-08-01 16:15:14 -07:00
Tyera Eulberg
e9499ac5b8 Update PublicKey AsRef to slice 2018-08-01 16:15:14 -07:00
Tyera Eulberg
7ff721e563 Replace pub field with AsRef impl 2018-08-01 16:15:14 -07:00
Tyera Eulberg
fda3b9bbd4 Use new PublicKey format instead of hex 2018-08-01 16:15:14 -07:00
Tyera Eulberg
cf70e5ff2f Handle wrapped PublicKey struct 2018-08-01 16:15:14 -07:00
Tyera Eulberg
a86618faf3 Add PublicKey wrapper
Add custom formatting for PublicKey display and debug
2018-08-01 16:15:14 -07:00
Stephen Akridge
6693386bc5 Lower errors to warnings so they don't print during tests
Negative tests should trigger the warnings, but errors look like
something is wrong.
2018-08-01 16:56:12 -06:00
Michael Vines
4a8a0d03a3 Correct localhost address 2018-08-01 15:49:48 -07:00
Michael Vines
2c9d288ca9 Add a CI metric data point upload timeout to prevent CI build stalls
5 seconds is somewhat arbitrary, seems like enough
2018-08-01 15:49:48 -07:00
Greg Fitzgerald
bb0aabae75 Add cmake, which is needed to build cargo-audit 2018-08-01 16:43:49 -06:00
Michael Vines
5cda0ed964 Airdrop from the leader 2018-08-01 15:21:20 -07:00
Rob Walker
0aba74935b fixups 2018-08-01 14:42:58 -07:00
Rob Walker
4eb666d4f9 provide ledger::copy() 2018-08-01 14:42:58 -07:00
Rob Walker
d5e0cf81ff fixups 2018-08-01 14:42:58 -07:00
Rob Walker
3ea784aff7 clippy fixups 2018-08-01 14:42:58 -07:00
Rob Walker
fef93958c8 fixups, tests 2018-08-01 14:42:58 -07:00
Rob Walker
cae88c90b1 add a persistent ledger of index and data files 2018-08-01 14:42:58 -07:00
Rob Walker
1a8da769b6 ... 2018-08-01 14:42:58 -07:00
Michael Vines
2b259aeb41 testnet now deploys successfully on days of the month < 10 2018-08-01 14:10:52 -07:00
Michael Vines
de7e9b4b4c Remove retry
This was introduced to mask the occasional failure of racy tests.  But this is misguided as it helps hid the true problem, the racy test, and it causes tries builds that fail deterministically to retry only to fail once again.
2018-08-01 12:02:39 -07:00
Michael Vines
0f95031b99 CI builds no longer turn red if a metrics write fails 2018-08-01 11:35:19 -07:00
Pankaj Garg
d622742b84 Mark test-multinode-basic as ignore 2018-08-01 10:13:05 -06:00
Rob Walker
ff254fbe5f re-instate traces 2018-08-01 09:08:38 -07:00
Rob Walker
05153e4884 de-trace this function, new blob is not a dup 2018-08-01 09:08:38 -07:00
Rob Walker
2ece27ee3a fix leak 2018-08-01 09:08:38 -07:00
Greg Fitzgerald
a58df52205 Fix build
Last two PRs crossed in flight. A keypair is now required for all
types of FullNode, not just validators.
2018-08-01 08:53:21 -07:00
sakridge
2ea6f86199 Submit leader's vote after observing 2/3 validator votes ()
* fixup!

* fixups!

* send the vote and count it

* actually vote

* test

* Spelling fixes

* Process the voting transaction in the leader's bank

* Send tokens to the leader

* Give leader tokens in more cases

* Test for write_stage::leader_vote

* Request airdrop inside fullnode and not the script

* Change readme to indicate that drone should be up before leader

And start drone before leader in snap scripts

* Rename _kp => _keypair for keypairs and other review fixups

* Remove empty else
* tweak test_leader_vote numbers to be closer to testing 2/3 boundary
* combine creating blob and transaction for leader/validator
2018-07-31 22:07:53 -07:00
pgarg66
7c5172a65e Converted sigverify disable flag to runtime check instead of "cfg" () 2018-07-31 16:54:24 -07:00
Michael Vines
821e3bc3ca Avoid race between test_lograte and test_lograte_env 2018-07-31 16:08:01 -07:00
Rob Walker
5dd2f737a3 clear out old blobs in find_next_missing 2018-07-31 15:54:32 -07:00
Michael Vines
c9bb5c1f5b Update snap log file documentation 2018-07-31 13:13:27 -07:00
Michael Vines
5d936e5c8a Trap SIGINT for clean ^C shutdown 2018-07-30 17:15:50 -07:00
Michael Vines
e985c2e7d5 .gitignore more generated files 2018-07-30 17:15:50 -07:00
anatoly yakovenko
308b6c3371 Follow Shared prefix convention for Window alias ()
Follow Shared prefix convention for Window alias.
2018-07-30 16:56:01 -07:00
Rob Walker
ea7fa11b3e use size_of() instead of serialized_size() and magic number 8 2018-07-30 16:48:58 -07:00
Michael Vines
5a40ea3fd7 Only map HOME when in CI 2018-07-30 16:36:26 -07:00
Michael Vines
102510ac0e Clear apt cache to reduce image size 2018-07-30 16:36:26 -07:00
Michael Vines
2158329058 Switch to docker-rust image 2018-07-30 16:36:26 -07:00
Michael Vines
bc484ffe5f Add docker-rust image 2018-07-30 16:36:26 -07:00
Michael Vines
6fcf4584d5 Propagate more BUILDKITE environment variables into containers 2018-07-30 16:36:26 -07:00
Michael Vines
1adc83d148 Add localnet-sanity.sh 2018-07-30 16:36:26 -07:00
Michael Vines
647053e973 Terminate child process when main script is interrupted 2018-07-30 16:36:26 -07:00
Michael Vines
95b98b3845 Fix --addr option 2018-07-30 16:36:26 -07:00
Michael Vines
f27613754a Report number of nodes found on failure too 2018-07-30 16:36:26 -07:00
Michael Vines
3e351b0b13 Drop -t 2018-07-30 16:13:51 -07:00
Michael Vines
79ece53e3c Don't panic the tokio worker thread when deserialize() fails 2018-07-30 14:56:53 -07:00
Rob Walker
f341b2ec10 fixups 2018-07-30 14:26:44 -07:00
Rob Walker
167b079e29 fixups 2018-07-30 14:26:44 -07:00
Rob Walker
7ded5a70be fixups 2018-07-30 14:26:44 -07:00
Rob Walker
fc476ff979 implement iterator for parsing length + data ledger 2018-07-30 14:26:44 -07:00
Rob Walker
c3279c8a00 chugga 2018-07-30 14:26:44 -07:00
Rob Walker
e471ea41da fixups 2018-07-30 14:26:44 -07:00
Rob Walker
552d4adff5 use a binary ledger: newline-separated, newline-escaped entries instead of json 2018-07-30 14:26:44 -07:00
pgarg66
0c33c9e0d7 Dynamic network test changes ()
- No sigverify if feature sigverify_cpu_disable is used
- Purge validators in the test if lag count increases beyond
  SOLANA_DYNAMIC_NODES_PURGE_LAG environment variable
- Other useful log messages in the test
2018-07-30 13:57:10 -07:00
Michael Vines
fae9fff24c Unify logging initialization 2018-07-29 19:08:27 -07:00
Michael Vines
79924e407c Include nanoseconds in log timestamp 2018-07-29 19:08:27 -07:00
Michael Vines
18d4da0076 Fetch env_logger from github until 0.5.12 is available 2018-07-29 19:08:27 -07:00
Michael Vines
416c141775 export SKIP_INSTALL=1 to reset the network without reinstalling the snap 2018-07-28 18:04:13 -07:00
Michael Vines
af1a2e83bc Don't panic again when waiting for a panicked validator thread 2018-07-28 16:35:35 -07:00
Michael Vines
4cdb9a73f8 Skip testnet-sanity on manual deploy 2018-07-28 12:37:29 -07:00
Michael Vines
4433730610 Add support for deploying a locally built snap 2018-07-28 12:37:29 -07:00
Michael Vines
71eb5bdecc Factor out vm_foreach 2018-07-28 12:37:29 -07:00
Michael Vines
029e2db2cf Improve assert message 2018-07-28 10:40:50 -07:00
Michael Vines
81db333490 Guard against rsyncing TBs of ledger 2018-07-27 23:53:20 -07:00
Michael Vines
c68ee0040d No need to support migrating from the old ledger format anymore 2018-07-27 23:53:20 -07:00
Michael Vines
d96e267624 Keep around 3GB of logs, 160MB is just not enough 2018-07-27 22:40:21 -07:00
sakridge
0b47404ba6 Check for default leader and use cmp::max for a bit nicer code () 2018-07-27 15:53:31 -07:00
Stephen Akridge
7f4844f426 More stats in dynamic multinode test 2018-07-27 11:55:09 -07:00
Rob Walker
50e1e0ae47 use rust's rotate (in place, yay!) 2018-07-27 11:44:02 -07:00
Michael Vines
538c3b63e1 Log the last_id being voted on 2018-07-27 11:27:51 -07:00
Michael Vines
678b2870ff i 2018-07-27 11:11:37 -07:00
Michael Vines
308d8c254d poll_get_balance no longer fails intermittently for zero balance accounts
While polling for a non-zero balance, it's not uncommon for one of the
get_balance requests to fail with EWOULDBLOCK.  Previously when a get_balance
request failure occurred on the last iteration of the polling loop,
poll_get_balance returned an error even though the N-1 iterations may have
successfully retrieved a balance of 0.
2018-07-26 21:41:07 -07:00
Michael Vines
f11aa4a57b Ensure non-zero exit code if 'balance' command fails 2018-07-26 21:41:07 -07:00
Michael Vines
c52d4eca0b Stop validator first to stop voting before the leader stops 2018-07-26 17:29:32 -07:00
Michael Vines
7672506b45 Validators now vote once a second regardless 2018-07-26 17:07:42 -07:00
Michael Vines
80a02359f7 Add script to audit for security vulnerabilities 2018-07-26 13:42:12 -07:00
Michael Vines
ab3968e3bf Dedup 2018-07-26 11:45:58 -07:00
Michael Vines
42ebf9502a Agent cleaning is now performed in a separate pipeline 2018-07-26 11:37:36 -07:00
Michael Vines
bd4fcf4ac6 Clean out stale buildkite agent build directories 2018-07-26 11:37:36 -07:00
Michael Vines
4dceb73909 Reinstall client nodes in the background to speed up deploys 2018-07-26 09:49:00 -07:00
Rob Walker
dd819cec3d fix off by one in packet.rs 2018-07-26 09:24:44 -07:00
Rob Walker
5115cd7798 large network back to erasure 2018-07-25 20:45:16 -07:00
Rob Walker
cbb8dee360 rework broadcast to understand a separate transmit index for coding blobs 2018-07-25 20:45:16 -07:00
Rob Walker
e0cdcb0973 employ the simple choice for broadcast table of coding blobs: round-robin 2018-07-25 20:45:16 -07:00
Rob Walker
a6a2a745ae fix broadcast of erasure coding blobs
erasure coding blobs were being counted as window slots, skewing transmit_index

erasure coding blobs were being skipped over for broadcast, because they're
  only generated when the last data blob in an erasure block is added to the
  window.... rewind the index to pick up and broadcast those coding blobs
2018-07-25 20:45:16 -07:00
Rob Walker
297896bc49 honor environment variable SOLANA_DYNAMIC_NODES, obviating need to edit+compile for re-test 2018-07-25 20:45:16 -07:00
Michael Vines
f372840354 Collect some datapoints while bench-tps is running 2018-07-25 20:15:43 -07:00
Michael Vines
4c4659be13 Add more stdout 2018-07-25 16:38:21 -07:00
Michael Vines
1b79fe73a1 Emit a metrics datapoint if bench-tps terminates 2018-07-25 15:55:02 -07:00
Michael Vines
5fa072cf16 Avoid quotes around net name 2018-07-25 15:55:02 -07:00
Stephen Akridge
212874e155 Use BlobError for get_size return 2018-07-25 15:54:04 -07:00
Rob Walker
75212f40e7 fix off by one for send_to() of blob 2018-07-25 15:16:56 -07:00
Rob Walker
6fde65577e fixes 2018-07-25 11:07:03 -07:00
Stephen Akridge
80ecef2832 Add --sustained to ci testnet deploy script 2018-07-25 10:16:46 -07:00
Stephen Akridge
edf2ffaf4e Reduce complexity of main for clippy
...and readability
2018-07-25 10:16:46 -07:00
Stephen Akridge
6c275ea5ef More knobs. Arg for tx count per batch and also sustained mode
sustained mode overlaps tx generation with transfer. This mode seems
to have lower peak performance but higher average performance
2018-07-25 10:16:46 -07:00
Stephen Akridge
23ed65b339 Transfer and sign at the same time in bench-tps 2018-07-25 10:16:46 -07:00
Rob Walker
9c7913ac9e trying to raise an error 2018-07-25 08:12:20 -07:00
Rob Walker
8b01e6ac0b implement Blob::get_size(), the counterpart of Blob::set_size() 2018-07-25 08:12:20 -07:00
Rob Walker
ff5854396a deserialize using get_data_size(), which refers to blob.data()'s length,
instead of using msg.meta.size, which refers to the entire blob's length

fixes 
2018-07-25 08:12:20 -07:00
Michael Vines
f0725b4900 Avoid panicking if poll_get_balance() fails while in the transaction loop 2018-07-24 23:31:28 -07:00
Michael Vines
327ba5301d Log token balance throughout the transfer loop 2018-07-24 22:40:12 -07:00
Michael Vines
dcce475f0b Progagate logging configuration to client nodes 2018-07-24 21:40:02 -07:00
Michael Vines
aa2104a21b Reclaim tokens before exiting to avoid leaking tokens 2018-07-24 21:40:02 -07:00
Michael Vines
0206020104 Make airdrops more robust 2018-07-24 21:40:02 -07:00
Rob Walker
33bd1229d9 make next_entries() smarter about fitting Transactions into a Blob 2018-07-24 21:38:06 -07:00
Michael Vines
195098ca2b Failure test case 2018-07-24 21:38:06 -07:00
pgarg66
9daa7bdbe2 Replace rayon with threads for dynamic network test () 2018-07-24 17:54:29 -07:00
Stephen Akridge
6bd18e18ea Add error messages to ledger verify 2018-07-24 17:35:41 -07:00
Rob Walker
8f046cb1f8 disable erasure for large network testing 2018-07-24 16:54:52 -07:00
Michael Vines
735a0ee16d Switch back to running bench-tps in 10 minute iterations 2018-07-24 15:43:25 -07:00
Michael Vines
537be6a29d export SOLANA_DEFAULT_METRICS_RATE 2018-07-24 15:43:25 -07:00
Rob Walker
2b528e2225 fixups 2018-07-24 13:04:34 -07:00
Rob Walker
75505bbd72 fixups 2018-07-24 13:04:34 -07:00
Rob Walker
e1fc7444f9 fixups 2018-07-24 13:04:34 -07:00
Rob Walker
940caf7876 test large network with erasure 2018-07-24 13:04:34 -07:00
Rob Walker
fcdb0403ba eliminate unused parameter received, this branch fixes 2018-07-24 13:04:34 -07:00
Rob Walker
caeb55d066 placate clippy and reduce replicode 2018-07-24 13:04:34 -07:00
Rob Walker
f11e60b801 fix major bug: re-used blobs need to have their flags cleared
plus: lots of additional debug-ability
2018-07-24 13:04:34 -07:00
Rob Walker
54f2146429 fixups 2018-07-24 13:04:34 -07:00
Rob Walker
f60ee87a52 zero the tails of data blobs during generate() and recover() to enable blob reuse 2018-07-24 13:04:34 -07:00
Rob Walker
9c06fe25df enhance unit test to fail when erasure encodes stray bytes of data blobs 2018-07-24 13:04:34 -07:00
Rob Walker
1eec8bf57f fixups 2018-07-24 13:04:34 -07:00
Rob Walker
ddb24ebb61 fixups 2018-07-24 13:04:34 -07:00
Rob Walker
a58c83d999 prevent infinite loop on window wraparound 2018-07-24 13:04:34 -07:00
Rob Walker
6656ec816c protect generate and recover from u64->usize casting issues 2018-07-24 13:04:34 -07:00
Rob Walker
8d2bd43100 fixups 2018-07-24 13:04:34 -07:00
Rob Walker
429ea98ace mutable-coding-blocks 2018-07-24 13:04:34 -07:00
Rob Walker
3d80926508 fixups 2018-07-24 13:04:34 -07:00
Rob Walker
d713e3c2cf send coding in broadcast(), fixups 2018-07-24 13:04:34 -07:00
Rob Walker
5d20d1ddbf get test_window_recover_basic() passing 2018-07-24 13:04:34 -07:00
Rob Walker
257acdcda1 building now 2018-07-24 13:04:34 -07:00
Rob Walker
dab98dcd81 coded => coding 2018-07-24 13:04:34 -07:00
Rob Walker
99653a4d04 rework erasure to have data and coding blobs side-by-side in window 2018-07-24 13:04:34 -07:00
Rob Walker
dda563a169 document process_blob() 2018-07-24 13:04:34 -07:00
Michael Vines
782aa7b23b Cap at 4 threads 2018-07-24 11:35:03 -07:00
Michael Vines
813e438d18 Improve panic message 2018-07-24 11:20:13 -07:00
Michael Vines
7a71adaa8c Adjust threads by the number of cpus 2018-07-23 21:17:36 -07:00
Michael Vines
ce8796bc2e Correctly calculate the expected number of full nodes 2018-07-23 19:55:09 -07:00
Michael Vines
c7e1409f7b Not so much |set -x| 2018-07-23 19:55:09 -07:00
Michael Vines
9de9379925 Add support more more than 1 client node 2018-07-23 19:34:34 -07:00
Michael Vines
7d68b6edc8 Fixup arg processing 2018-07-23 16:51:39 -07:00
Michael Vines
48b5344586 Check for 0 TPS explicitly 2018-07-23 16:51:39 -07:00
Michael Vines
686b7d3737 Report panics if metrics are setup 2018-07-23 16:51:39 -07:00
Michael Vines
7c65e2fbfc Rename variable to improve readability 2018-07-23 16:51:39 -07:00
Michael Vines
96a6e09050 Enable metrics in the TPS client 2018-07-23 16:51:39 -07:00
Stephen Akridge
b3f823d544 Alternate between token reclaim and distribution 2018-07-23 13:17:52 -07:00
Stephen Akridge
ea21c7a43e Limit bench-tps last_id poll to prevent infinite loop 2018-07-23 13:17:52 -07:00
Stephen Akridge
437fb1a8d7 Add -a argument to client in case you want to override the address
advertised by the client
2018-07-23 13:17:52 -07:00
pgarg66
166099b9d9 Start validators in parallel in multinode test () 2018-07-23 09:27:06 -07:00
Michael Vines
c707b3d2e7 Display the total number of transactions for each node once complete 2018-07-22 23:19:33 -07:00
Michael Vines
f7d294de90 Don't rsync leader.json on every iteration 2018-07-22 17:25:00 -07:00
Michael Vines
4ecd0a0e45 Improve bench-tps logging 2018-07-22 16:26:49 -07:00
Michael Vines
7ebbaaeb2d Use bench-tps default duration 2018-07-22 16:26:49 -07:00
Michael Vines
cdcf59ede0 Display a list of all discovered nodes 2018-07-22 11:32:44 -07:00
Michael Vines
5d065133ef Add data point for testnet startup and shutdown 2018-07-21 23:27:24 -07:00
Michael Vines
d403808564 Restart solana.bench-tps every 10 minutes to work around memory leak
cc: 
2018-07-21 19:48:13 -07:00
Michael Vines
3ffdca193d Rename client-demo to bench-tps catchup 2018-07-21 15:46:03 -07:00
Greg Fitzgerald
69688a18c7 Fix clippy warnings
Seems clippy is not linting any of the benches.
2018-07-21 11:36:20 -04:00
Greg Fitzgerald
7193bf28b6 Move streamer bench into standalone executable
It doesn't make use of criterion (or libtest)
2018-07-21 11:36:20 -04:00
Greg Fitzgerald
637f890b91 Rename client-demo to bench-tps 2018-07-21 11:36:20 -04:00
Michael Vines
009d5adcba Tell the client to transact for 1 hour blocks 2018-07-20 17:52:19 -07:00
Michael Vines
52c55a0335 Log to /tmp/solana.log for easy runtime inspection of client activity 2018-07-20 17:45:13 -07:00
Greg Fitzgerald
23428b0381 Migrate drone to poll_for_signature 2018-07-20 20:33:55 -04:00
Greg Fitzgerald
0e305bd7dd Add poll_for_signature 2018-07-20 20:33:55 -04:00
Greg Fitzgerald
c068ca4cb7 Return Signature from transfer_signed and send_airdrop 2018-07-20 20:33:55 -04:00
Greg Fitzgerald
6a8379109d Sleep between retries
Don't congest a congested network.
2018-07-20 20:33:55 -04:00
Michael Vines
120add0e82 Add support for a client node running continuous transactions on the net 2018-07-20 17:07:36 -07:00
Michael Vines
b92ee51c2d Add --loop flag to easily send transactions continuously 2018-07-20 17:07:36 -07:00
Stephen Akridge
cba3b35ac9 Change not_enough_peers to the default log rate 2018-07-20 11:37:12 -07:00
Stephen Akridge
313fed375c Add counter for tx count and limit error messages 2018-07-20 11:37:12 -07:00
Greg Fitzgerald
1e63702c36 cargo fmt 2018-07-20 13:09:01 -04:00
Martius Lim
478ee9a1c4 move tests for 'is_valid_address()' into its own test 2018-07-20 13:09:01 -04:00
Martius Lim
eb1e5dcce4 add test for 'is_valid_address()' 2018-07-20 13:09:01 -04:00
Martius Lim
84225beeef replace 'daddr' checks with 'is_valid_address()' 2018-07-20 13:09:01 -04:00
Michael Vines
9cf0bd9b88 Adjust variable name 2018-07-20 09:50:24 -07:00
Michael Vines
9d25d7611a Protect against unsupported configurations to prevent non-obvious errors later 2018-07-20 09:47:01 -07:00
Michael Vines
1abefb2c7a Pass expected node count to testnet-sanity 2018-07-20 09:47:01 -07:00
Greg Fitzgerald
bcc247f25f Clarify code comment 2018-07-20 12:31:23 -04:00
Michael Vines
68ca9b2cb8 Generalize testnet deployment scripts 2018-07-20 09:10:28 -07:00
pgarg66
686e61d50c Display max TPS from all nodes at end of client demo ()
- Also lists node with 0 TPS and overall average TPS
2018-07-19 20:09:57 -07:00
Michael Vines
17d927ac74 Count testnet nodes as a part of sanity 2018-07-19 12:05:21 -07:00
Michael Vines
966c55f58e Trim CUDA runtime 2018-07-19 11:47:31 -07:00
Michael Vines
d76d3162e5 Slow down deployment more 2018-07-19 10:11:04 -07:00
Michael Vines
d0a2d46923 Don't shellcheck in target/ 2018-07-19 09:41:09 -07:00
Michael Vines
a67f58e9a5 Add -c option to easily interrogate the number of nodes 2018-07-19 09:41:09 -07:00
anatoly yakovenko
fece91c4d1 Test multi node dynamic network ci ()
Buildkite automation for multinode test.  This test is ignored by default because it requires a large cpu machine to run.
2018-07-19 07:50:44 -07:00
Greg Fitzgerald
9d2d9a0189 Update Cargo.toml 2018-07-19 09:02:54 -04:00
Greg Fitzgerald
6d3afc774a Version bump
And authors update
2018-07-19 09:01:06 -04:00
Greg Fitzgerald
88646bf27d Version bump 2018-07-19 08:41:18 -04:00
Rob Walker
0696f9f497 flush writer, makes partial deserialization a bit less likely 2018-07-18 22:53:33 -07:00
Michael Vines
b2ea2455e2 Disable rolling updates 2018-07-18 21:16:03 -07:00
Michael Vines
3f659a69fd Prevent nodes from gossiping with themselves with different ids 2018-07-18 19:38:38 -07:00
anatoly yakovenko
2c62be951f boot invalid height () 2018-07-18 18:10:53 -07:00
Stephen Akridge
2348733d6c remove drone port magic number 2018-07-19 02:01:23 +02:00
pgarg66
cc229b535d Remote multinode script cleanup ()
- Create a known_hosts file if it doesn't exist
  Otherwise ssh-keygen exits
- Move some common rsync code to common_start_setup
- Build the project before deploying it
2018-07-18 16:02:05 -07:00
anatoly yakovenko
7f810a29ff Purge leader ()
* purge leader

* fixup!

* fixup!
2018-07-18 14:39:43 -07:00
Michael Vines
fc1dfd86d2 Disable coverage again :-/ 2018-07-18 12:54:50 -07:00
Michael Vines
5deb34e5bd Little more trace! logging 2018-07-18 12:54:50 -07:00
Michael Vines
39df087902 Permit more than the requested amount of nodes 2018-07-18 12:07:50 -07:00
Greg Fitzgerald
6ff46540b6 Install llvm-cov on nightly to revive coverage
Towards 
2018-07-18 12:52:13 -04:00
Michael Vines
dbab8792e4 Use real default value 2018-07-18 08:23:59 -07:00
Michael Vines
4eb676afaa Tunnel SOLANA_DEFAULT_METRICS_RATE into Snap nodes 2018-07-18 08:23:59 -07:00
Greg Fitzgerald
a6cb2f1bcf Version bump 2018-07-18 09:07:25 -04:00
Greg Fitzgerald
28af9a39b4 Don't clone before borrowing
Clippy told us to change function parameters to references, but
wasn't able to then tell us that the clone() before borrowing
was superfluous. This patch removes those by hand.

No expectation of a performance improvement here, since we were
just cloning reference counts. Just removes a bunch of noise.
2018-07-18 08:04:31 -04:00
anatoly yakovenko
8cf5620b87 crdt_insert_new_entry_counter () 2018-07-17 22:55:53 -07:00
Michael Vines
85d6627ee6 Deploy in one ssh login in a further attempt to avoid hitting GCP login quota 2018-07-17 20:45:52 -07:00
Michael Vines
611a005ec9 Avoid |wait| as it masks failures 2018-07-17 19:52:39 -07:00
Michael Vines
90b3b90391 -p 2018-07-17 19:42:00 -07:00
Michael Vines
fd4f294fd3 Rotate logs at 16MB 2018-07-17 19:42:00 -07:00
Michael Vines
145274c001 Ensure log directories are go+r 2018-07-17 18:16:40 -07:00
Michael Vines
df5d6693f6 Don't cache leader.json to make it easier to switch between nets 2018-07-17 18:16:40 -07:00
Anatoly Yakovenko
05c5603879 error counter 2018-07-17 17:28:23 -07:00
Anatoly Yakovenko
c2c48a5c3c write stage broadcast counters 2018-07-17 17:28:23 -07:00
pgarg66
4af556f70e Added tests for bad gossip address () 2018-07-17 16:27:46 -07:00
anatoly yakovenko
8bad411962 env variable for default metrics rate that gets set for counters ()
* env variable for default metrics rate that gets set for counters

* ignore if env rate is set to 0

* use a slow rate by default

* fixed test
2018-07-17 15:26:10 -07:00
Michael Vines
5b0418793e Keep Snap fullnode/drone logs out of syslog, we're too spammy 2018-07-17 15:08:35 -07:00
pgarg66
4423ee6902 Renamed start_nodes.sh to remote_nodes.sh () 2018-07-17 15:01:53 -07:00
pgarg66
f0c39cc84d Remote multinode scripts cleanup ()
- Also added support for stop nodes
2018-07-17 13:48:25 -07:00
Anatoly Yakovenko
3d45b04da8 review comments 2018-07-17 15:51:32 -04:00
Anatoly Yakovenko
9e2f26a5d2 review comments 2018-07-17 15:51:32 -04:00
Anatoly Yakovenko
a016f6e82e bulds 2018-07-17 15:51:32 -04:00
Anatoly Yakovenko
eb3e5fd204 server too 2018-07-17 15:51:32 -04:00
Anatoly Yakovenko
72282dc493 fast exit dynamic test 2018-07-17 15:51:32 -04:00
Michael Vines
47a22c66b4 Include program name in panic metric 2018-07-17 12:13:22 -07:00
Michael Vines
fb11d8a909 Install panic hook 2018-07-17 12:13:22 -07:00
Michael Vines
7d872f52f4 Add set_panic_hook 2018-07-17 12:13:22 -07:00
Michael Vines
d882bfe65c Ignore/log RequestWindowIndex from self 2018-07-17 12:12:54 -07:00
pgarg66
103584ef27 Use public IP for client gossip, if UPnP fails () 2018-07-17 11:23:32 -07:00
anatoly yakovenko
1fb537deb9 Do not generate gossip requests to unspecified addresses ()
* Do not generate gossip requests to unspecified addresses

* review comments
2018-07-17 09:44:48 -07:00
Michael Vines
2bd48b4207 Display better deploy logs 2018-07-17 09:10:55 -07:00
Michael Vines
f5a6db3dc0 Add daemon plugs 2018-07-17 08:24:37 -07:00
anatoly yakovenko
dd0c1ac5b2 Error counters for streamer ()
* error counters for streamer

* more counters
2018-07-17 08:20:35 -07:00
anatoly yakovenko
d8c9655128 Dynamic test assert ()
* log responder error to warn

* log responder error to warn

* fixup!

* fixed assert

* fixed bad ports issue

* comments

* test for dummy address in Crdt::new instaad of NodeInfo::new

* return error if ContactInfo supplied to Crdt::new cannot be used to connect to network

* comments
2018-07-16 19:31:52 -07:00
anatoly yakovenko
09f2d273c5 less intrusive counters ()
* less intrusive counters

* fixed arg

* tests

* comments
2018-07-16 18:33:50 -07:00
Michael Vines
f6eb85e7a3 Permit Snap RUST_LOG to be overridden 2018-07-16 17:44:54 -07:00
pgarg66
0d85b43901 Fix input parameter processing for client num nodes () 2018-07-16 17:23:35 -07:00
Michael Vines
fdf94a77b4 CUDA is now configurable 2018-07-16 16:23:45 -07:00
pgarg66
af40ab0c04 Split start_nodes script ssh commands to individual scripts () 2018-07-16 16:21:32 -07:00
anatoly yakovenko
015b7a1ddb dash for namespaces () 2018-07-16 15:55:54 -07:00
Anatoly Yakovenko
ab3e460e64 insert votes as they are observed 2018-07-16 13:39:20 -07:00
Michael Vines
194a84c8dd Add testnet-sanity.sh 2018-07-16 12:17:39 -07:00
Michael Vines
51d932dad1 Connect validators to the right leader 2018-07-16 11:05:01 -07:00
Michael Vines
561d31cc13 Add support for master.testnet.s.c 2018-07-16 10:08:58 -07:00
Michael Vines
d6a8e437bb Temporarily disable erasure 2018-07-16 08:15:47 -07:00
Anatoly Yakovenko
4631af5011 counters for vote not found 2018-07-15 20:31:23 -06:00
Michael Vines
5d28729b2a Use ed25519_init() for faster failures 2018-07-15 20:30:32 -06:00
Pankaj Garg
8c08e614b7 Start validator nodes in parallel
- This speeds up overall network startup time
2018-07-15 19:11:52 -06:00
Michael Vines
e76bf1438b A validator and leader running from the same workspace no longer share an identity 2018-07-15 13:34:48 -07:00
Michael Vines
4e177877c9 Add more error checking, better logging, avoid hitting GCP login quota 2018-07-15 09:27:25 -07:00
Michael Vines
60848b9d95 Testnet sanity test failures will now turn the build red 2018-07-14 21:27:27 -07:00
Michael Vines
79b3564a26 Log metrics params to stderr
Keep stdout clean for the actual program.  This is a specific concern for the
wallet command, where there exists tests that capture stdout from the wallet to
confirm transactions.
2018-07-14 21:24:22 -07:00
Michael Vines
1e8c36c555 Be less noisy 2018-07-14 20:42:00 -07:00
Michael Vines
94d015b089 Demote log level 2018-07-14 20:42:00 -07:00
Michael Vines
cfb3736372 Update buildkite-snap.yml 2018-07-14 17:55:03 -07:00
Michael Vines
2b77f62233 Poll longer while waiting for an airdrop 2018-07-14 17:10:44 -07:00
Michael Vines
e8d23c17ca timeout++ 2018-07-14 15:51:32 -07:00
Michael Vines
a7ed2a304a Add CUDA libraries 2018-07-14 15:27:24 -07:00
Michael Vines
0025b42c26 Locate perf libs 2018-07-14 10:24:20 -07:00
Michael Vines
3f7f492cc0 Fix snap client-demo fixes 2018-07-14 00:18:54 -07:00
Michael Vines
490d7875dd Snap client-demo fixes 2018-07-13 23:51:33 -07:00
Michael Vines
4240edf710 solana.client-demo now runs client.sh for the bash extras 2018-07-13 22:57:38 -07:00
Michael Vines
30e50d0f70 Log airdrop amount and client public key 2018-07-13 22:41:52 -07:00
Michael Vines
751c1eba32 Run wallet-sanity against the new testnet 2018-07-13 22:21:41 -07:00
Michael Vines
d349d6aa98 USE_SNAP=1 is now supported 2018-07-13 22:21:41 -07:00
Michael Vines
1f9152dc72 Detect and report airdrop failures 2018-07-13 18:08:28 -07:00
Michael Vines
1b9d50172b Correct log message 2018-07-13 18:08:28 -07:00
Michael Vines
084dbd7f58 Fail gracefully when leader.json is missing 2018-07-13 17:24:25 -07:00
Rob Walker
58c0508f94 add drone information to multinode demo instructions 2018-07-13 17:16:55 -07:00
Michael Vines
dcf82c024f Surface hidden call to solana-keygen 2018-07-13 16:16:46 -07:00
Greg Fitzgerald
b253ed0c46 Version bump 2018-07-13 15:10:45 -06:00
Greg Fitzgerald
61db53fc19 Version bump 2018-07-13 15:04:10 -06:00
Michael Vines
b0ead086a1 Fix bad copy'n'paste 2018-07-13 13:04:38 -07:00
Greg Fitzgerald
a3b22d0d33 Faster benchmarking for CI
Increase that sample size manually when you're doing performance
work.
2018-07-13 14:03:50 -06:00
Michael Vines
28d24497a3 Wait for the leader to initialize before starting the validators 2018-07-13 12:32:24 -07:00
Michael Vines
05cea4c1da dedup 2018-07-13 11:48:17 -07:00
Michael Vines
260f5edfd6 Use correct leader.json 2018-07-13 11:48:17 -07:00
Michael Vines
7105136595 Enable CUDA for the leader node 2018-07-13 11:36:12 -07:00
Michael Vines
54db379bf2 Refresh in parallel 2018-07-13 11:19:31 -07:00
Michael Vines
effbf0b978 Add script to refresh testnet nodes 2018-07-13 11:19:31 -07:00
Michael Vines
8e7a2a9587 Validators now request an airdrop of 1 token before starting up 2018-07-13 10:02:19 -07:00
Michael Vines
18e6ff4167 Fail gracefully when keypair file is unreadable 2018-07-13 10:00:55 -07:00
Michael Vines
fa1cdaa91a Add home plugs to enable Snap access to ~/.config/solana/id.json 2018-07-13 09:32:37 -07:00
Michael Vines
b538b67524 Bump timeout for stable build
When the CI build machine caches are empty stable occasionally needs more than 20m
2018-07-13 09:17:45 -07:00
Michael Vines
2b0f6355af Tagged snap builds now correctly publish to the beta channel 2018-07-12 23:43:59 -07:00
Rob Walker
11b9a0323d fixups 2018-07-12 22:51:55 -07:00
Rob Walker
710fa822a0 fixups 2018-07-12 22:51:55 -07:00
Rob Walker
aaf6ce5aea fixups 2018-07-12 22:51:55 -07:00
Rob Walker
34ea483736 step two: supply a ledger file argument to fullnode in the demo
(also whack unused "myip.sh", even though it was pretty)
2018-07-12 22:51:55 -07:00
Rob Walker
a3ff40476e Banish stdin/stdout for ledger
step one: accept a "ledger file" argument instead of "outfile"
2018-07-12 22:51:55 -07:00
Greg Fitzgerald
4cca3ff454 Fix keypair option in scripts
Thanks @CriesofCarrots!
2018-07-12 21:50:28 -06:00
Greg Fitzgerald
3d9acdd970 Fix nightly 2018-07-12 21:50:28 -06:00
Greg Fitzgerald
428f220b88 Battle shellcheck 2018-07-12 21:50:28 -06:00
Greg Fitzgerald
10add6a8ac Cleanup setup.sh 2018-07-12 21:50:28 -06:00
Greg Fitzgerald
f06a8dceda Fix keygen docs
Thanks @rob-solana
2018-07-12 21:50:28 -06:00
Greg Fitzgerald
545f4f1c87 Pass the owner's keypair to fullnode-config 2018-07-12 21:50:28 -06:00
Greg Fitzgerald
77543d83ff Fix default keypair paths 2018-07-12 21:50:28 -06:00
Greg Fitzgerald
eb6a30cb7c In Wallet, make --tokens required and --to optional 2018-07-12 21:50:28 -06:00
Greg Fitzgerald
97372b8e63 Add --outfile option to solana-keygen 2018-07-12 21:50:28 -06:00
Greg Fitzgerald
cea29ed772 More keygen 2018-07-12 21:50:28 -06:00
Greg Fitzgerald
b5006b8f2b Migrate to solana-keygen
Most of 
2018-07-12 21:50:28 -06:00
Greg Fitzgerald
81c44c605b Add solana-keygen
Same as solana-mint, but without a tokens field.
2018-07-12 14:06:43 -06:00
Greg Fitzgerald
0b66a6626a Use docker image's clippy 2018-07-12 09:40:40 -06:00
Greg Fitzgerald
e8be4d7eae Add clippy to CI 2018-07-12 09:40:40 -06:00
Greg Fitzgerald
30f0c25b65 Fix all remaining clippy warnings
Fixes 
2018-07-12 09:40:40 -06:00
Greg Fitzgerald
73ae3c3301 Apply most of clippy's feedback 2018-07-12 09:40:40 -06:00
Greg Fitzgerald
f98e9aba48 Apply clippy feedback to CLI apps 2018-07-12 09:40:40 -06:00
Michael Vines
84c28a077a Use custom rust nightly image with cargo-cov and clippy pre-installed 2018-07-12 07:25:56 -06:00
Tyera Eulberg
350cf62b90 Sequence client outgoing and incoming txs 2018-07-12 07:24:15 -06:00
Tyera Eulberg
aa4f30c491 Repay transactions from test accounts to client 2018-07-12 07:24:15 -06:00
Tyera Eulberg
3de979aa7c Check client balance and only airdrop if less than TPS quota 2018-07-12 07:24:15 -06:00
Pankaj Garg
5bc133985b Start drone on remote leader node
- Also, enables CUDA for leader node
2018-07-11 20:08:18 -06:00
Greg Fitzgerald
87156e1364 Fix flaky test
The test would fail any time the original value was coincidently
the same as the new bogus value.
2018-07-11 14:16:21 -07:00
Pankaj Garg
45ff142871 Optimized start_nodes script to speed up GCE nodes start time
- Reduced dependency on local network
- Validators get binaries (solana bins and scripts) from leader node
2018-07-11 11:42:25 -06:00
Greg Fitzgerald
2710ff271e cargo fmt 2018-07-11 11:38:41 -06:00
OEM Configuration (temporary user)
468ac9facd Refactor the "ReplicatedData" struct
Rename the "ReplicatedData" struct to the "NodeInfo" struct.
Also refactors and renames the members in this struct.
2018-07-11 11:38:41 -06:00
Rob Walker
705720f086 fixups 2018-07-11 10:37:47 -07:00
Rob Walker
a219e78f00 fixups 2018-07-11 10:37:47 -07:00
Rob Walker
7a41868173 fixups 2018-07-11 10:37:47 -07:00
Rob Walker
e16acec901 fixups 2018-07-11 10:37:47 -07:00
Rob Walker
de44d7475e fixups 2018-07-11 10:37:47 -07:00
Rob Walker
c2dd009e0b fixups 2018-07-11 10:37:47 -07:00
Rob Walker
5a8da75d06 optimize process_ledger() 2018-07-11 10:37:47 -07:00
Greg Fitzgerald
848c6e2371 Reduce sample size to restore number of transactions 2018-07-11 11:18:18 -06:00
Greg Fitzgerald
e3882950cf Run benchmarks from Rust stable CI 2018-07-11 11:18:18 -06:00
Greg Fitzgerald
28f6fbee23 Port all benchmarks to Criterion 2018-07-11 11:18:18 -06:00
Greg Fitzgerald
3144a70b18 Move all benchmarks to benches/ 2018-07-11 11:18:18 -06:00
Stephen Akridge
bed5438831 Improved streamer debug messages
distinguish between threads
2018-07-11 18:26:16 +02:00
Anatoly Yakovenko
6f991b3c11 Send keypair args for validators instead of leaders 2018-07-11 07:54:38 -06:00
Anatoly Yakovenko
03a8a5ed55 only submit to influx when we log
test accumilated value logging

lots of counters

higher influx rate

fix counter name

replicate-transactions
2018-07-11 07:53:39 -06:00
Greg Fitzgerald
0c6d2ef1f4 Fix typo 2018-07-10 19:38:29 -06:00
Greg Fitzgerald
d2be79f38c Use iter_with_setup() to improve precision 2018-07-10 19:38:29 -06:00
Greg Fitzgerald
cc89801b12 Port bank benchmark to Criterion 2018-07-10 19:38:29 -06:00
Greg Fitzgerald
dfa05a8742 Move bank benchmark outside src
This will make it available to third party benchmarking tools.
2018-07-10 19:38:29 -06:00
Pankaj Garg
d7d985365b Add script to create/delete multiple GCE instances
- This script outputs the IP address array that can be used
  with start_nodes script to launch multinode demo
- Changes to start_nodes to compress files for rsync
2018-07-10 18:16:05 -06:00
anatoly yakovenko
0d4e4b18c2 Quiet counter ()
* only submit to influx when we log

* test accumulated value logging
2018-07-10 15:14:59 -07:00
Anatoly Yakovenko
7687436bef some cleanup on messages 2018-07-10 13:32:31 -06:00
Anatoly Yakovenko
d531b9645d review comments 2018-07-10 13:32:31 -06:00
Anatoly Yakovenko
6a1b5a222a rebase builds 2018-07-10 13:32:31 -06:00
Anatoly Yakovenko
be2bf69c93 initial vote stage
wip

voting

wip

move voting into the replicate stage

update

fixup!

fixup!

fixup!

fixup!

fixup!

fixup!

fixup!

fixup!

fixup!

fixup!

update

fixup!

fixup!

fixup!

tpu processing votes in entries before write stage

fixup!

fixup!

txs

make sure validators have an account

fixup!

fixup!

fixup!

exit fullnode correctly

exit on exit not err

try 50

add delay for voting

300

300

startup logs

par start

100

no rayon

retry longer

log leader drop

fix distance

50 nodes

100

handle deserialize error

update

fix broadcast

new table every time

tweaks

table

update

try shuffle table

skip kill

skip add

purge test

fixed tests

rebase 2

fixed tests

fixed rebase

cleanup

ok for blobs to be longer then window

fix init window

60 nodes
2018-07-10 13:32:31 -06:00
Greg Fitzgerald
0672794692 Cleanup leader restarts
Try to avoid adding a runtime codepath when all paths are
statically known.
2018-07-10 11:11:36 -06:00
Greg Fitzgerald
c65c0d9b23 Expose fewer exit variables 2018-07-10 11:11:36 -06:00
Michael Vines
0ee86ff313 Map counters to metrics 2018-07-10 11:11:21 -06:00
Pankaj Garg
3b1aa846b5 Fixed issues with configuring new GCE instances
- New nodes cloned from a working node can be used with the script
- Script takes care of installing SSH keys, and package dependencies correctly
2018-07-10 10:31:03 -06:00
Michael Vines
0a34cb8023 Include hh:mm in image name 2018-07-09 23:07:07 -06:00
Michael Vines
227aa38c8a Add image --family arg 2018-07-09 23:02:46 -06:00
Rob Walker
1dd467ed7d fix issue 2018-07-09 22:27:11 -06:00
Anatoly Yakovenko
922dffb122 fix erasure 2018-07-09 20:40:14 -06:00
Anatoly Yakovenko
63985d4595 renamed to contact_info 2018-07-09 20:40:14 -06:00
Anatoly Yakovenko
97dd1834d7 fix tests, more logs 2018-07-09 20:40:14 -06:00
Anatoly Yakovenko
2ea030be48 stick all the addrs into one struct 2018-07-09 20:40:14 -06:00
Tyera Eulberg
606cfbfe1e Migrate fullnode and fullnode-config to clap for CLI arguments 2018-07-09 20:38:32 -06:00
Rob Walker
90a4ab7e57 fixes issue 2018-07-09 14:50:14 -07:00
Rob Walker
412e15fbdc add test for populated window 2018-07-09 14:50:14 -07:00
Rob Walker
ed0a590549 support an initial window filled with last up-to-WINDOW_SIZE blobs 2018-07-09 14:50:14 -07:00
Greg Fitzgerald
71f05cb23e Vet timestamp source from contract, not leader
Per @aeyakovenko, contracts shouldn't trust the network for
timestamps. Instead, pass the verified public key to the
contract and let it decide if that's a public key it wants
to trust the timestamp from.

Fixes 
2018-07-09 08:40:07 -06:00
Greg Fitzgerald
5f99657523 Remove last_time from bank
We had a test for this, but without `Bank::time_sources` (removed in the last
commit), there's no last_time that can be trusted.
2018-07-09 08:40:07 -06:00
Greg Fitzgerald
587ae1bf3c Remove time_sources from bank
I wrote this, but per
https://github.com/solana-labs/solana#code-coverage, if it doesn't
break a test, it's fair game to delete.
2018-07-09 08:40:07 -06:00
Michael Vines
461dea69d9 Add SOLANA_METRICS_CONFIG environment variable 2018-07-07 19:40:09 -07:00
Michael Vines
22c0e3cd54 Metrics v0.1 2018-07-07 19:40:09 -07:00
Greg Fitzgerald
3ed9567f96 Remove exit variable from RequestStage 2018-07-05 17:32:41 -06:00
Greg Fitzgerald
c4fa841aa9 Remove exit variable from respond [stage]
And drop the sender that feeds input to the responder.
2018-07-05 17:32:41 -06:00
Greg Fitzgerald
f284af1c3d Remove exit variable from WindowStage and retransmit [stage] 2018-07-05 17:32:41 -06:00
Greg Fitzgerald
46602ba9c3 Remove exit variable from ReplicateStage 2018-07-05 17:32:41 -06:00
Greg Fitzgerald
81477246be Remove exit variable from VerifyStage 2018-07-05 17:32:41 -06:00
Greg Fitzgerald
9bd63867aa No longer need to ignore downstream send errors
By removing the exit variables, the downstream stages wait for
upstream stages to drop their senders before exiting.
2018-07-05 17:32:41 -06:00
Greg Fitzgerald
d1c317fd5f Remove exit variable from broadcast [stage] 2018-07-05 17:32:41 -06:00
Greg Fitzgerald
cbd664ba4b Remove exit variable from BankingStage 2018-07-05 17:32:41 -06:00
Greg Fitzgerald
4bb7cefa15 Remove exit variable from WriteStage 2018-07-05 17:32:41 -06:00
Greg Fitzgerald
82c86daa78 Exit write_stage on channel errors 2018-07-05 17:32:41 -06:00
Greg Fitzgerald
b95db62be3 Handle errors consistently
Error handling is still clumsy. We should switch to something like
`error-chain` or `Result<T, Box<Error>>`, but until then, we can
at least be consistent across modules.
2018-07-05 17:32:41 -06:00
Greg Fitzgerald
0f7fdd71cc Remove executable bit from nat.rs 2018-07-05 17:32:41 -06:00
Greg Fitzgerald
af1a7da0d5 Fix code comments 2018-07-05 17:32:41 -06:00
Tyera Eulberg
d698b3da3a Revert tps_demo marker 2018-07-05 15:15:23 -06:00
Tyera Eulberg
6d275d571c Clean up commented code 2018-07-05 15:15:23 -06:00
Tyera Eulberg
63acb82c87 Update drone airdrop test for tps_demo functionality 2018-07-05 15:15:23 -06:00
Tyera Eulberg
4d05b74314 Port solana-client-demo to clap crate for CLI arguments 2018-07-05 15:15:23 -06:00
Tyera Eulberg
37dd511356 Pass client.json location as argument 2018-07-05 15:15:23 -06:00
Tyera Eulberg
96c321da76 Update drone to allow TPS-sized airdrops 2018-07-05 15:15:23 -06:00
Tyera Eulberg
4701540cc9 Migrate solana-client-demo to use drone 2018-07-05 15:15:23 -06:00
Pankaj Garg
f54615b4e3 UDP port for client demo in range
* This change will allow clients to run behind a firewall
  with only certain port range opened for access
2018-07-05 10:17:35 -06:00
Greg Fitzgerald
9c456b2fb0 Fixup the integration tests 2018-07-04 16:40:34 -06:00
Greg Fitzgerald
77bf17064a Add Service trait
Added a consistent interface to all the microservices.
2018-07-04 16:40:34 -06:00
Tyera Eulberg
44150b2e85 Remove unused crate from wallet CLI 2018-07-04 16:39:26 -06:00
Tyera Eulberg
8ec2fe15f3 Port solana-drone to clap crate for CLI arguments 2018-07-04 16:39:26 -06:00
Michael Vines
687af3e3a4 Document source of magic net.core.rmem_max value 2018-07-04 14:24:01 -07:00
Michael Vines
72ab83cd45 Collect timing metrics for CI jobs 2018-07-04 11:00:56 -07:00
Greg Fitzgerald
4b07772e22 Add helper functions for reading entries
```rust
let entries = entry_writer::read_entries_from_str(entries_str).unwrap();
let entries_len = entries.len();
assert_eq!(entries_len, 7);
let bank = Bank::default();
bank.process_ledger(entries).unwrap();
assert_eq!(bank.transaction_count(), entries_len - 2);
```
2018-07-03 19:32:01 -06:00
Anatoly Yakovenko
22d2c962b2 ignore 2018-07-03 18:10:16 -06:00
Anatoly Yakovenko
e771d36278 Better logs 2018-07-03 18:10:16 -06:00
Rob Walker
800c2dd370 make the leader append to the ledger file 2018-07-03 17:17:52 -06:00
Greg Fitzgerald
f38842822f Cleanup code duplication 2018-07-03 16:33:36 -06:00
Greg Fitzgerald
88a6fb86bf Clean up read_entries() and its usage 2018-07-03 16:33:36 -06:00
Greg Fitzgerald
f6fe998ed4 Revert 1dd8c5ed36
Per @sakridge, this might cause a performance degradation. Need
to benchmark it.
2018-07-03 14:00:53 -06:00
Anatoly Yakovenko
16337d7c1e unstable test 2018-07-03 14:00:39 -06:00
Anatoly Yakovenko
ae309f80f7 boot from file test 2018-07-03 14:00:39 -06:00
Anatoly Yakovenko
fa70b3bf70 split out files, fixed a bug @garious! 2018-07-03 14:00:39 -06:00
Anatoly Yakovenko
3a90f138b2 dynamit network test
* cleaned up fullnode api
* added debug_id to ReplicatedData and crdt for debugging
2018-07-03 14:00:39 -06:00
Michael Vines
033f6dcbcb Demote 'sorted leader' log 2018-07-03 08:24:28 -07:00
Greg Fitzgerald
5d8b2f899a Fix wallet doc 2018-07-02 19:21:03 -07:00
Greg Fitzgerald
490205ab84 Fix sanity check
...that my last PR broke
2018-07-02 19:21:03 -07:00
Michael Vines
2c0e704c82 Confirm the payment 2018-07-02 17:59:50 -07:00
Michael Vines
253048f72d Only tune networking for leader/validator 2018-07-02 17:59:50 -07:00
Michael Vines
e09b8430ce Add |wallet reset| command 2018-07-02 17:59:50 -07:00
Michael Vines
9ae283dc3a Expose wallet.sh as a Snap program temporarily 2018-07-02 17:59:50 -07:00
Michael Vines
f95a79d145 Default to using testnet.s.c when running as a Snap 2018-07-02 17:59:50 -07:00
Greg Fitzgerald
0dabdfd48e Use zero to represent a nonexistent account
This also fixes a bug in the thin client where a nonexistent account
would have triggered a panic because we were using `balances[k]` instead
of `balances.get(key)`.

Fixes 
2018-07-02 18:48:40 -06:00
Greg Fitzgerald
d2bb4dc14a Purge empty accounts 2018-07-02 18:48:40 -06:00
Michael Vines
b4dc180592 More quotes to pacify shellcheck 2018-07-02 16:41:22 -07:00
Michael Vines
263577773f Set client config directory correctly in a Snap 2018-07-02 16:41:22 -07:00
Michael Vines
7d708be121 Drone now grabs mint.json locally 2018-07-02 16:41:22 -07:00
Michael Vines
feb1669d39 Correct locate rsync when running as a Snap 2018-07-02 15:57:30 -07:00
Michael Vines
2cbfe41422 Abort nicer on drone connection failure 2018-07-02 15:57:30 -07:00
Michael Vines
b7653865b1 Support testnet.solana.com as first argument 2018-07-02 15:57:30 -07:00
Michael Vines
c72dced8fa Report error when an invalid confirmation signature or public key is provided 2018-07-02 15:57:30 -07:00
Anatoly Yakovenko
6feed5fd56 rebased 2018-07-02 16:34:49 -06:00
Anatoly Yakovenko
b8fe5ae076 rename server to fullnode 2018-07-02 16:34:49 -06:00
Anatoly Yakovenko
7e657d65f3 merged f2ab08c65e 2018-07-02 16:34:49 -06:00
Anatoly Yakovenko
a166bb816e wtfr 2018-07-02 16:34:49 -06:00
Anatoly Yakovenko
2952027d04 wtfr 2018-07-02 16:34:49 -06:00
Anatoly Yakovenko
430d9d9314 fixup! 2018-07-02 16:34:49 -06:00
Anatoly Yakovenko
fa247196c0 fullnode lib 2018-07-02 16:34:49 -06:00
Greg Fitzgerald
5d17c2b58f Return output receivers from each stage
Reaching into the stages' structs for their receivers is, in hindsight,
more awkward than returning multiple values from constructors. By
returning the receiver, the caller can name the receiver whatever it
wants (as you would with any return value), and doesn't need to
reach into the struct for the field (which is super awkward in
combination with move semantics).
2018-07-02 16:18:32 -06:00
Rob Walker
6ee45d282e some auto-detect of wallet commands 2018-07-02 15:51:12 -06:00
Michael Vines
cfc3bd0696 Add manual wallet sanity test 2018-07-02 14:38:01 -07:00
Michael Vines
3e0e09555a Undo UPnP UDP port binding 2018-07-02 14:38:01 -07:00
Michael Vines
1d8bb5144e Drop -demo suffix 2018-07-02 14:38:01 -07:00
Michael Vines
67e0100866 Bind to 0.0.0.0 2018-07-02 14:38:01 -07:00
Michael Vines
f2ab08c65e Reuse request UDP port for responses 2018-07-02 14:38:01 -07:00
Greg Fitzgerald
04a93050e7 No need to share a write lock across single-threaded methods 2018-07-02 15:25:16 -06:00
Michael Vines
03401041db Correct signature argument name 2018-07-02 11:24:13 -07:00
Michael Vines
6eac744a05 Only rsync leader.json once 2018-07-02 10:59:09 -07:00
Michael Vines
ae29e2085f Init env_logger 2018-07-02 10:59:09 -07:00
Michael Vines
7ce0b58af8 Document pkg-config dependency 2018-07-02 10:42:56 -07:00
Michael Vines
ea5663c0da Demote log 2018-07-02 10:28:43 -07:00
Michael Vines
a61bfae8a4 Document libssl-dev dependency 2018-07-02 10:28:43 -07:00
Michael Vines
5716898216 setup.sh can now be more picky about the kind of config it creates 2018-07-02 09:22:26 -07:00
Michael Vines
c0f9e452f2 mint.json is now private 2018-07-02 09:22:26 -07:00
Greg Fitzgerald
4e3526394e Use IntoInterator to simplify write_entries() usage 2018-07-02 09:51:39 -06:00
Greg Fitzgerald
6806a14a3f Use Cursor instead of tempfile.
Faster and one less dependency.
2018-07-02 09:51:39 -06:00
Greg Fitzgerald
ec7e50b37d Consolidate ledger serialization code
The new read_entries() works, but is overly-contrained. Not
using that function yet, but adding it here in the hopes some
Rust guru will tell us how to get that lifetime constraint out
of there.

Fixes 
2018-07-02 09:51:39 -06:00
Greg Fitzgerald
e7b7dfebf5 Add tests for process_ledger() 2018-07-02 09:51:39 -06:00
Michael Vines
a9e0b27772 Speed up snap build
1. Use pre-installed host rust toolchain
2. Build reference/performance fullnode in same part to avoid rebuilding libraries
3. Merge scripts into same part
2018-07-01 17:47:51 -07:00
Greg Fitzgerald
669164bada Boot EntryWriter's Mutex
Finally!
2018-07-01 17:29:24 -06:00
Greg Fitzgerald
4f3a291391 Move the writer into EntryWriter 2018-07-01 17:29:24 -06:00
Greg Fitzgerald
56e37ad2f4 Limit sticky mutex to WriteStage 2018-07-01 17:29:24 -06:00
Greg Fitzgerald
17de79a83a Remove dead code 2018-07-01 17:29:24 -06:00
Greg Fitzgerald
09e9139855 Move channel code to write stage 2018-07-01 17:29:24 -06:00
Greg Fitzgerald
76fc5822c9 Send Vec<Entry> between stages instead of Entry
Might see a performance boost here.
2018-07-01 17:29:24 -06:00
Greg Fitzgerald
c767a854ed Remove useless Arc 2018-07-01 11:35:32 -07:00
Greg Fitzgerald
b60802ddff Refactor such that genesis can use entry_writer 2018-07-01 11:35:32 -07:00
Greg Fitzgerald
1c35d59f26 Receive entries first, then write 2018-07-01 11:35:32 -07:00
Greg Fitzgerald
adcaf715c6 Cleanup write_entries 2018-07-01 11:35:32 -07:00
Greg Fitzgerald
1f9494221b Make space for a write_entry() that only writes entries 2018-07-01 11:35:32 -07:00
Greg Fitzgerald
466d6f76b9 Don't hide error in write_entry() 2018-07-01 11:35:32 -07:00
Greg Fitzgerald
b05e6ce3db Cleanup solana-genesis 2018-07-01 11:35:32 -07:00
Michael Vines
1d812e78d5 Use hard linking to speed up target cache save/restore 2018-07-01 08:59:42 -07:00
Michael Vines
fba494343f Save/restore target/ directory between builds 2018-06-30 22:30:57 -07:00
Michael Vines
0b878eccf8 Map HOME to grant docker containers access to the ~/.cargo registry cache 2018-06-30 21:50:15 -07:00
Michael Vines
98772b16d6 Generalize solana-snap build trigger 2018-06-30 21:50:15 -07:00
Michael Vines
bb82ff0c80 Don't wanna wait 2018-06-30 20:05:27 -07:00
Michael Vines
71af03dc98 Skip snap build for PRs if nothing under snap/ is modified
Additionally relegate push snap build to a secondary CI pipeline
2018-06-30 20:05:27 -07:00
Michael Vines
5671da4a0a Generate a client-specific mint.json 2018-06-30 15:28:17 -07:00
Michael Vines
d63493a852 Grant the snap build more time 2018-06-30 12:20:22 -07:00
Greg Fitzgerald
c06582ba40 Wallet no longer uses global mint.json 2018-06-29 22:26:42 -07:00
Michael Vines
450f271cf7 Move public IP address detection out of bash 2018-06-29 21:12:05 -07:00
Greg Fitzgerald
a31889f129 Readme version bump 2018-06-29 21:39:41 -06:00
Tyera Eulberg
ba6a6f5227 Use clap crate for wallet CLI subcommands and arguments 2018-06-29 21:30:20 -06:00
Greg Fitzgerald
9a38d61048 Version bump 2018-06-29 21:23:50 -06:00
Michael Vines
903ec27754 Add BROKEN_NAT env variable to select Udp sender port workaround 2018-06-29 20:02:28 -07:00
Michael Vines
0b56d603c2 Client NAT traversal 0.1
UPnP is now used to request a port on the NAT be forwarded to the local machine.
This obviously only works for NATs that support UPnP, and thus is not a panacea
for all NAT-related connectivity issues.

Notable hacks in this patch include a transmit/receive UDP socket pair to work
around current protocol limitations whereby the full node assumes its peer can
receive on the same UDP port it transmitted from.
2018-06-29 17:36:26 -07:00
Michael Vines
4ffb5d157a Disable coverage until issue is resolved 2018-06-29 17:36:26 -07:00
Greg Fitzgerald
816246ebee Add doc 2018-06-29 17:28:12 -06:00
Greg Fitzgerald
a9881aee05 Add base58-encoded addresses 2018-06-29 17:28:12 -06:00
Greg Fitzgerald
7b5b989cfe Print usage is a command is not provided 2018-06-29 17:28:12 -06:00
Greg Fitzgerald
c4b62e19f2 Do Proof of History verification before appending entries to the bank
Note: replicate_stage is still using `process_entries()` because
changing it to `process_blocks()` causes the `test_replicate` test to
fail.
2018-06-29 15:35:39 -06:00
Pankaj Garg
79a97ada04 Fix more shellchecks
Also, stops current nodes before pushing updates
2018-06-29 15:19:28 -06:00
Pankaj Garg
da215d1a21 Fix failed shellchecks 2018-06-29 15:19:28 -06:00
Pankaj Garg
9ffc50bead Address review comments 2018-06-29 15:19:28 -06:00
Pankaj Garg
f8352bac2f Address review comments
* Only public IP address in the list
* formatting and other comments
2018-06-29 15:19:28 -06:00
Pankaj Garg
27c1410fdc Script to deploy multiple nodes (one as leader, others as validators)
* The built code is loaded to the nodes
* ssh_keys can be copied to the nodes for internode comm
* The nodes are started with their respective roles
* The client demo is started on the last node
2018-06-29 15:19:28 -06:00
Greg Fitzgerald
9a4733bde7 Remove interactive behavior from wallet 2018-06-29 13:22:20 -06:00
Stephen Akridge
f3df5df52c add validator catchup to multi-node test 2018-06-29 10:39:41 -07:00
Greg Fitzgerald
517d08c637 Cleanup 2018-06-29 09:51:13 -07:00
Greg Fitzgerald
90dd794ae5 cargo fmt
rustfmt 0.6.1-stable (49279d71 2018-05-08)
2018-06-29 09:51:13 -07:00
Tyera Eulberg
e0dbbba8a3 fmt 2018-06-29 09:51:13 -07:00
Tyera Eulberg
705df55a7f Fix program name 2018-06-29 09:51:13 -07:00
Tyera Eulberg
d354e85a9a Return bool on signature check 2018-06-29 09:51:13 -07:00
Tyera Eulberg
e4e1f8ec1e Missing -m parameter handling 2018-06-29 09:51:13 -07:00
Tyera Eulberg
0112a24179 Add confirm command to wallet, and update RPU to check bank for a signature 2018-06-29 09:51:13 -07:00
Tyera Eulberg
d680f6b3a5 Fix bash scripts:
* Use wallet name everywhere
* Update drone to use mint.json
2018-06-29 09:51:13 -07:00
Anatoly Yakovenko
47e732717f more notes 2018-06-29 10:18:36 -06:00
Michael Vines
ec56abfccb Correct setup.sh args 2018-06-29 07:59:16 -07:00
Anatoly Yakovenko
e7cdb402fb highlight 2018-06-29 07:16:03 -06:00
Anatoly Yakovenko
a3fe1965fb spelling 2018-06-29 07:16:03 -06:00
Anatoly Yakovenko
5256e6833e update 2018-06-29 07:16:03 -06:00
Anatoly Yakovenko
051cd2e1ff more examples 2018-06-29 07:16:03 -06:00
Anatoly Yakovenko
51929e7df8 rfcs 2018-06-29 07:16:03 -06:00
Stephen Akridge
a094507bb8 Lower default benchmarking numbers to make CI timeout 2018-06-29 07:14:47 -06:00
Stephen Akridge
8effa4e3e0 Clear old blobs before putting in the new one
Otherwise we will just warn about overrun and not insert new blob
Also, break if the index we find is less than consumed otherwise
we can infinite loop
2018-06-29 07:14:47 -06:00
Stephen Akridge
1c9e7dbc45 Don't recycle in the replicate stage
Windowing stage owns all the blobs now
2018-06-29 07:14:47 -06:00
Stephen Akridge
799b249f02 Don't null blob window until we have to 2018-06-29 07:14:47 -06:00
Greg Fitzgerald
7b4a378c92 Add public-ip option to snap validator with cuda 2018-06-28 21:14:29 -06:00
Michael Vines
47917d00d1 Always bind to 0.0.0.0 regardless of what's being advertised to other nodes 2018-06-28 19:13:36 -07:00
Michael Vines
a4c49af859 Add public-ip argument to setup.sh 2018-06-28 19:13:36 -07:00
Michael Vines
1c1d7d1e0e Log get_last_id errors 2018-06-28 19:13:36 -07:00
Stephen Akridge
d28536d76e Fix spelling of signature 2018-06-28 16:31:33 -07:00
Greg Fitzgerald
63cfbb9497 Only register last entry after a split 2018-06-28 16:54:06 -06:00
Greg Fitzgerald
231040b93e Add tests 2018-06-28 12:28:43 -07:00
Greg Fitzgerald
7c74afc35a Relax recycler
Instead of asserting ref count is 1 before recycling, allow users
to recycle items early. If it turns out that was too early, and
allocate() wants to return it, then boot it and take a memory
allocation performance hit instead.
2018-06-28 12:28:43 -07:00
Greg Fitzgerald
7878a011eb Use a Mint to configure the wallet
* Send transactions from the mint's private key
* By default, send full balance to oneself
* By default, request the mint's number of tokens for airdrops
2018-06-27 17:35:50 -06:00
Greg Fitzgerald
c05416e27d Turn simple-client-demo into a simpler wallet 2018-06-27 17:35:50 -06:00
Michael Vines
ee200d8fa0 Add DEBUG= flag to select debug binaries 2018-06-27 15:34:31 -07:00
Rob Walker
2f42658cd4 ... 2018-06-27 14:51:18 -07:00
Rob Walker
d95e8030fc ... 2018-06-27 14:51:18 -07:00
Greg Fitzgerald
4aedd3f1b6 Cleanup type aliases and imports 2018-06-27 15:06:18 -06:00
Greg Fitzgerald
bb89d6f54d Get back to 500k transactions 2018-06-27 13:50:27 -07:00
Greg Fitzgerald
ed10841e3d No longer spin up accounts for client-demo
Now that the Bank is single-threaded again, we can spin up new
accounts on the fly without concern of thread contention. Likewise,
we can send all transactions from a single account, which was also
problematic in the multi-threaded bank. Sending from one account will
also make client-demo straightforward to port to solana-drone.
2018-06-27 13:50:27 -07:00
Tyera Eulberg
6dac87f2a7 Add entry to snapscraft yaml; cleanup bash header 2018-06-27 13:01:29 -06:00
Tyera Eulberg
a167d0d331 CI cleanup 2018-06-27 13:01:29 -06:00
Tyera Eulberg
eed37820b5 Comments 2018-06-27 13:01:29 -06:00
Tyera Eulberg
124e1fa350 Bash scripts to go with simple-client-demo 2018-06-27 13:01:29 -06:00
Tyera Eulberg
ac40434cdf Initial simple client demo commit 2018-06-27 13:01:29 -06:00
Rob Walker
39354c06f8 take multiple log files, allow restart of leader, validator 2018-06-27 11:41:25 -07:00
Michael Vines
faedb88de0 s/local/declare/g 2018-06-26 19:11:31 -07:00
Michael Vines
5cd1fb486f Automatically add rsync:// prefix to URLs that need it 2018-06-26 17:45:53 -07:00
Rob Walker
5b5df49e6c make client.sh behave like the others, i.e. no tee to a log 2018-06-26 17:02:24 -07:00
Michael Vines
86f9277e2d Add USE_SNAP flag 2018-06-26 16:32:55 -07:00
Greg Fitzgerald
56b09bf0ac cargo fmt 2018-06-26 16:51:07 -06:00
Stephen Akridge
f4c4b9df9c Only free in replicate if we did not hold the reference in window stage
And then free when we are consuming blobs
2018-06-26 16:51:07 -06:00
Greg Fitzgerald
6e568c69a7 Preemptive strike
Should that blob have been passed to a recycler, it would have
had too high a reference count.
2018-06-26 16:51:07 -06:00
Greg Fitzgerald
14d624ee40 Fix benchmarks too
This change will make these benchmarks way slower, because its now
cloning the transaction vector each iteration instead of the ref
counts. We need to rethink these.
2018-06-26 16:51:07 -06:00
Greg Fitzgerald
d5c0557891 Fix test_replicate too 2018-06-26 16:51:07 -06:00
Greg Fitzgerald
1691060a22 Assert recycler is given last reference to data
This patch likely fixes the sporadic failures in the following tests:

```
test server::tests::validator_exit ... FAILED
test streamer::test::streamer_send_test ... FAILED
test thin_client::tests::test_bad_sig ... FAILED
test drone::tests::test_send_airdrop ... FAILED
test thin_client::tests::test_thin_client ... FAILED
```
2018-06-26 16:51:07 -06:00
Rob Walker
a5ce578c72 ... 2018-06-26 16:23:41 -06:00
Greg Fitzgerald
05edfad13a Fix compiler warnings 2018-06-26 15:03:15 -07:00
Greg Fitzgerald
136b43f461 Fix whitespace
TODO: Why didn't "cargo fmt" fail the build.
2018-06-26 15:03:15 -07:00
Rob Walker
ac40c1818f .. 2018-06-26 13:57:10 -07:00
Rob Walker
eb63dbcd2a an Entry needs to be multiple of 4 bytes long 2018-06-26 13:57:10 -07:00
Rob Walker
4e2f1a519e whack next_entries_batched 2018-06-26 13:57:10 -07:00
Rob Walker
55ec7f9fe9 add entry.has_more
* quick fix for really big genesis
 * longer term fix for possible parallel verification over multiple
      Blobs/Entries
2018-06-26 13:57:10 -07:00
Michael Vines
b7ddefdbf9 Empty plug array is not accepted by the snap store 2018-06-26 12:49:40 -07:00
Michael Vines
ce361c2cdc Add Snap fullnode daemon 2018-06-26 12:32:33 -07:00
Michael Vines
ed6ba55261 Add snap/ README 2018-06-26 12:32:33 -07:00
Michael Vines
ec333d2bd6 Revert "-v was renamed to -t"
This reverts commit 8f4ce1e8d0.
2018-06-26 12:32:33 -07:00
OEM Configuration (temporary user)
551f639259 Some pull request fixes(linting + documentation) 2018-06-26 12:31:04 -06:00
OEM Configuration (temporary user)
da3bb6fb93 ran linter 2018-06-26 12:31:04 -06:00
OEM Configuration (temporary user)
08bcb62016 added remote table to update respones 2018-06-26 12:31:04 -06:00
Michael Vines
8f4ce1e8d0 -v was renamed to -t 2018-06-25 20:48:26 -07:00
Greg Fitzgerald
4a534d6abb Don't clone() Arc before recycling
This might fix an awful bug where the streamer reuses a Blob
before the current user is done with it. Recycler should probably
assert ref count is one?

* Also don't collect() an iterator before iterating over it.
2018-06-25 17:33:07 -06:00
Stephen Akridge
b48a8c0555 Chunk blobs into window size to avoid window overrun
Fixes 
2018-06-25 17:33:07 -06:00
Rob Walker
1919ec247b add a clock to validator windows (part 3 of ) ()
* count entries processed by Bank
 * initialize windows with initial height of Entries
2018-06-25 15:07:48 -07:00
Rob Walker
3966eb5374 support MacOS bash and ifconfig properly 2018-06-25 13:14:36 -06:00
Stephen Akridge
c22ef50cae Client fixes, poll for unique last id and cache clients
So we don't keep running up the port range
2018-06-25 10:02:29 -06:00
Michael Vines
be5f2ef9b9 Consolidate CI jobs 2018-06-24 22:28:24 -07:00
Greg Fitzgerald
adfcb79387 Force install cargo-cov 2018-06-24 15:34:30 -06:00
Greg Fitzgerald
73c4c0ac5f Revert "cargo-cov installed by default in nightly?"
This reverts commit 6fc601f696.
2018-06-24 15:34:30 -06:00
Greg Fitzgerald
6fc601f696 cargo-cov installed by default in nightly? 2018-06-24 12:17:42 -06:00
Greg Fitzgerald
07111fb7bb Use llvm-cov instead of gcov
@marco-c called this a hack, but since grcov isn't working
out-of-the-box (panics on call to gcov), we'll take a stab at using
llvm-cov.
2018-06-24 12:17:42 -06:00
Greg Fitzgerald
a06d2170b0 No need for rustfmt on nightly 2018-06-24 12:17:42 -06:00
Greg Fitzgerald
7f53ea3bf3 Generate coverage with Rust nightly
Fixes 

Thanks @marco-c!
2018-06-24 12:17:42 -06:00
Michael Vines
b2accd1c2a Run snap build sooner to better mask the delay 2018-06-24 10:24:32 -07:00
Anatoly Yakovenko
8ef8a8dea7 borrow checker 2018-06-24 11:17:55 -06:00
Anatoly Yakovenko
e929404676 comments 2018-06-24 11:17:55 -06:00
Anatoly Yakovenko
c2258bedae fixed! 2018-06-24 11:17:55 -06:00
Anatoly Yakovenko
215fdbb7ed nits 2018-06-24 11:17:55 -06:00
Anatoly Yakovenko
ee998f6882 fix docs 2018-06-24 11:17:55 -06:00
Anatoly Yakovenko
826e95afca fix logs 2018-06-24 11:17:55 -06:00
Anatoly Yakovenko
47583d48e7 get rid of dummy test 2018-06-24 11:17:55 -06:00
Anatoly Yakovenko
e759cdf061 tests 2018-06-24 11:17:55 -06:00
Anatoly Yakovenko
88503c2a09 generic array fail case 2018-06-24 11:17:55 -06:00
Tyera Eulberg
d5be23dffe fmt 2018-06-24 10:44:17 -06:00
Tyera Eulberg
80c01dc085 Use leader.json or ReplicatedData to get ports for drone 2018-06-24 10:44:17 -06:00
Tyera Eulberg
45b2549fa9 Reset bad TestNode edit 2018-06-24 10:44:17 -06:00
Greg Fitzgerald
c7ce454188 Use pnet_datalink instead of all of pnet
pnet_transport takes a long time to build. It's been especially
painful from within a docker container for reasons I don't care
to understand. pnet_datalink is the only part of pnet we're using
so booting the rest.
2018-06-24 10:39:59 -06:00
Anatoly Yakovenko
7059ea42d6 comments 2018-06-24 09:19:05 -06:00
Anatoly Yakovenko
8ea1c29c9b more notes 2018-06-24 09:19:05 -06:00
Michael Vines
33bbfdbc9b Retry flaky coverage/cuda builds on initial failure 2018-06-23 16:17:25 -07:00
Michael Vines
5de54f8853 Make cuda/erasure build logs public 2018-06-23 16:17:25 -07:00
Michael Vines
a1ac41218a Document CUDA version 2018-06-23 16:17:25 -07:00
Rob Walker
55fc647568 fix more shellcheck 2018-06-23 16:00:17 -07:00
Rob Walker
e83e898eed fix shellcheck's concerns 2018-06-23 16:00:17 -07:00
Rob Walker
eb07e4588b remove IPADDR, which was making Rob feel ill
IPADDR is simple, but not exactly what we need for testnet, where NAT'd
  folks need to join in, need to advertize themselves as on the interweb.

  myip() helps, but there's some TODOs: fullnode-config probably needs to
  be told where it lives in the real world (machine interfaces tell us dick),
  or incorporate something like the "ifconfig.co" code in myip.sh
2018-06-23 16:00:17 -07:00
Michael Vines
563f834c96 Document how to update the snap 2018-06-23 15:29:22 -07:00
Michael Vines
183178681d Simply fetching perf libs 2018-06-23 12:54:38 -07:00
anatoly yakovenko
8dba53e494 debit undo () 2018-06-23 06:14:52 -07:00
Michael Vines
e4782b19a3 Document GCP setup 2018-06-23 02:12:20 -07:00
Michael Vines
ec86b1dffa Adapt to GCP-based CI 2018-06-23 02:12:20 -07:00
anatoly yakovenko
6cb8266c7b cleanup () 2018-06-22 23:26:42 -07:00
Greg Fitzgerald
9c50302a39 Update rfc-001-smart-contracts-engine.md 2018-06-22 22:45:22 -07:00
Anatoly Yakovenko
3313c69898 remove ccal 2018-06-22 22:43:54 -07:00
Anatoly Yakovenko
530c6ca7ec a bunch of updates 2018-06-22 22:43:54 -07:00
Anatoly Yakovenko
07ed2fb523 cleanup 2018-06-22 22:43:54 -07:00
Anatoly Yakovenko
d9ec380a15 cleanup 2018-06-22 22:43:54 -07:00
Anatoly Yakovenko
b60eb3a899 edits 2018-06-22 22:43:54 -07:00
Anatoly Yakovenko
b4df69791b cleanup 2018-06-22 22:43:54 -07:00
Anatoly Yakovenko
c21b8a22b9 update 2018-06-22 22:43:54 -07:00
Anatoly Yakovenko
475a76e656 wip 2018-06-22 22:43:54 -07:00
Anatoly Yakovenko
7ba5d5ef86 first! 2018-06-22 22:43:54 -07:00
Greg Fitzgerald
737dc1ddde Per rustc 1.27.0, we can ensure nested results are used 2018-06-22 22:42:47 -07:00
Greg Fitzgerald
164bf19b36 Update LICENSE 2018-06-22 22:41:04 -07:00
Greg Fitzgerald
25976771d9 Version bump 2018-06-22 22:38:18 -07:00
Greg Fitzgerald
f2198c2e9a cargo fmt
rustc 1.27.0

```
$ cargo fmt --version
rustfmt 0.6.1-stable (49279d71 2018-05-08)
```
2018-06-22 22:23:55 -07:00
Rob Walker
eec19c6d2c move genesis to new Entry generation 2018-06-22 17:46:45 -07:00
Michael Vines
30e03feb5f Add initial CI subsystem documentation 2018-06-22 15:30:29 -07:00
Michael Vines
58cd3bde9f Add drone to snap package 2018-06-22 15:27:25 -07:00
Tyera Eulberg
662bfb7b88 fmt 2018-06-22 14:52:36 -07:00
Tyera Eulberg
5f3e3a17d3 Fix test_send_airdrop cap; add helpful panic msgs 2018-06-22 14:52:36 -07:00
Tyera Eulberg
feba2d9975 Set request cap to a reasonable number, based on 30min reset noted in issue 2018-06-22 14:52:36 -07:00
Tyera Eulberg
e3e3a1c457 Better drone request cap handling 2018-06-22 14:52:36 -07:00
Tyera Eulberg
90628f3c8d Edit TestNode port logic to be consistent with new_leader (fixes hanging test_send_airdrop) 2018-06-22 14:52:36 -07:00
Tyera Eulberg
f6bcadb79d Make airdrop amount variable 2018-06-22 14:52:36 -07:00
Tyera Eulberg
d4ac16773c fmt 2018-06-22 14:52:36 -07:00
Tyera Eulberg
96f044d2bf Clean up; add new_from_server_addr routine 2018-06-22 14:52:36 -07:00
Tyera Eulberg
f31868b913 Rename drone bin; fix usage statement 2018-06-22 14:52:36 -07:00
Tyera Eulberg
73b0ff5b55 Add request-count check and tests; fmt 2018-06-22 14:52:36 -07:00
Tyera Eulberg
64cf69045a Add request-count check; Clean up solana-drone and fmt 2018-06-22 14:52:36 -07:00
Tyera Eulberg
e57dae0f31 Update config and dependencies for solana-drone 2018-06-22 14:52:36 -07:00
Tyera Eulberg
6386e7d5cf Leave some tokens in the mint for solana-drone 2018-06-22 14:52:36 -07:00
Tyera Eulberg
4bad103da9 Add solana-drone CLI 2018-06-22 14:52:36 -07:00
Tyera Eulberg
30a26adb7c Add solana-drone module to library 2018-06-22 14:52:36 -07:00
Stephen Akridge
8be4adfc0a Rename tr => tx and add back comments 2018-06-22 14:34:46 -07:00
Stephen Akridge
fed4cc3965 Remove commented code/imports 2018-06-22 14:34:46 -07:00
Stephen Akridge
7d1e074683 bump last_ids 2018-06-22 14:34:46 -07:00
Stephen Akridge
00516e50a1 last_ids opt 2018-06-22 14:34:46 -07:00
Stephen Akridge
e83d76fbd9 Remove mutexes 2018-06-22 14:34:46 -07:00
Stephen Akridge
304f152315 rwlock balances table 2018-06-22 14:34:46 -07:00
Stephen Akridge
3a82ebf7fd Add multiple source accounts for benchmark 2018-06-22 14:34:46 -07:00
Pankaj Garg
0253d34467 Address review comments 2018-06-22 14:18:45 -07:00
Pankaj Garg
9209f9acde Run multiple instances from same workspace
* Support running leader and validators from multiple machines
  using the same NFS mounted workspace.
* Changes to setup, leader and validator scripts
2018-06-22 14:18:45 -07:00
Rob Walker
3dbbb398df use next_entries() in recorder, recycle blobs in reconstruct_from_blobs 2018-06-22 14:17:36 -07:00
Michael Vines
17e8ad110f Temporarily disable failing CI to get back to green 2018-06-22 11:29:31 -07:00
Rob Walker
5e91d31ed3 issue 309 part 1
* limit the number of Tntries per Blob to at most one
* limit the number of Transactions per Entry such that an Entry will
    always fit in a Blob

With a one-to-one map of Entries to Blobs, recovery of a validator
  is a simple fast-forward from the end of the initial genesis.log
  and tx-*.logs Entries.

TODO: initialize validators' blob index with initial # of Entries.
2018-06-22 09:58:51 -07:00
Greg Fitzgerald
fad9d20820 Add assertion for now next_entry must be called 2018-06-21 21:24:32 -07:00
Greg Fitzgerald
fe9a1c8580 Fix comment 2018-06-21 21:24:32 -07:00
Greg Fitzgerald
cd6d7d5198 Remove redundant clones 2018-06-21 21:24:32 -07:00
Michael Vines
771478bc68 Add simple CUDA version check, warn on mismatch 2018-06-21 13:42:06 -07:00
Michael Vines
c4a59896f8 Run test-erasure in a container 2018-06-21 13:00:40 -07:00
Michael Vines
3eb1608403 Skip --user if SOLANA_DOCKER_RUN_NOSETUID is set 2018-06-21 12:24:52 -07:00
Michael Vines
8fde70d4dc Erasure tests do not require a CUDA agent 2018-06-21 11:42:37 -07:00
Michael Vines
5a047833ed Run snap publishing directly on CUDA agent
This is necessary until we build a docker image that also contains a CUDA
installation
2018-06-21 11:42:37 -07:00
Michael Vines
f6c28e6be1 Update snapcraft docker image contain snapcraft 2.42.1 2018-06-21 11:42:37 -07:00
Michael Vines
0ebf10d19d Snap cuda fullnode 2018-06-21 11:42:37 -07:00
Pankaj Garg
d3005d3ef3 Updated setup and leader scripts
* Setup will us -b to set validator ports
* Leader script fixed to append .log to the log file
* Updated readme file
2018-06-20 19:05:38 -07:00
Anatoly Yakovenko
effcef2184 fixed sleep bug 2018-06-20 16:58:10 -07:00
Michael Vines
89fc0ad7a9 Add convenience script to download performance libraries 2018-06-20 16:48:32 -07:00
Greg Fitzgerald
410272ee1d Update generic_array
Warning: this may have performance implications.
2018-06-20 11:41:54 -07:00
Greg Fitzgerald
1c97bf50b6 Fix nightly
No longer ignore failures in the nightly build.
2018-06-19 17:38:04 -07:00
Rob Walker
4ecd2c9d0b update demo scripts
* add setup to combine init steps, configurable initial mint
  * bash -e -> bash and be explicit about errors with || exit $?
  * feed transaction logs to validator, too
2018-06-19 17:04:44 -07:00
Michael Vines
e592243a09 De-double quote 2018-06-19 13:20:47 -07:00
Greg Fitzgerald
2f4a92e352 Cleanup test 2018-06-19 12:36:02 -07:00
OEM Configuration (temporary user)
ceafc29040 fix linting errors, add retransmission fix to repair requests 2018-06-19 12:36:02 -07:00
OEM Configuration (temporary user)
b20efabfd2 added retransmission of repair messages 2018-06-19 12:36:02 -07:00
Michael Vines
85b6e7293c Add cleanup script to manage build agent disk space 2018-06-19 12:22:45 -07:00
Rob Walker
6aced927ad improve ledger initialization for fullnode
* use a line iterator on stdin instead of a line iterator on a buffer
 * move some unwrap() to expect(), documenting failures
 * bind entry type earlier (for kicks)
2018-06-19 09:28:35 -07:00
Michael Vines
75997e6c08 Allow BUILDKITE_BRANCH in containers 2018-06-18 22:51:30 -07:00
Michael Vines
9040d00110 Package solana as a snap 2018-06-18 17:36:03 -07:00
Michael Vines
8ebc5c6b07 Suggest different validator port by default to coexist with leader port on the same machine 2018-06-18 17:36:03 -07:00
Michael Vines
d4807790ff Add snapcraft login credentials
This file was created as follows:
$ snapcraft export-login --snaps solana --channels beta,edge snapcraft.credentials
$ openssl aes-256-cbc -e -in snapcraft.credentials -out snapcraft.credentials.enc
2018-06-18 17:36:03 -07:00
Rob Walker
0de5e7a285 attempt to understand entry 2018-06-18 16:48:59 -07:00
Greg Fitzgerald
c40000aeda Fix compiler warning 2018-06-18 15:49:41 -07:00
Stephen Akridge
31198bc105 Fix cargo bench nightly 2018-06-18 13:20:39 -07:00
Michael Vines
92599acfca Abort when -l is not present or unreadable 2018-06-16 09:55:03 -07:00
Greg Fitzgerald
f6e70779fe Don't panic if sent a bad packet 2018-06-16 09:51:45 -06:00
Greg Fitzgerald
3017bde686 Update README.md 2018-06-16 09:43:23 -06:00
Greg Fitzgerald
9d84ec4bb3 Delete TODO
That comment predates the separation of RPU and TPU.
2018-06-16 08:59:30 -06:00
Anatoly Yakovenko
586141adb2 Cleanup TVU docs 2018-06-15 22:45:35 -06:00
Michael Vines
3f763f99e2 Fail fast in CI when |cargo fmt| says no 2018-06-15 17:10:00 -07:00
Michael Vines
15c7f36ea3 Improve error reporting 2018-06-15 17:10:00 -07:00
Michael Vines
04d1a083fa Skip |sudo sysctl ...| on macOS 2018-06-15 17:10:00 -07:00
Greg Fitzgerald
327ee1dae8 Apply feedback from @aeyakovenko 2018-06-15 17:01:38 -06:00
Greg Fitzgerald
22885c3e64 Add TVU ASCII art 2018-06-15 17:01:38 -06:00
Stephen Akridge
94ededb54c Add comments and limit digits for tps prints 2018-06-15 11:54:01 -06:00
Stephen Akridge
af6a07697a Change client-demo to run continuosly for some amount of time
Also retry for get_last_id/transaction_count if dropped.
2018-06-15 11:54:01 -06:00
Stephen Akridge
5f1d8c95eb Fix blob data size 2018-06-15 11:54:01 -06:00
Anatoly Yakovenko
7d9e032407 make sure we test large tables 2018-06-15 06:56:35 -06:00
Anatoly Yakovenko
bc918a5ad5 purger 2018-06-15 06:56:35 -06:00
Anatoly Yakovenko
ee54ce4727 min table size before purge 2018-06-15 06:56:35 -06:00
Anatoly Yakovenko
e85bf2f2d5 tests pass 2018-06-15 06:56:35 -06:00
Anatoly Yakovenko
a7460ffbd1 purge validators we havent seen for a long time 2018-06-15 06:56:35 -06:00
Rob Walker
7fe1fd2f95 clean up fullnode cmdline
* fix documentation, other opt parameters
 * add support for a named output file, remove hardcoded "leader.log"
 * resurrect stdout as the default output
2018-06-15 00:41:07 -07:00
Rob Walker
d30670e92e clean up demo bash scripts
* allow other level of RUST logging
 * avoid "echo" in favor of printf (builtin)
 * single quotes for literals, double quotes for variables
2018-06-14 23:12:11 -06:00
Greg Fitzgerald
9b202c6e1e No longer flood log with emtpy entries 2018-06-14 18:04:36 -06:00
Stephen Akridge
87946eafd5 Lower processing transaction message to debug by default 2018-06-14 17:08:11 -06:00
Greg Fitzgerald
7575d3c726 Add timestamp to log messages
Upgraded env_logger and now we have timestamps and colorful messages.

Fixes 
2018-06-14 17:07:58 -06:00
Stephen Akridge
8b9713a934 Skip link_local v4 addresses and v6 address when v6 is not enabled 2018-06-14 16:10:31 -06:00
Stephen Akridge
ec713c18c4 Revert client.sh script to use cargo 2018-06-14 11:56:36 -06:00
anatoly yakovenko
c24b0a1a3f TVU rework ()
Refactored TVU, into stages
* blob fetch stage for blobs
* window stage for maintaining the blob window
* pulled out NCP out of the TVU so they can be separate units
TVU is now just the fetch -> window -> request and bank processing
2018-06-13 21:52:23 -07:00
Robert Kelly
34e0cb0092 cargo fmt 2018-06-13 19:17:21 -07:00
Robert Kelly
7b7c7cba21 changed atty library 2018-06-13 19:17:21 -07:00
Anatoly Yakovenko
c45343dd30 comments 2018-06-13 16:11:44 -06:00
Anatoly Yakovenko
b7f6603c1f fix coverage build 2018-06-13 16:11:44 -06:00
anatoly yakovenko
2d3b052dea allow for insertion of dummy entry points into the local table ()
* Needed for . Create a dummy entry with public key 0..., but with a valid gossip address that we can ask for updates. This will allow validators to discover the full network by just knowing a single node's gossip address without knowing anything else about their identity.
* once we start removing dead validators this entry should get purged since we will never see a message from public key 0, 
2018-06-13 11:42:30 -07:00
Tyera Eulberg
dcb6234771 Fix relative link to client demo 2018-06-12 23:29:04 -06:00
Greg Fitzgerald
e44d423e83 Make version syntax consistent
Using no symbol implies its a symver caret requirement.

https://doc.rust-lang.org/cargo/reference/specifying-dependencies.html
2018-06-11 15:51:25 -06:00
Greg Fitzgerald
5435bb734c Upgrade rand 2018-06-11 15:51:25 -06:00
Michael Vines
13f59adf61 Update build status badge link to publicly available build log 2018-06-10 22:12:09 -07:00
Stephen Akridge
0fce3368d3 Fix json perf counter print and add script to generate a chart 2018-06-09 10:55:22 -07:00
Stephen Akridge
1ee5c81267 Fix benchmarking banking_stage 2018-06-08 15:50:36 -07:00
Stephen Akridge
3bb9d5eb50 Use timing::timestamp for counters 2018-06-08 15:50:36 -07:00
Grimes
efb23f7cf9 Ensure stuck builds eventually timeout 2018-06-07 19:08:03 -07:00
Grimes
013f4674de Target cuda agents 2018-06-07 19:08:03 -07:00
Greg Fitzgerald
6966b25d9c Don't mark a build as failed if line coverage drops
It's not always a problem if line coverage drops. For example,
coverage will drop if you make well-tested code more succinct.
It just means the uncovered code is just a larger percentage of
the codebase.
2018-06-07 19:09:25 -06:00
Greg Fitzgerald
d513f56c8c Readme version bump 2018-06-07 17:32:07 -06:00
Greg Fitzgerald
7aa05618a3 data_replicator -> ncp
Fixes 
2018-06-07 17:11:17 -06:00
Greg Fitzgerald
cdfbbe5e60 Fix diagram typos 2018-06-07 17:11:17 -06:00
Greg Fitzgerald
fe7d1cb81c Race -> Or
Thanks for the suggestion @FishmanL!
2018-06-07 17:11:03 -06:00
Anatoly Yakovenko
c2a9395a4b perf counters 2018-06-07 14:59:21 -07:00
Greg Fitzgerald
586279bcfc Add server diagrams 2018-06-07 15:24:44 -06:00
Greg Fitzgerald
8bd10e7c4c Cleanup top-level lib doc 2018-06-07 15:24:44 -06:00
Greg Fitzgerald
928e6165bc Add TPU & RPU diagrams 2018-06-07 15:24:44 -06:00
anatoly yakovenko
77c9e801aa fixed client demo ()
* fixed client demo
2018-06-07 13:51:15 -07:00
Anatoly Yakovenko
c78132417f fix deadlock 2018-06-07 13:52:33 -06:00
Anatoly Yakovenko
849928887e undo 2018-06-07 13:52:33 -06:00
Anatoly Yakovenko
ba1163d49f fix logs 2018-06-07 13:52:33 -06:00
Anatoly Yakovenko
6f9c89af39 fix deadlock 2018-06-07 13:52:33 -06:00
Greg Fitzgerald
246b8b1242 No longer cat scripts
Because we keep changing those scripts and not updating the readme.

Also, this removes the "-b 9000" starting validators. Is that right?
Or should we be passing that to the validator config?
2018-06-07 12:17:43 -06:00
Stephen Akridge
f0db68cb75 Add note about validator.json and -d flag to config generating scripts 2018-06-07 11:15:41 -06:00
Greg Fitzgerald
f0d1fdfb46 Add missing module descriptions 2018-06-07 09:25:36 -06:00
Greg Fitzgerald
3b8b2e030a Better docs for transaction 2018-06-07 09:25:36 -06:00
Greg Fitzgerald
b4fee677a5 Better docs for payment_plan 2018-06-07 09:25:36 -06:00
Greg Fitzgerald
fe706583f9 Better docs for sigverify_stage 2018-06-07 09:25:36 -06:00
Greg Fitzgerald
d0e0c17ece Better docs for rpu 2018-06-07 09:25:36 -06:00
Greg Fitzgerald
5aaa38bcaf Better docs for write_stage 2018-06-07 09:25:36 -06:00
Greg Fitzgerald
6ff9b27f8e Better docs for entry 2018-06-07 09:25:36 -06:00
Greg Fitzgerald
3f4e035506 Better docs for budget 2018-06-07 09:25:36 -06:00
Greg Fitzgerald
57d9fbb927 Better docs for banking_stage 2018-06-07 09:25:36 -06:00
Greg Fitzgerald
ee44e51b30 Better docs for the bank 2018-06-07 09:25:36 -06:00
Greg Fitzgerald
5011f24123 Move more interesting content into first header
The first header and its content is the only text displayed on
GitHub's mobile page. Reorder so that the disclaimer is the only
information people see.

Disclaimer: IANAL and assume reordering these doesn't matter. :)
2018-06-07 09:25:36 -06:00
Anatoly Yakovenko
d1eda334f3 gdb 2018-06-07 09:25:08 -06:00
Hleb Albau
2ae5ce9f2c Do not use cuda for multinode-demo validator component 2018-06-07 07:04:33 -06:00
Greg Fitzgerald
4f5ac78b7e Add readme to crates.io 2018-06-06 15:00:25 -06:00
Stephen Akridge
074c9af020 Shellcheck again 2018-06-05 15:32:25 -06:00
Stephen Akridge
2da2d4e365 More shellcheck 2018-06-05 15:32:25 -06:00
Stephen Akridge
8eb76ab2a5 Fix shellcheck 2018-06-05 15:32:25 -06:00
Stephen Akridge
a710d95243 Fix non-erasure blob nulling 2018-06-05 15:32:25 -06:00
Stephen Akridge
a06535d7ed cargo fmt 2018-06-05 15:32:25 -06:00
Stephen Akridge
f511ac9be7 Fixes for receiving old blobs and nulling the window with coding 2018-06-05 15:32:25 -06:00
Stephen Akridge
e28ad2177e Receive fixes 2018-06-05 15:32:25 -06:00
Stephen Akridge
cb16fe84cd Rework to fix coding blob insertion 2018-06-05 15:32:25 -06:00
Stephen Akridge
ec3569aa39 Move receive_index to correct place 2018-06-05 15:32:25 -06:00
Stephen Akridge
246edecf53 Add receive_index for broadcast blobs and fix blobs_len position 2018-06-05 15:32:25 -06:00
Stephen Akridge
34834c5af9 Store another size in the data block so it is coded as well 2018-06-05 15:32:25 -06:00
Stephen Akridge
b845245614 Restore more of the blob window and add is_coding helper 2018-06-05 15:32:25 -06:00
Stephen Akridge
5711fb9969 Generate coding for the current blob set not just the first coding set 2018-06-05 15:32:25 -06:00
Stephen Akridge
d1eaecde9a Fix deadlock and only push to contq if it's not a coding blob 2018-06-05 15:32:25 -06:00
Stephen Akridge
00c8505d1e Handle set_flags error 2018-06-05 15:32:25 -06:00
Stephen Akridge
33f01efe69 Fixes for erasure coding 2018-06-05 15:32:25 -06:00
Stephen Akridge
377d312c81 Revert log levels 2018-06-05 15:32:25 -06:00
Stephen Akridge
badf5d5412 Add window recovery 2018-06-05 15:32:25 -06:00
Stephen Akridge
0339f90b40 Fix gf-complete url and symlinks 2018-06-05 15:32:25 -06:00
Stephen Akridge
5455e8e6a9 Review comments 2018-06-05 15:32:25 -06:00
Stephen Akridge
6843b71a0d Debug erasure ci script 2018-06-05 15:32:25 -06:00
Stephen Akridge
634408b5e8 Add erasure build to ci 2018-06-05 15:32:25 -06:00
Stephen Akridge
d053f78b74 Erasure refinements, fix generating orders table 2018-06-05 15:32:25 -06:00
Stephen Akridge
93b6fceb2f generate coding after indexing 2018-06-05 15:32:25 -06:00
Stephen Akridge
ac7860c35d indexing blobs then coding 2018-06-05 15:32:25 -06:00
Stephen Akridge
b0eab8729f Add erasure ci script 2018-06-05 15:32:25 -06:00
Stephen Akridge
cb81f80b31 Enable logging for client demo 2018-06-05 15:32:25 -06:00
Stephen Akridge
ea97529185 Fix erasure compilation 2018-06-05 15:32:25 -06:00
Greg Fitzgerald
f1075191fe Clean up comments: Event -> Transaction 2018-06-04 21:43:46 -06:00
Greg Fitzgerald
74c479fbc9 Delete bitrotted docs 2018-06-04 21:43:46 -06:00
Greg Fitzgerald
7e788d3a17 No longer need explicit refs in rustc 1.26 2018-06-04 21:43:46 -06:00
anatoly yakovenko
69b3c75f0d Power of two chance ()
* fix validator script
* 1/2^30 that we fail due to random returning the same value
2018-06-04 13:32:34 -07:00
Anatoly Yakovenko
b2c2fa40a2 comments 2018-06-03 22:08:25 -06:00
Anatoly Yakovenko
50458d9524 more tests 2018-06-03 22:08:25 -06:00
Anatoly Yakovenko
9679e3e356 more tests 2018-06-03 22:08:25 -06:00
Anatoly Yakovenko
6db9f92b8a crdt gossip tests 2018-06-03 22:08:25 -06:00
Stephen Akridge
4a44498d45 Fix args in validator script, readme version, client-demo perf print 2018-06-02 21:55:27 -06:00
anatoly yakovenko
216510c573 repair socket and receiver thread ()
repair socket and receiver thread
2018-06-02 08:32:51 -07:00
212 changed files with 26868 additions and 4865 deletions
.buildkite/hooks
.clippy.toml.codecov.yml
.github
.gitignoreCONTRIBUTING.mdCargo.tomlLICENSEREADME.mdRELEASE.md_config.yml
benches
build.rs
ci
common
doc
fetch-perf-libs.sh
multinode-demo
net
programs
rfcs
scripts
snap
src
tests

@@ -0,0 +1,2 @@
CI_BUILD_START=$(date +%s)
export CI_BUILD_START

@@ -0,0 +1 @@
post-checkout

@@ -0,0 +1,45 @@
#!/bin/bash -e
#
# Save target/ for the next CI build on this machine
#
if [[ -n $CARGO_TARGET_CACHE_NAME ]]; then
(
d=$HOME/cargo-target-cache/"$CARGO_TARGET_CACHE_NAME"
mkdir -p "$d"
set -x
rsync -a --delete --link-dest="$PWD" target "$d"
du -hs "$d"
)
fi
#
# Add job_stats data point
#
if [[ -z $CI_BUILD_START ]]; then
echo Error: CI_BUILD_START empty
else
CI_BUILD_DURATION=$(( $(date +%s) - CI_BUILD_START + 1 ))
CI_LABEL=${BUILDKITE_LABEL:-build label missing}
PR=false
if [[ $BUILDKITE_BRANCH =~ pull/* ]]; then
PR=true
fi
SUCCESS=true
if [[ $BUILDKITE_COMMAND_EXIT_STATUS != 0 ]]; then
SUCCESS=false
fi
point_tags="pipeline=$BUILDKITE_PIPELINE_SLUG,job=$CI_LABEL,pr=$PR,success=$SUCCESS"
point_tags="${point_tags// /\\ }" # Escape spaces
point_fields="duration=$CI_BUILD_DURATION"
point_fields="${point_fields// /\\ }" # Escape spaces
point="job_stats,$point_tags $point_fields"
scripts/metrics-write-datapoint.sh "$point" || true
fi

@@ -0,0 +1 @@
post-command

@@ -0,0 +1,27 @@
#!/bin/bash -e
# Ensure the pattern "+++ ..." never occurs when |set -x| is set, as buildkite
# interprets this as the start of a log group.
# Ref: https://buildkite.com/docs/pipelines/managing-log-output
export PS4="++"
#
# Restore target/ from the previous CI build on this machine
#
[[ -n "$CARGO_TARGET_CACHE_NAME" ]] || (
d=$HOME/cargo-target-cache/"$CARGO_TARGET_CACHE_NAME"
if [[ -d $d ]]; then
du -hs "$d"
read -r cacheSizeInGB _ < <(du -s --block-size=1000000000 "$d")
if [[ $cacheSizeInGB -gt 5 ]]; then
echo "$d has gotten too large, removing it"
rm -rf "$d"
fi
fi
mkdir -p "$d"/target
set -x
rsync -a --delete --link-dest="$d" "$d"/target .
)

@@ -0,0 +1 @@
pre-command

1
.clippy.toml Normal file

@@ -0,0 +1 @@
too-many-arguments-threshold = 9

@@ -1,2 +1,5 @@
ignore: ignore:
- "src/bin" - "src/bin"
coverage:
status:
patch: off

28
.github/RELEASE_TEMPLATE.md vendored Normal file

@@ -0,0 +1,28 @@
# Release v0.X.Y <milestone name>
fun blurb about the name, what's in the release
## Major Features And Improvements
* bulleted
* list of features and improvements
## Breaking Changes
* bulleted
* list
* of
* protocol changes/breaks
* API breaks
* CLI changes
* etc.
## Bug Fixes and Other Changes
* can be pulled from commit log, or synthesized
## Thanks to our Contributors
This release contains contributions from many people at Solana, as well as:
pull from commit log

12
.gitignore vendored

@@ -1,4 +1,16 @@
Cargo.lock Cargo.lock
/target/ /target/
**/*.rs.bk **/*.rs.bk
.cargo .cargo
# node configuration files
/config/
/config-private/
/config-drone/
/config-validator/
/config-client/
/multinode-demo/test/config-client/
# test temp files, ledgers, etc.
/farf/

53
CONTRIBUTING.md Normal file

@@ -0,0 +1,53 @@
Solana Coding Guidelines
===
The goal of these guidelines is to improve developer productivity by allowing developers to
jump any file in the codebase and not need to adapt to inconsistencies in how the code is
written. The codebase should appear as if it had been authored by a single developer. If you
don't agree with a convention, submit a PR patching this document and let's discuss! Once
the PR is accepted, *all* code should be updated as soon as possible to reflect the new
conventions.
Rust coding conventions
---
* All Rust code is formatted using the latest version of `rustfmt`. Once installed, it will be
updated automatically when you update the compiler with `rustup`.
* All Rust code is linted with Clippy. If you'd prefer to ignore its advice, do so explicitly:
```rust
#[cfg_attr(feature = "cargo-clippy", allow(too_many_arguments))]
```
Note: Clippy defaults can be overridden in the top-level file `.clippy.toml`.
* For variable names, when in doubt, spell it out. The mapping from type names to variable names
is to lowercase the type name, putting an underscore before each capital letter. Variable names
should *not* be abbreviated unless being used as closure arguments and the brevity improves
readability. When a function has multiple instances of the same type, qualify each with a
prefix and underscore (i.e. alice_keypair) or a numeric suffix (i.e. tx0).
* For function and method names, use `<verb>_<subject>`. For unit tests, that verb should
always be `test` and for benchmarks the verb should always be `bench`. Avoid namespacing
function names with some arbitrary word. Avoid abreviating words in function names.
* As they say, "When in Rome, do as the Romans do." A good patch should acknowledge the coding
conventions of the code that surrounds it, even in the case where that code has not yet been
updated to meet the conventions described here.
Terminology
---
Inventing new terms is allowed, but should only be done when the term is widely used and
understood. Avoid introducing new 3-letter terms, which can be confused with 3-letter acronyms.
Some terms we currently use regularly in the codebase:
* fullnode: n. A fully participating network node.
* hash: n. A SHA-256 Hash.
* keypair: n. A Ed25519 key-pair, containing a public and private key.
* pubkey: n. The public key of a Ed25519 key-pair.
* sigverify: v. To verify a Ed25519 digital signature.

@@ -1,20 +1,41 @@
[package] [package]
name = "solana" name = "solana"
description = "Blockchain Rebuilt for Scale" description = "Blockchain, Rebuilt for Scale"
version = "0.6.0" version = "0.9.0"
documentation = "https://docs.rs/solana" documentation = "https://docs.rs/solana"
homepage = "http://solana.com/" homepage = "http://solana.com/"
readme = "README.md"
repository = "https://github.com/solana-labs/solana" repository = "https://github.com/solana-labs/solana"
authors = [ authors = [
"Anatoly Yakovenko <anatoly@solana.com>", "Anatoly Yakovenko <anatoly@solana.com>",
"Greg Fitzgerald <greg@solana.com>", "Greg Fitzgerald <greg@solana.com>",
"Stephen Akridge <stephen@solana.com>", "Stephen Akridge <stephen@solana.com>",
"Michael Vines <mvines@solana.com>",
"Rob Walker <rob@solana.com>",
"Pankaj Garg <pankaj@solana.com>",
"Tyera Eulberg <tyera@solana.com>",
] ]
license = "Apache-2.0" license = "Apache-2.0"
[[bin]] [[bin]]
name = "solana-client-demo" name = "solana-upload-perf"
path = "src/bin/client-demo.rs" path = "src/bin/upload-perf.rs"
[[bin]]
name = "solana-bench-streamer"
path = "src/bin/bench-streamer.rs"
[[bin]]
name = "solana-bench-tps"
path = "src/bin/bench-tps.rs"
[[bin]]
name = "solana-drone"
path = "src/bin/drone.rs"
[[bin]]
name = "solana-replicator"
path = "src/bin/replicator.rs"
[[bin]] [[bin]]
name = "solana-fullnode" name = "solana-fullnode"
@@ -29,16 +50,16 @@ name = "solana-genesis"
path = "src/bin/genesis.rs" path = "src/bin/genesis.rs"
[[bin]] [[bin]]
name = "solana-genesis-demo" name = "solana-ledger-tool"
path = "src/bin/genesis-demo.rs" path = "src/bin/ledger-tool.rs"
[[bin]] [[bin]]
name = "solana-mint" name = "solana-keygen"
path = "src/bin/mint.rs" path = "src/bin/keygen.rs"
[[bin]] [[bin]]
name = "solana-mint-demo" name = "solana-wallet"
path = "src/bin/mint-demo.rs" path = "src/bin/wallet.rs"
[badges] [badges]
codecov = { repository = "solana-labs/solana", branch = "master", service = "github" } codecov = { repository = "solana-labs/solana", branch = "master", service = "github" }
@@ -48,24 +69,80 @@ unstable = []
ipv6 = [] ipv6 = []
cuda = [] cuda = []
erasure = [] erasure = []
test = []
[dependencies] [dependencies]
atty = "0.2"
bincode = "1.0.0"
bs58 = "0.2.0"
byteorder = "1.2.1"
bytes = "0.4"
chrono = { version = "0.4.0", features = ["serde"] }
clap = "2.31"
dirs = "1.0.2"
env_logger = "0.5.12"
generic-array = { version = "0.12.0", default-features = false, features = ["serde"] }
getopts = "0.2"
influx_db_client = "0.3.4"
solana-jsonrpc-core = "0.1"
solana-jsonrpc-http-server = "0.1"
solana-jsonrpc-macros = "0.1"
ipnetwork = "0.12.7"
itertools = "0.7.8"
libc = "0.2.43"
libloading = "0.5.0"
log = "0.4.2"
matches = "0.1.6"
nix = "0.11.0"
pnet_datalink = "0.21.0"
rand = "0.5.1"
rayon = "1.0.0" rayon = "1.0.0"
reqwest = "0.9.0"
ring = "0.13.2"
sha2 = "0.7.0" sha2 = "0.7.0"
generic-array = { version = "0.9.0", default-features = false, features = ["serde"] }
serde = "1.0.27" serde = "1.0.27"
serde_cbor = "0.9.0"
serde_derive = "1.0.27" serde_derive = "1.0.27"
serde_json = "1.0.10" serde_json = "1.0.10"
ring = "0.12.1" socket2 = "0.3.8"
untrusted = "0.5.1" solana_program_interface = { path = "common", version="0.1.0" }
bincode = "1.0.0" sys-info = "0.5.6"
chrono = { version = "0.4.0", features = ["serde"] } tokio = "0.1"
log = "^0.4.1" tokio-codec = "0.1"
env_logger = "^0.4.1" untrusted = "0.6.2"
matches = "^0.1.6"
byteorder = "^1.2.1" [dev-dependencies]
libc = "^0.2.1" noop = { path = "programs/noop", version="0.1.0" }
getopts = "^0.2" print = { path = "programs/print", version="0.1.0" }
isatty = "0.1" move_funds = { path = "programs/move_funds", version="0.1.0" }
rand = "0.4.2"
pnet = "^0.21.0" [[bench]]
name = "bank"
[[bench]]
name = "banking_stage"
[[bench]]
name = "ledger"
[[bench]]
name = "signature"
[[bench]]
name = "sigverify"
[workspace]
members = [
".",
"common",
"programs/noop",
"programs/print",
"programs/move_funds",
]
default-members = [
".",
"common",
"programs/noop",
"programs/print",
"programs/move_funds",
]

@@ -1,4 +1,4 @@
Copyright 2018 Anatoly Yakovenko, Greg Fitzgerald and Stephen Akridge Copyright 2018 Solana Labs, Inc.
Licensed under the Apache License, Version 2.0 (the "License"); Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License. you may not use this file except in compliance with the License.

287
README.md

@@ -1,23 +1,27 @@
[![Solana crate](https://img.shields.io/crates/v/solana.svg)](https://crates.io/crates/solana) [![Solana crate](https://img.shields.io/crates/v/solana.svg)](https://crates.io/crates/solana)
[![Solana documentation](https://docs.rs/solana/badge.svg)](https://docs.rs/solana) [![Solana documentation](https://docs.rs/solana/badge.svg)](https://docs.rs/solana)
[![Build status](https://badge.buildkite.com/d4c4d7da9154e3a8fb7199325f430ccdb05be5fc1e92777e51.svg?branch=master)](https://buildkite.com/solana-labs/solana) [![Build status](https://badge.buildkite.com/d4c4d7da9154e3a8fb7199325f430ccdb05be5fc1e92777e51.svg?branch=master)](https://solana-ci-gate.herokuapp.com/buildkite_public_log?https://buildkite.com/solana-labs/solana/builds/latest/master)
[![codecov](https://codecov.io/gh/solana-labs/solana/branch/master/graph/badge.svg)](https://codecov.io/gh/solana-labs/solana) [![codecov](https://codecov.io/gh/solana-labs/solana/branch/master/graph/badge.svg)](https://codecov.io/gh/solana-labs/solana)
Blockchain, Rebuilt for Scale
===
Solana&trade; is a new blockchain architecture built from the ground up for scale. The architecture supports
up to 710 thousand transactions per second on a gigabit network.
Disclaimer Disclaimer
=== ===
All claims, content, designs, algorithms, estimates, roadmaps, specifications, and performance measurements described in this project are done with the author's best effort. It is up to the reader to check and validate their accuracy and truthfulness. Furthermore nothing in this project constitutes a solicitation for investment. All claims, content, designs, algorithms, estimates, roadmaps, specifications, and performance measurements described in this project are done with the author's best effort. It is up to the reader to check and validate their accuracy and truthfulness. Furthermore nothing in this project constitutes a solicitation for investment.
Solana: Blockchain Rebuilt for Scale
===
Solana&trade; is a new blockchain architecture built from the ground up for scale. The architecture supports
up to 710 thousand transactions per second on a gigabit network.
Introduction Introduction
=== ===
It's possible for a centralized database to process 710,000 transactions per second on a standard gigabit network if the transactions are, on average, no more than 176 bytes. A centralized database can also replicate itself and maintain high availability without significantly compromising that transaction rate using the distributed system technique known as Optimistic Concurrency Control [H.T.Kung, J.T.Robinson (1981)]. At Solana, we're demonstrating that these same theoretical limits apply just as well to blockchain on an adversarial network. The key ingredient? Finding a way to share time when nodes can't trust one-another. Once nodes can trust time, suddenly ~40 years of distributed systems research becomes applicable to blockchain! Furthermore, and much to our surprise, it can implemented using a mechanism that has existed in Bitcoin since day one. The Bitcoin feature is called nLocktime and it can be used to postdate transactions using block height instead of a timestamp. As a Bitcoin client, you'd use block height instead of a timestamp if you don't trust the network. Block height turns out to be an instance of what's being called a Verifiable Delay Function in cryptography circles. It's a cryptographically secure way to say time has passed. In Solana, we use a far more granular verifiable delay function, a SHA 256 hash chain, to checkpoint the ledger and coordinate consensus. With it, we implement Optimistic Concurrency Control and are now well in route towards that theoretical limit of 710,000 transactions per second. It's possible for a centralized database to process 710,000 transactions per second on a standard gigabit network if the transactions are, on average, no more than 176 bytes. A centralized database can also replicate itself and maintain high availability without significantly compromising that transaction rate using the distributed system technique known as Optimistic Concurrency Control [\[H.T.Kung, J.T.Robinson (1981)\]](http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.65.4735). At Solana, we're demonstrating that these same theoretical limits apply just as well to blockchain on an adversarial network. The key ingredient? Finding a way to share time when nodes can't trust one-another. Once nodes can trust time, suddenly ~40 years of distributed systems research becomes applicable to blockchain!
> Perhaps the most striking difference between algorithms obtained by our method and ones based upon timeout is that using timeout produces a traditional distributed algorithm in which the processes operate asynchronously, while our method produces a globally synchronous one in which every process does the same thing at (approximately) the same time. Our method seems to contradict the whole purpose of distributed processing, which is to permit different processes to operate independently and perform different functions. However, if a distributed system is really a single system, then the processes must be synchronized in some way. Conceptually, the easiest way to synchronize processes is to get them all to do the same thing at the same time. Therefore, our method is used to implement a kernel that performs the necessary synchronization--for example, making sure that two different processes do not try to modify a file at the same time. Processes might spend only a small fraction of their time executing the synchronizing kernel; the rest of the time, they can operate independently--e.g., accessing different files. This is an approach we have advocated even when fault-tolerance is not required. The method's basic simplicity makes it easier to understand the precise properties of a system, which is crucial if one is to know just how fault-tolerant the system is. [\[L.Lamport (1984)\]](http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.71.1078)
Furthermore, and much to our surprise, it can be implemented using a mechanism that has existed in Bitcoin since day one. The Bitcoin feature is called nLocktime and it can be used to postdate transactions using block height instead of a timestamp. As a Bitcoin client, you'd use block height instead of a timestamp if you don't trust the network. Block height turns out to be an instance of what's being called a Verifiable Delay Function in cryptography circles. It's a cryptographically secure way to say time has passed. In Solana, we use a far more granular verifiable delay function, a SHA 256 hash chain, to checkpoint the ledger and coordinate consensus. With it, we implement Optimistic Concurrency Control and are now well in route towards that theoretical limit of 710,000 transactions per second.
Testnet Demos Testnet Demos
@@ -58,94 +62,81 @@ your odds of success if you check out the
before proceeding: before proceeding:
```bash ```bash
$ git checkout v0.6.0 $ git checkout v0.8.0
```
Configuration Setup
---
The network is initialized with a genesis ledger and leader/validator configuration files.
These files can be generated by running the following script.
```bash
$ ./multinode-demo/setup.sh
```
Drone
---
In order for the leader, client and validators to work, we'll need to
spin up a drone to give out some test tokens. The drone delivers Milton
Friedman-style "air drops" (free tokens to requesting clients) to be used in
test transactions.
Start the drone on the leader node with:
```bash
$ ./multinode-demo/drone.sh
``` ```
Singlenode Testnet Singlenode Testnet
--- ---
The fullnode server is initialized with a ledger from stdin and
generates new ledger entries on stdout. To create the input ledger, we'll need
to create *the mint* and use it to generate a *genesis ledger*. It's done in
two steps because the mint-demo.json file contains private keys that will be
used later in this demo.
```bash
$ echo 1000000000 | cargo run --release --bin solana-mint-demo > mint-demo.json
$ cat mint-demo.json | cargo run --release --bin solana-genesis-demo > genesis.log
```
Before you start a fullnode, make sure you know the IP address of the machine you Before you start a fullnode, make sure you know the IP address of the machine you
want to be the leader for the demo, and make sure that udp ports 8000-10000 are want to be the leader for the demo, and make sure that udp ports 8000-10000 are
open on all the machines you want to test with. open on all the machines you want to test with.
Generate a leader configuration file with: Now start the server in a separate shell:
```bash ```bash
cargo run --release --bin solana-fullnode-config > leader.json $ ./multinode-demo/leader.sh
``` ```
Now start the server: Wait a few seconds for the server to initialize. It will print "leader ready..." when it's ready to
receive transactions. The leader will request some tokens from the drone if it doesn't have any.
```bash The drone does not need to be running for subsequent leader starts.
$ cat ./multinode-demo/leader.sh
#!/bin/bash
export RUST_LOG=solana=info
sudo sysctl -w net.core.rmem_max=26214400
cargo run --release --bin solana-fullnode -- -l leader.json < genesis.log
$ ./multinode-demo/leader.sh > leader-txs.log
```
To run a performance-enhanced fullnode on Linux, download `libcuda_verify_ed25519.a`. Enable
it by adding `--features=cuda` to the line that runs `solana-fullnode` in `leader.sh`.
```bash
$ wget https://solana-build-artifacts.s3.amazonaws.com/v0.6.0/libcuda_verify_ed25519.a
cargo run --release --features=cuda --bin solana-fullnode -- -l leader.json < genesis.log
```
Wait a few seconds for the server to initialize. It will print "Ready." when it's ready to
receive transactions.
Multinode Testnet Multinode Testnet
--- ---
To run a multinode testnet, after starting a leader node, spin up some validator nodes: To run a multinode testnet, after starting a leader node, spin up some validator nodes in
separate shells:
```bash ```bash
$ cat ./multinode-demo/validator.sh $ ./multinode-demo/validator.sh
#!/bin/bash
rsync -v -e ssh $1/mint-demo.json .
rsync -v -e ssh $1/leader.json .
rsync -v -e ssh $1/genesis.log .
export RUST_LOG=solana=info
sudo sysctl -w net.core.rmem_max=26214400
cargo run --release --bin solana-fullnode -- -l validator.json -v leader.json -b 9000 -d < genesis.log
$ ./multinode-demo/validator.sh ubuntu@10.0.1.51:~/solana > validator-txs.log #The leader machine
``` ```
As with the leader node, you can run a performance-enhanced validator fullnode by adding To run a performance-enhanced leader or validator (on Linux),
`--features=cuda` to the line that runs `solana-fullnode` in `validator.sh`. [CUDA 9.2](https://developer.nvidia.com/cuda-downloads) must be installed on
your system:
```bash ```bash
cargo run --release --features=cuda --bin solana-fullnode -- -l validator.json -v leader.json -b 9000 -d < genesis.log $ ./fetch-perf-libs.sh
$ SOLANA_CUDA=1 ./multinode-demo/leader.sh
$ SOLANA_CUDA=1 ./multinode-demo/validator.sh
``` ```
Testnet Client Demo Testnet Client Demo
--- ---
Now that your singlenode or multinode testnet is up and running, in a separate shell, let's send it some transactions! Note we pass in Now that your singlenode or multinode testnet is up and running let's send it
the JSON configuration file here, not the genesis ledger. some transactions!
In a separate shell start the client:
```bash ```bash
$ cat ./multinode-demo/client.sh $ ./multinode-demo/client.sh # runs against localhost by default
#!/bin/bash
export RUST_LOG=solana=info
rsync -v -e ssh $1/leader.json .
rsync -v -e ssh $1/mint-demo.json .
cat mint-demo.json | cargo run --release --bin solana-client-demo -- -l leader.json
$ ./multinode-demo/client.sh ubuntu@10.0.1.51:~/solana #The leader machine
``` ```
What just happened? The client demo spins up several threads to send 500,000 transactions What just happened? The client demo spins up several threads to send 500,000 transactions
@@ -157,6 +148,120 @@ demo completes after it has convinced itself the testnet won't process any addit
transactions. You should see several TPS measurements printed to the screen. In the transactions. You should see several TPS measurements printed to the screen. In the
multinode variation, you'll see TPS measurements for each validator node as well. multinode variation, you'll see TPS measurements for each validator node as well.
Public Testnet
--------------
In this example the client connects to our public testnet. To run validators on the testnet you would need to open udp ports `8000-10000`.
```bash
$ ./multinode-demo/client.sh --network $(dig +short testnet.solana.com):8001 --identity config-private/client-id.json --duration 60
```
You can observe the effects of your client's transactions on our [dashboard](https://metrics.solana.com:3000/d/testnet/testnet-hud?orgId=2&from=now-30m&to=now&refresh=5s&var-testnet=testnet)
Linux Snap
---
A Linux [Snap](https://snapcraft.io/) is available, which can be used to
easily get Solana running on supported Linux systems without building anything
from source. The `edge` Snap channel is updated daily with the latest
development from the `master` branch. To install:
```bash
$ sudo snap install solana --edge --devmode
```
(`--devmode` flag is required only for `solana.fullnode-cuda`)
Once installed the usual Solana programs will be available as `solona.*` instead
of `solana-*`. For example, `solana.fullnode` instead of `solana-fullnode`.
Update to the latest version at any time with:
```bash
$ snap info solana
$ sudo snap refresh solana --devmode
```
### Daemon support
The snap supports running a leader, validator or leader+drone node as a system
daemon.
Run `sudo snap get solana` to view the current daemon configuration. To view
daemon logs:
1. Run `sudo snap logs -n=all solana` to view the daemon initialization log
2. Runtime logging can be found under `/var/snap/solana/current/leader/`,
`/var/snap/solana/current/validator/`, or `/var/snap/solana/current/drone/` depending
on which `mode=` was selected. Within each log directory the file `current`
contains the latest log, and the files `*.s` (if present) contain older rotated
logs.
Disable the daemon at any time by running:
```bash
$ sudo snap set solana mode=
```
Runtime configuration files for the daemon can be found in
`/var/snap/solana/current/config`.
#### Leader daemon
```bash
$ sudo snap set solana mode=leader
```
If CUDA is available:
```bash
$ sudo snap set solana mode=leader enable-cuda=1
```
`rsync` must be configured and running on the leader.
1. Ensure rsync is installed with `sudo apt-get -y install rsync`
2. Edit `/etc/rsyncd.conf` to include the following
```
[config]
path = /var/snap/solana/current/config
hosts allow = *
read only = true
```
3. Run `sudo systemctl enable rsync; sudo systemctl start rsync`
4. Test by running `rsync -Pzravv rsync://<ip-address-of-leader>/config
solana-config` from another machine. **If the leader is running on a cloud
provider it may be necessary to configure the Firewall rules to permit ingress
to port tcp:873, tcp:9900 and the port range udp:8000-udp:10000**
To run both the Leader and Drone:
```bash
$ sudo snap set solana mode=leader+drone
```
#### Validator daemon
```bash
$ sudo snap set solana mode=validator
```
If CUDA is available:
```bash
$ sudo snap set solana mode=validator enable-cuda=1
```
By default the validator will connect to **testnet.solana.com**, override
the leader IP address by running:
```bash
$ sudo snap set solana mode=validator leader-address=127.0.0.1 #<-- change IP address
```
It's assumed that the leader will be running `rsync` configured as described in
the previous **Leader daemon** section.
Developing Developing
=== ===
@@ -177,6 +282,12 @@ If your rustc version is lower than 1.26.1, please update it:
$ rustup update $ rustup update
``` ```
On Linux systems you may need to install libssl-dev, pkg-config, zlib1g-dev, etc. On Ubuntu:
```bash
$ sudo apt-get install libssl-dev pkg-config zlib1g-dev
```
Download the source code: Download the source code:
```bash ```bash
@@ -194,6 +305,7 @@ $ cargo test
``` ```
To emulate all the tests that will run on a Pull Request, run: To emulate all the tests that will run on a Pull Request, run:
```bash ```bash
$ ./ci/run-local.sh $ ./ci/run-local.sh
``` ```
@@ -202,13 +314,28 @@ Debugging
--- ---
There are some useful debug messages in the code, you can enable them on a per-module and per-level There are some useful debug messages in the code, you can enable them on a per-module and per-level
basis with the normal RUST\_LOG environment variable. Run the fullnode with this syntax: basis. Before running a leader or validator set the normal RUST\_LOG environment variable.
For example, to enable info everywhere and debug only in the solana::banking_stage module:
```bash ```bash
$ RUST_LOG=solana::streamer=debug,solana::server=info cat genesis.log | ./target/release/solana-fullnode > transactions0.log $ export RUST_LOG=info,solana::banking_stage=debug
``` ```
to see the debug and info sections for streamer and server respectively. Generally
we are using debug for infrequent debug messages, trace for potentially frequent messages and Generally we are using debug for infrequent debug messages, trace for potentially frequent
info for performance-related logging. messages and info for performance-related logging.
You can also attach to a running process with GDB. The leader's process is named
_solana-fullnode_:
```bash
$ sudo gdb
attach <PID>
set logging on
thread apply all bt
```
This will dump all the threads stack traces into gdb.txt
Benchmarking Benchmarking
--- ---
@@ -225,15 +352,29 @@ Run the benchmarks:
$ cargo +nightly bench --features="unstable" $ cargo +nightly bench --features="unstable"
``` ```
Release Process
---
The release process for this project is described [here](rfcs/rfc-005-branches-tags-and-channels.md).
Code coverage Code coverage
--- ---
To generate code coverage statistics, run kcov via Docker: To generate code coverage statistics, install cargo-cov. Note: the tool currently only works
in Rust nightly.
```bash ```bash
$ ./ci/coverage.sh $ cargo +nightly install cargo-cov
``` ```
The coverage report will be written to `./target/cov/index.html`
Run cargo-cov and generate a report:
```bash
$ cargo +nightly cov test
$ cargo +nightly cov report --open
```
The coverage report will be written to `./target/cov/report/index.html`
Why coverage? While most see coverage as a code quality metric, we see it primarily as a developer Why coverage? While most see coverage as a code quality metric, we see it primarily as a developer

32
RELEASE.md Normal file

@@ -0,0 +1,32 @@
# Solana Release process
## Introduction
Solana uses a channel-oriented, date-based branching process described [here](https://github.com/solana-labs/solana/blob/master/rfcs/rfc-005-branches-tags-and-channels.md).
## Release Steps
### Changing channels
When cutting a new channel branch these pre-steps are required:
1. Pick your branch point for release on master.
2. Create the branch. The name should be "v" + the first 2 "version" fields from Cargo.toml. For example, a Cargo.toml with version = "0.9.0" implies the next branch name is "v0.9".
3. Update Cargo.toml to the next semantic version (e.g. 0.9.0 -> 0.10.0).
4. Push your new branch to solana.git
5. Land your Carto.toml change as a master PR.
At this point, ci/channel-info.sh should show your freshly cut release branch as "BETA_CHANNEL" and the previous release branch as "STABLE_CHANNEL".
### Updating channels (i.e. "making a release")
We use [github's Releases UI](https://github.com/solana-labs/solana/releases) for tagging a release.
1. Go [there ;)](https://github.com/solana-labs/solana/releases).
2. Click "Draft new release".
3. If the first major release on the branch (e.g. v0.8.0), paste in [this template](https://raw.githubusercontent.com/solana-labs/solana/master/.github/RELEASE_TEMPLATE.md) and fill it in.
4. Test the release by generating a tag using semver's rules. First try at a release should be <branchname>.X-rc.0.
5. Verify release automation:
1. [Crates.io](https://crates.io/crates/solana) should have an updated Solana version.
2. ...
6. After testnet deployment, verify that testnets are running correct software. http://metrics.solana.com should show testnet running on a hash from your newly created branch.

@@ -1 +0,0 @@
theme: jekyll-theme-slate

55
benches/bank.rs Normal file

@@ -0,0 +1,55 @@
#![feature(test)]
extern crate bincode;
extern crate rayon;
extern crate solana;
extern crate test;
use bincode::serialize;
use rayon::prelude::*;
use solana::bank::*;
use solana::hash::hash;
use solana::mint::Mint;
use solana::signature::{Keypair, KeypairUtil};
use solana::system_transaction::SystemTransaction;
use solana::transaction::Transaction;
use test::Bencher;
#[bench]
fn bench_process_transaction(bencher: &mut Bencher) {
let mint = Mint::new(100_000_000);
let bank = Bank::new(&mint);
// Create transactions between unrelated parties.
let transactions: Vec<_> = (0..4096)
.into_par_iter()
.map(|i| {
// Seed the 'from' account.
let rando0 = Keypair::new();
let tx = Transaction::system_move(
&mint.keypair(),
rando0.pubkey(),
10_000,
mint.last_id(),
0,
);
assert!(bank.process_transaction(&tx).is_ok());
// Seed the 'to' account and a cell for its signature.
let last_id = hash(&serialize(&i).unwrap()); // Unique hash
bank.register_entry_id(&last_id);
let rando1 = Keypair::new();
let tx = Transaction::system_move(&rando0, rando1.pubkey(), 1, last_id, 0);
assert!(bank.process_transaction(&tx).is_ok());
// Finally, return the transaction to the benchmark.
tx
}).collect();
bencher.iter(|| {
// Since benchmarker runs this multiple times, we need to clear the signatures.
bank.clear_signatures();
let results = bank.process_transactions(&transactions);
assert!(results.iter().all(Result::is_ok));
})
}

110
benches/banking_stage.rs Normal file

@@ -0,0 +1,110 @@
#![feature(test)]
extern crate bincode;
extern crate rand;
extern crate rayon;
extern crate solana;
extern crate solana_program_interface;
extern crate test;
use rand::{thread_rng, Rng};
use rayon::prelude::*;
use solana::bank::Bank;
use solana::banking_stage::{BankingStage, NUM_THREADS};
use solana::entry::Entry;
use solana::mint::Mint;
use solana::packet::to_packets_chunked;
use solana::signature::{KeypairUtil, Signature};
use solana::system_transaction::SystemTransaction;
use solana::transaction::Transaction;
use solana_program_interface::pubkey::Pubkey;
use std::iter;
use std::sync::mpsc::{channel, Receiver};
use std::sync::Arc;
use std::time::Duration;
use test::Bencher;
fn check_txs(receiver: &Receiver<Vec<Entry>>, ref_tx_count: usize) {
let mut total = 0;
loop {
let entries = receiver.recv_timeout(Duration::new(1, 0));
if let Ok(entries) = entries {
for entry in &entries {
total += entry.transactions.len();
}
} else {
break;
}
if total >= ref_tx_count {
break;
}
}
assert_eq!(total, ref_tx_count);
}
#[bench]
fn bench_banking_stage_multi_accounts(bencher: &mut Bencher) {
let txes = 1000 * NUM_THREADS;
let mint_total = 1_000_000_000_000;
let mint = Mint::new(mint_total);
let (verified_sender, verified_receiver) = channel();
let bank = Arc::new(Bank::new(&mint));
let dummy = Transaction::system_move(
&mint.keypair(),
mint.keypair().pubkey(),
1,
mint.last_id(),
0,
);
let transactions: Vec<_> = (0..txes)
.into_par_iter()
.map(|_| {
let mut new = dummy.clone();
let from: Vec<u8> = (0..64).map(|_| thread_rng().gen()).collect();
let to: Vec<u8> = (0..64).map(|_| thread_rng().gen()).collect();
let sig: Vec<u8> = (0..64).map(|_| thread_rng().gen()).collect();
new.keys[0] = Pubkey::new(&from[0..32]);
new.keys[1] = Pubkey::new(&to[0..32]);
new.signature = Signature::new(&sig[0..64]);
new
}).collect();
// fund all the accounts
transactions.iter().for_each(|tx| {
let fund = Transaction::system_move(
&mint.keypair(),
tx.keys[0],
mint_total / txes as i64,
mint.last_id(),
0,
);
assert!(bank.process_transaction(&fund).is_ok());
});
//sanity check, make sure all the transactions can execute sequentially
transactions.iter().for_each(|tx| {
let res = bank.process_transaction(&tx);
assert!(res.is_ok(), "sanity test transactions");
});
bank.clear_signatures();
//sanity check, make sure all the transactions can execute in parallel
let res = bank.process_transactions(&transactions);
for r in res {
assert!(r.is_ok(), "sanity parallel execution");
}
bank.clear_signatures();
let verified: Vec<_> = to_packets_chunked(&transactions.clone(), 192)
.into_iter()
.map(|x| {
let len = x.read().unwrap().packets.len();
(x, iter::repeat(1).take(len).collect())
}).collect();
let (_stage, signal_receiver) = BankingStage::new(&bank, verified_receiver, Default::default());
bencher.iter(move || {
for v in verified.chunks(verified.len() / NUM_THREADS) {
verified_sender.send(v.to_vec()).unwrap();
}
check_txs(&signal_receiver, txes);
bank.clear_signatures();
// make sure the tx last id is still registered
bank.register_entry_id(&mint.last_id());
});
}

25
benches/ledger.rs Normal file

@@ -0,0 +1,25 @@
#![feature(test)]
extern crate solana;
extern crate test;
use solana::hash::{hash, Hash};
use solana::ledger::{next_entries, reconstruct_entries_from_blobs, Block};
use solana::signature::{Keypair, KeypairUtil};
use solana::system_transaction::SystemTransaction;
use solana::transaction::Transaction;
use test::Bencher;
#[bench]
fn bench_block_to_blobs_to_block(bencher: &mut Bencher) {
let zero = Hash::default();
let one = hash(&zero.as_ref());
let keypair = Keypair::new();
let tx0 = Transaction::system_move(&keypair, keypair.pubkey(), 1, one, 0);
let transactions = vec![tx0; 10];
let entries = next_entries(&zero, 1, transactions);
bencher.iter(|| {
let blobs = entries.to_blobs();
assert_eq!(reconstruct_entries_from_blobs(blobs).unwrap(), entries);
});
}

12
benches/signature.rs Normal file

@@ -0,0 +1,12 @@
#![feature(test)]
extern crate solana;
extern crate test;
use solana::signature::GenKeys;
use test::Bencher;
#[bench]
fn bench_gen_keys(b: &mut Bencher) {
let mut rnd = GenKeys::new([0u8; 32]);
b.iter(|| rnd.gen_n_keypairs(1000));
}

23
benches/sigverify.rs Normal file

@@ -0,0 +1,23 @@
#![feature(test)]
extern crate bincode;
extern crate rayon;
extern crate solana;
extern crate test;
use solana::packet::to_packets;
use solana::sigverify;
use solana::system_transaction::test_tx;
use test::Bencher;
#[bench]
fn bench_sigverify(bencher: &mut Bencher) {
let tx = test_tx();
// generate packet vector
let batches = to_packets(&vec![tx; 128]);
// verify packets
bencher.iter(|| {
let _ans = sigverify::ed25519_verify(&batches);
})
}

@@ -1,15 +1,34 @@
use std::env; use std::env;
use std::fs;
fn main() { fn main() {
println!("cargo:rustc-link-search=native=."); println!("cargo:rerun-if-changed=target/perf-libs");
if !env::var("CARGO_FEATURE_CUDA").is_err() { println!("cargo:rerun-if-changed=build.rs");
// Ensure target/perf-libs/ exists. It's been observed that
// a cargo:rerun-if-changed= directive with a non-existent
// directory triggers a rebuild on every |cargo build| invocation
fs::create_dir("target/perf-libs").unwrap_or_else(|err| {
if err.kind() != std::io::ErrorKind::AlreadyExists {
panic!("Unable to create target/perf-libs: {:?}", err);
}
});
let cuda = !env::var("CARGO_FEATURE_CUDA").is_err();
let erasure = !env::var("CARGO_FEATURE_ERASURE").is_err();
if cuda || erasure {
println!("cargo:rustc-link-search=native=target/perf-libs");
}
if cuda {
println!("cargo:rustc-link-lib=static=cuda_verify_ed25519"); println!("cargo:rustc-link-lib=static=cuda_verify_ed25519");
println!("cargo:rustc-link-search=native=/usr/local/cuda/lib64"); println!("cargo:rustc-link-search=native=/usr/local/cuda/lib64");
println!("cargo:rustc-link-lib=dylib=cudart"); println!("cargo:rustc-link-lib=dylib=cudart");
println!("cargo:rustc-link-lib=dylib=cuda"); println!("cargo:rustc-link-lib=dylib=cuda");
println!("cargo:rustc-link-lib=dylib=cudadevrt"); println!("cargo:rustc-link-lib=dylib=cudadevrt");
} }
if !env::var("CARGO_FEATURE_ERASURE").is_err() { if erasure {
println!("cargo:rustc-link-lib=dylib=Jerasure"); println!("cargo:rustc-link-lib=dylib=Jerasure");
println!("cargo:rustc-link-lib=dylib=gf_complete");
} }
} }

1
ci/.gitignore vendored

@@ -1,2 +1,3 @@
/node_modules/ /node_modules/
/package-lock.json /package-lock.json
/snapcraft.credentials

89
ci/README.md Normal file

@@ -0,0 +1,89 @@
Our CI infrastructure is built around [BuildKite](https://buildkite.com) with some
additional GitHub integration provided by https://github.com/mvines/ci-gate
## Agent Queues
We define two [Agent Queues](https://buildkite.com/docs/agent/v3/queues):
`queue=default` and `queue=cuda`. The `default` queue should be favored and
runs on lower-cost CPU instances. The `cuda` queue is only necessary for
running **tests** that depend on GPU (via CUDA) access -- CUDA builds may still
be run on the `default` queue, and the [buildkite artifact
system](https://buildkite.com/docs/builds/artifacts) used to transfer build
products over to a GPU instance for testing.
## Buildkite Agent Management
### Buildkite GCP Setup
CI runs on Google Cloud Platform via two Compute Engine Instance groups:
`ci-default` and `ci-cuda`. Autoscaling is currently disabled and the number of
VM Instances in each group is manually adjusted.
#### Updating a CI Disk Image
Each Instance group has its own disk image, `ci-default-vX` and
`ci-cuda-vY`, where *X* and *Y* are incremented each time the image is changed.
The process to update a disk image is as follows (TODO: make this less manual):
1. Create a new VM Instance using the disk image to modify.
2. Once the VM boots, ssh to it and modify the disk as desired.
3. Stop the VM Instance running the modified disk. Remember the name of the VM disk
4. From another machine, `gcloud auth login`, then create a new Disk Image based
off the modified VM Instance:
```
$ gcloud compute images create ci-default-$(date +%Y%m%d%H%M) --source-disk xxx --source-disk-zone us-east1-b --family ci-default
```
or
```
$ gcloud compute images create ci-cuda-$(date +%Y%m%d%H%M) --source-disk xxx --source-disk-zone us-east1-b --family ci-cuda
```
5. Delete the new VM instance.
6. Go to the Instance templates tab, find the existing template named
`ci-default-vX` or `ci-cuda-vY` and select it. Use the "Copy" button to create
a new Instance template called `ci-default-vX+1` or `ci-cuda-vY+1` with the
newly created Disk image.
7. Go to the Instance Groups tag and find the applicable group, `ci-default` or
`ci-cuda`. Edit the Instance Group in two steps: (a) Set the number of
instances to 0 and wait for them all to terminate, (b) Update the Instance
template and restore the number of instances to the original value.
8. Clean up the previous version by deleting it from Instance Templates and
Images.
## Reference
### Buildkite AWS CloudFormation Setup
**AWS CloudFormation is currently inactive, although it may be restored in the
future**
AWS CloudFormation can be used to scale machines up and down based on the
current CI load. If no machine is currently running it can take up to 60
seconds to spin up a new instance, please remain calm during this time.
#### AMI
We use a custom AWS AMI built via https://github.com/solana-labs/elastic-ci-stack-for-aws/tree/solana/cuda.
Use the following process to update this AMI as dependencies change:
```bash
$ export AWS_ACCESS_KEY_ID=my_access_key
$ export AWS_SECRET_ACCESS_KEY=my_secret_access_key
$ git clone https://github.com/solana-labs/elastic-ci-stack-for-aws.git -b solana/cuda
$ cd elastic-ci-stack-for-aws/
$ make build
$ make build-ami
```
Watch for the *"amazon-ebs: AMI:"* log message to extract the name of the new
AMI. For example:
```
amazon-ebs: AMI: ami-07118545e8b4ce6dc
```
The new AMI should also now be visible in your EC2 Dashboard. Go to the desired
AWS CloudFormation stack, update the **ImageId** field to the new AMI id, and
*apply* the stack changes.

32
ci/audit.sh Executable file

@@ -0,0 +1,32 @@
#!/bin/bash -e
#
# Audits project dependencies for security vulnerabilities
#
cd "$(dirname "$0")/.."
export RUST_BACKTRACE=1
rustc --version
cargo --version
_() {
echo "--- $*"
"$@"
}
maybe_cargo_install() {
for cmd in "$@"; do
set +e
cargo "$cmd" --help > /dev/null 2>&1
declare exitcode=$?
set -e
if [[ $exitcode -eq 101 ]]; then
_ cargo install cargo-"$cmd"
fi
done
}
maybe_cargo_install audit tree
_ cargo tree
_ cargo audit || true

@@ -0,0 +1,7 @@
steps:
- command: "ci/snap.sh"
timeout_in_minutes: 40
name: "snap [public]"
- command: "ci/docker-solana/build.sh"
timeout_in_minutes: 20
name: "docker-solana"

@@ -1,16 +1,50 @@
steps: steps:
- command: "ci/coverage.sh" - command: "ci/docker-run.sh solanalabs/rust:1.29.1 ci/test-stable.sh"
name: "coverage [public]"
- command: "ci/docker-run.sh rust ci/test-stable.sh"
name: "stable [public]" name: "stable [public]"
- command: "ci/docker-run.sh rustlang/rust:nightly ci/test-nightly.sh || true" env:
name: "nightly - FAILURES IGNORED [public]" CARGO_TARGET_CACHE_NAME: "stable"
- command: "ci/docker-run.sh rust ci/test-ignored.sh" timeout_in_minutes: 30
name: "ignored [public]" - command: "ci/docker-run.sh solanalabs/rust-nightly ci/test-bench.sh"
- command: "ci/test-cuda.sh" name: "bench [public]"
name: "cuda" env:
CARGO_TARGET_CACHE_NAME: "nightly"
timeout_in_minutes: 30
- command: "ci/shellcheck.sh" - command: "ci/shellcheck.sh"
name: "shellcheck [public]" name: "shellcheck [public]"
timeout_in_minutes: 20
- command: "ci/docker-run.sh solanalabs/rust-nightly:2018-09-03 ci/test-nightly.sh || true"
name: "nightly [public]"
env:
CARGO_TARGET_CACHE_NAME: "nightly"
timeout_in_minutes: 30
- command: "ci/test-stable-perf.sh"
name: "stable-perf [public]"
env:
CARGO_TARGET_CACHE_NAME: "stable-perf"
timeout_in_minutes: 20
agents:
- "queue=cuda"
# TODO: Fix and re-enable test-large-network.sh
# - command: "ci/test-large-network.sh || true"
# name: "large-network [public] [ignored]"
# env:
# CARGO_TARGET_CACHE_NAME: "stable"
# timeout_in_minutes: 20
# agents:
# - "queue=large"
- command: "ci/pr-snap.sh"
timeout_in_minutes: 20
name: "snap [public]"
- wait - wait
- command: "ci/publish.sh" - command: "ci/publish-crate.sh"
name: "publish release artifacts" timeout_in_minutes: 20
name: "publish crate [public]"
- trigger: "solana-secondary"
branches: "!pull/*"
async: true
build:
message: "${BUILDKITE_MESSAGE}"
commit: "${BUILDKITE_COMMIT}"
branch: "${BUILDKITE_BRANCH}"
env:
TRIGGERED_BUILDKITE_TAG: "${BUILDKITE_TAG}"

91
ci/channel-info.sh Executable file

@@ -0,0 +1,91 @@
#!/bin/bash
#
# Computes the current branch names of the edge, beta and stable
# channels, as well as the latest tagged release for beta and stable.
#
# stdout of this script may be eval-ed
#
here="$(dirname "$0")"
# shellcheck source=ci/semver_bash/semver.sh
source "$here"/semver_bash/semver.sh
remote=https://github.com/solana-labs/solana.git
# Fetch all vX.Y.Z tags
#
# NOTE: pre-release tags are explicitly ignored
#
# shellcheck disable=SC2207
tags=( \
$(git ls-remote --tags $remote \
| cut -c52- \
| grep '^v[[:digit:]][[:digit:]]*\.[[:digit:]][[:digit:]]*.[[:digit:]][[:digit:]]*$' \
| cut -c2- \
) \
)
# Fetch all the vX.Y branches
#
# shellcheck disable=SC2207
heads=( \
$(git ls-remote --heads $remote \
| cut -c53- \
| grep '^v[[:digit:]][[:digit:]]*\.[[:digit:]][[:digit:]]*$' \
| cut -c2- \
) \
)
# Figure the beta channel by looking for the largest vX.Y branch
beta=
for head in "${heads[@]}"; do
if [[ -n $beta ]]; then
if semverLT "$head.0" "$beta.0"; then
continue
fi
fi
beta=$head
done
# Figure the stable channel by looking for the second largest vX.Y branch
stable=
for head in "${heads[@]}"; do
if [[ $head = "$beta" ]]; then
continue
fi
if [[ -n $stable ]]; then
if semverLT "$head.0" "$stable.0"; then
continue
fi
fi
stable=$head
done
for tag in "${tags[@]}"; do
if [[ -n $beta && $tag = $beta* ]]; then
if [[ -n $beta_tag ]]; then
if semverLT "$tag" "$beta_tag"; then
continue
fi
fi
beta_tag=$tag
fi
if [[ -n $stable && $tag = $stable* ]]; then
if [[ -n $stable_tag ]]; then
if semverLT "$tag" "$stable_tag"; then
continue
fi
fi
stable_tag=$tag
fi
done
echo EDGE_CHANNEL=master
echo BETA_CHANNEL="${beta:+v$beta}"
echo STABLE_CHANNEL="${stable:+v$stable}"
echo BETA_CHANNEL_LATEST_TAG="${beta_tag:+v$beta_tag}"
echo STABLE_CHANNEL_LATEST_TAG="${stable_tag:+v$stable_tag}"
exit 0

@@ -1,21 +0,0 @@
#!/bin/bash -e
cd "$(dirname "$0")/.."
ci/docker-run.sh evilmachines/rust-cargo-kcov \
bash -exc "\
export RUST_BACKTRACE=1; \
cargo build --verbose; \
cargo kcov --lib --verbose; \
"
echo Coverage report:
ls -l target/cov/index.html
if [[ -z "$CODECOV_TOKEN" ]]; then
echo CODECOV_TOKEN undefined
else
bash <(curl -s https://codecov.io/bash)
fi
exit 0

@@ -1,41 +1,85 @@
#!/bin/bash -e #!/bin/bash -e
usage() { usage() {
echo "Usage: $0 [docker image name] [command]" echo "Usage: $0 [--nopull] [docker image name] [command]"
echo echo
echo Runs command in the specified docker image with echo Runs command in the specified docker image with
echo a CI-appropriate environment echo a CI-appropriate environment.
echo
echo "--nopull Skip the dockerhub image update"
echo "--shell Skip command and enter an interactive shell"
echo echo
} }
cd "$(dirname "$0")/.." cd "$(dirname "$0")/.."
INTERACTIVE=false
if [[ $1 = --shell ]]; then
INTERACTIVE=true
shift
fi
NOPULL=false
if [[ $1 = --nopull ]]; then
NOPULL=true
shift
fi
IMAGE="$1" IMAGE="$1"
if [[ -z "$IMAGE" ]]; then if [[ -z "$IMAGE" ]]; then
echo Error: image not defined echo Error: image not defined
exit 1 exit 1
fi fi
docker pull "$IMAGE" $NOPULL || docker pull "$IMAGE"
shift shift
ARGS=(--workdir /solana --volume "$PWD:/solana" --rm) ARGS=(
--workdir /solana
--volume "$PWD:/solana"
--rm
)
ARGS+=(--env "CARGO_HOME=/solana/.cargo") if [[ -n $CI ]]; then
# Share the real ~/.cargo between docker containers in CI for speed
ARGS+=(--volume "$HOME:/home")
else
# Avoid sharing ~/.cargo when building locally to avoid a mixed macOS/Linux
# ~/.cargo
ARGS+=(--volume "$PWD:/home")
fi
ARGS+=(--env "CARGO_HOME=/home/.cargo")
# kcov tries to set the personality of the binary which docker # kcov tries to set the personality of the binary which docker
# doesn't allow by default. # doesn't allow by default.
ARGS+=(--security-opt "seccomp=unconfined") ARGS+=(--security-opt "seccomp=unconfined")
# Ensure files are created with the current host uid/gid # Ensure files are created with the current host uid/gid
ARGS+=(--user "$(id -u):$(id -g)") if [[ -z "$SOLANA_DOCKER_RUN_NOSETUID" ]]; then
ARGS+=(--user "$(id -u):$(id -g)")
fi
# Environment variables to propagate into the container # Environment variables to propagate into the container
ARGS+=( ARGS+=(
--env BUILDKITE
--env BUILDKITE_AGENT_ACCESS_TOKEN
--env BUILDKITE_BRANCH
--env BUILDKITE_JOB_ID
--env BUILDKITE_TAG --env BUILDKITE_TAG
--env CODECOV_TOKEN --env CODECOV_TOKEN
--env CRATES_IO_TOKEN --env CRATES_IO_TOKEN
--env SNAPCRAFT_CREDENTIALS_KEY
) )
if $INTERACTIVE; then
if [[ -n $1 ]]; then
echo
echo "Note: '$*' ignored due to --shell argument"
echo
fi
set -x
exec docker run --interactive --tty "${ARGS[@]}" "$IMAGE" bash
fi
set -x set -x
docker run "${ARGS[@]}" "$IMAGE" "$@" exec docker run "${ARGS[@]}" "$IMAGE" "$@"

@@ -0,0 +1,10 @@
FROM solanalabs/rust
ARG date
RUN set -x && \
rustup install nightly-$date && \
rustup default nightly-$date && \
rustup component add clippy-preview --toolchain=nightly-$date && \
rustc --version && \
cargo --version && \
cargo +nightly-$date install cargo-cov

@@ -0,0 +1,36 @@
Docker image containing rust nightly and some preinstalled crates used in CI.
This image may be manually updated by running `CI=true ./build.sh` if you are a member
of the [Solana Labs](https://hub.docker.com/u/solanalabs/) Docker Hub
organization, but it is also automatically updated periodically by
[this automation](https://buildkite.com/solana-labs/solana-ci-docker-rust-nightly).
## Moving to a newer nightly
We pin the version of nightly (see the `ARG nightly=xyz` line in `Dockerfile`)
to avoid the build breaking at unexpected times, as occasionally nightly will
introduce breaking changes.
To update the pinned version:
1. Run `ci/docker-rust-nightly/build.sh` to rebuild the nightly image locally,
or potentially `ci/docker-rust-nightly/build.sh YYYY-MM-DD` if there's a
specific YYYY-MM-DD that is desired (default is today's build).
1. Run `SOLANA_DOCKER_RUN_NOSETUID=1 ci/docker-run.sh --nopull solanalabs/rust-nightly:YYYY-MM-DD ci/test-nightly.sh`
to confirm the new nightly image builds. Fix any issues as needed
1. Run `docker login` to enable pushing images to Docker Hub, if you're authorized.
1. Run `CI=true ci/docker-rust-nightly/build.sh YYYY-MM-DD` to push the new nightly image to dockerhub.com.
1. Modify the `solanalabs/rust-nightly:YYYY-MM-DD` reference in `ci/buildkite.yml` from the previous to
new *YYYY-MM-DD* value, send a PR with this change and any codebase adjustments needed.
## Troubleshooting
### Resource is denied
When running `CI=true ci/docker-rust-nightly/build.sh`, you see:
```
denied: requested access to the resource is denied
```
Run `docker login` to enable pushing images to Docker Hub. Contact @mvines or @garious
to get write access.

13
ci/docker-rust-nightly/build.sh Executable file

@@ -0,0 +1,13 @@
#!/bin/bash -ex
cd "$(dirname "$0")"
nightlyDate=${1:-$(date +%Y-%m-%d)}
docker build -t solanalabs/rust-nightly:"$nightlyDate" --build-arg date="$nightlyDate" .
maybeEcho=
if [[ -z $CI ]]; then
echo "Not CI, skipping |docker push|"
maybeEcho="echo"
fi
$maybeEcho docker push solanalabs/rust-nightly:"$nightlyDate"

24
ci/docker-rust/Dockerfile Normal file

@@ -0,0 +1,24 @@
# Note: when the rust version is changed also modify
# ci/buildkite.yml to pick up the new image tag
FROM rust:1.29.1
RUN set -x && \
apt update && \
apt-get install apt-transport-https && \
echo deb https://apt.buildkite.com/buildkite-agent stable main > /etc/apt/sources.list.d/buildkite-agent.list && \
echo deb http://apt.llvm.org/xenial/ llvm-toolchain-xenial-6.0 main > /etc/apt/sources.list.d/llvm.list && \
apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-keys 32A37959C2FA5C3C99EFBC32A79206696452D198 && \
wget -O - https://apt.llvm.org/llvm-snapshot.gpg.key | apt-key add - && \
apt update && \
apt install -y \
buildkite-agent \
cmake \
llvm-6.0 \
rsync \
sudo \
&& \
rustup component add rustfmt-preview && \
rustup component add clippy-preview && \
rm -rf /var/lib/apt/lists/* && \
rustc --version && \
cargo --version

6
ci/docker-rust/README.md Normal file

@@ -0,0 +1,6 @@
Docker image containing rust and some preinstalled packages used in CI.
This image may be manually updated by running `./build.sh` if you are a member
of the [Solana Labs](https://hub.docker.com/u/solanalabs/) Docker Hub
organization, but it is also automatically updated periodically by
[this automation](https://buildkite.com/solana-labs/solana-ci-docker-rust).

11
ci/docker-rust/build.sh Executable file

@@ -0,0 +1,11 @@
#!/bin/bash -ex
cd "$(dirname "$0")"
docker build -t solanalabs/rust .
read -r rustc version _ < <(docker run solanalabs/rust rustc --version)
[[ $rustc = rustc ]]
docker tag solanalabs/rust:latest solanalabs/rust:"$version"
docker push solanalabs/rust

@@ -0,0 +1,7 @@
FROM snapcraft/xenial-amd64
# Update snapcraft to latest version
RUN apt-get update -qq \
&& apt-get install -y snapcraft daemontools \
&& rm -rf /var/lib/apt/lists/* \
&& snapcraft --version

6
ci/docker-snapcraft/build.sh Executable file

@@ -0,0 +1,6 @@
#!/bin/bash -ex
cd "$(dirname "$0")"
docker build -t solanalabs/snapcraft .
docker push solanalabs/snapcraft

1
ci/docker-solana/.gitignore vendored Normal file

@@ -0,0 +1 @@
cargo-install/

@@ -0,0 +1,13 @@
FROM debian:stretch
# JSON RPC port
EXPOSE 8899/tcp
# Install libssl
RUN apt update && \
apt-get install -y libssl-dev && \
rm -rf /var/lib/apt/lists/*
COPY usr/bin /usr/bin/
ENTRYPOINT [ "/usr/bin/solana-entrypoint.sh" ]
CMD [""]

@@ -0,0 +1,17 @@
## Minimal Solana Docker image
This image is automatically updated by CI
https://hub.docker.com/r/solanalabs/solana/
### Usage:
Run the latest beta image:
```bash
$ docker run --rm -p 8899:8899 solanalabs/solana:beta
```
Run the latest edge image:
```bash
$ docker run --rm -p 8899:8899 solanalabs/solana:edge
```
Port *8899* is the JSON RPC port, which is used by clients to communicate with the network.

38
ci/docker-solana/build.sh Executable file

@@ -0,0 +1,38 @@
#!/bin/bash -ex
cd "$(dirname "$0")"
eval "$(../channel-info.sh)"
if [[ $BUILDKITE_BRANCH = "$STABLE_CHANNEL" ]]; then
CHANNEL=stable
elif [[ $BUILDKITE_BRANCH = "$EDGE_CHANNEL" ]]; then
CHANNEL=edge
elif [[ $BUILDKITE_BRANCH = "$BETA_CHANNEL" ]]; then
CHANNEL=beta
fi
if [[ -z $CHANNEL ]]; then
echo Unable to determine channel to publish into, exiting.
exit 0
fi
rm -rf usr/
../docker-run.sh solanalabs/rust:1.29.1 \
cargo install --path . --root ci/docker-solana/usr
cp -f entrypoint.sh usr/bin/solana-entrypoint.sh
docker build -t solanalabs/solana:$CHANNEL .
maybeEcho=
if [[ -z $CI ]]; then
echo "Not CI, skipping |docker push|"
maybeEcho="echo"
else
(
set +x
if [[ -n $DOCKER_PASSWORD && -n $DOCKER_USERNAME ]]; then
echo "$DOCKER_PASSWORD" | docker login --username "$DOCKER_USERNAME" --password-stdin
fi
)
fi
$maybeEcho docker push solanalabs/solana:$CHANNEL

23
ci/docker-solana/entrypoint.sh Executable file

@@ -0,0 +1,23 @@
#!/bin/bash -ex
export RUST_LOG=solana=info
export RUST_BACKTRACE=1
solana-keygen -o /config/leader-keypair.json
solana-keygen -o /config/drone-keypair.json
solana-genesis --tokens=1000000000 --ledger /ledger < /config/drone-keypair.json
solana-fullnode-config --keypair=/config/leader-keypair.json -l > /config/leader-config.json
solana-drone --keypair /config/drone-keypair.json --network 127.0.0.1:8001 &
drone=$!
solana-fullnode --identity /config/leader-config.json --ledger /ledger/ &
fullnode=$!
abort() {
kill "$drone" "$fullnode"
}
trap abort SIGINT SIGTERM
wait "$fullnode"
kill "$drone" "$fullnode"

81
ci/hoover.sh Executable file

@@ -0,0 +1,81 @@
#!/bin/bash
#
# Regular maintenance performed on a buildkite agent to control disk usage
#
echo --- Delete all exited containers first
(
set -x
exited=$(docker ps -aq --no-trunc --filter "status=exited")
if [[ -n "$exited" ]]; then
# shellcheck disable=SC2086 # Don't want to double quote "$exited"
docker rm $exited
fi
)
echo --- Delete untagged images
(
set -x
untagged=$(docker images | grep '<none>'| awk '{ print $3 }')
if [[ -n "$untagged" ]]; then
# shellcheck disable=SC2086 # Don't want to double quote "$untagged"
docker rmi $untagged
fi
)
echo --- Delete all dangling images
(
set -x
dangling=$(docker images --filter 'dangling=true' -q --no-trunc | sort | uniq)
if [[ -n "$dangling" ]]; then
# shellcheck disable=SC2086 # Don't want to double quote "$dangling"
docker rmi $dangling
fi
)
echo --- Remove unused docker networks
(
set -x
docker network prune -f
)
echo "--- Delete /tmp files older than 1 day owned by $(id -un)"
(
set -x
find /tmp -maxdepth 1 -user "$(id -un)" -mtime +1 -print0 | xargs -0 rm -rf
)
echo --- Deleting stale buildkite agent build directories
if [[ ! -d ../../../../builds/$BUILDKITE_AGENT_NAME ]]; then
# We might not be where we think we are, do nothing
echo Warning: Skipping flush of stale agent build directories
echo " PWD=$PWD"
else
# NOTE: this will be horribly broken if we ever decide to run multiple
# agents on the same machine.
(
for keepDir in "$BUILDKITE_PIPELINE_SLUG" \
"$BUILDKITE_ORGANIZATION_SLUG" \
"$BUILDKITE_AGENT_NAME"; do
cd .. || exit 1
for dir in *; do
if [[ -d $dir && $dir != "$keepDir" ]]; then
echo "Removing $dir"
rm -rf "${dir:?}"/
fi
done
done
)
fi
echo --- System Status
(
set -x
docker images
docker ps
docker network ls
df -h
)
exit 0

8
ci/is-pr.sh Executable file

@@ -0,0 +1,8 @@
#!/bin/bash -e
#
# The standard BUILDKITE_PULL_REQUEST environment variable is always "false" due
# to how solana-ci-gate is used to trigger PR builds rather than using the
# standard Buildkite PR trigger.
#
[[ $BUILDKITE_BRANCH =~ pull/* ]]

93
ci/localnet-sanity.sh Executable file

@@ -0,0 +1,93 @@
#!/bin/bash -e
#
# Perform a quick sanity test on a leader, drone, validator and client running
# locally on the same machine
#
cd "$(dirname "$0")"/..
source ci/upload_ci_artifact.sh
source scripts/configure-metrics.sh
multinode-demo/setup.sh
backgroundCommands="drone leader validator validator-x"
pids=()
for cmd in $backgroundCommands; do
echo "--- Start $cmd"
rm -f log-"$cmd".txt
multinode-demo/"$cmd".sh > log-"$cmd".txt 2>&1 &
declare pid=$!
pids+=("$pid")
echo "pid: $pid"
done
killBackgroundCommands() {
set +e
for pid in "${pids[@]}"; do
if kill "$pid"; then
wait "$pid"
else
echo -e "^^^ +++\\nWarning: unable to kill $pid"
fi
done
set -e
pids=()
}
shutdown() {
exitcode=$?
killBackgroundCommands
set +e
echo "--- Upload artifacts"
for cmd in $backgroundCommands; do
declare logfile=log-$cmd.txt
upload_ci_artifact "$logfile"
tail "$logfile"
done
exit $exitcode
}
trap shutdown EXIT INT
set -e
flag_error() {
echo Failed
echo "^^^ +++"
exit 1
}
echo "--- Wallet sanity"
(
set -x
scripts/wallet-sanity.sh
) || flag_error
echo "--- Node count"
(
source multinode-demo/common.sh
set -x
client_id=/tmp/client-id.json-$$
$solana_keygen -o $client_id
$solana_bench_tps --identity $client_id --num-nodes 3 --reject-extra-nodes --converge-only
rm -rf $client_id
) || flag_error
killBackgroundCommands
echo "--- Ledger verification"
(
source multinode-demo/common.sh
set -x
cp -R "$SOLANA_CONFIG_DIR"/ledger /tmp/ledger-$$
$solana_ledger_tool --ledger /tmp/ledger-$$ verify
rm -rf /tmp/ledger-$$
) || flag_error
echo +++
echo Ok
exit 0

18
ci/pr-snap.sh Executable file

@@ -0,0 +1,18 @@
#!/bin/bash -e
#
# Only run snap.sh for pull requests that modify files under /snap
#
cd "$(dirname "$0")"
if ./is-pr.sh; then
affected_files="$(buildkite-agent meta-data get affected_files)"
echo "Affected files in this PR: $affected_files"
if [[ ! ":$affected_files:" =~ :snap/ ]]; then
echo "Skipping snap build as no files under /snap were modified"
exit 0
fi
exec ./snap.sh
else
echo "Skipping snap build as this is not a pull request"
fi

@@ -16,4 +16,4 @@ if [[ ! -x $BKRUN ]]; then
fi fi
set -x set -x
./ci/node_modules/.bin/bkrun ci/buildkite.yml exec ./ci/node_modules/.bin/bkrun ci/buildkite.yml

26
ci/semver_bash/LICENSE Normal file

@@ -0,0 +1,26 @@
Copyright (c) 2013, Ray Bejjani
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
1. Redistributions of source code must retain the above copyright notice, this
list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
The views and conclusions contained in the software and documentation are those
of the authors and should not be interpreted as representing official policies,
either expressed or implied, of the FreeBSD Project.

31
ci/semver_bash/README.md Normal file

@@ -0,0 +1,31 @@
semver_bash is a bash parser for semantic versioning
====================================================
[Semantic Versioning](http://semver.org/) is a set of guidelines that help keep
version and version management sane. This is a bash based parser to help manage
a project's versions. Use it from a Makefile or any scripts you use in your
project.
Usage
-----
semver_bash can be used from the command line as:
$ ./semver.sh "3.2.1" "3.2.1-alpha"
3.2.1 -> M: 3 m:2 p:1 s:
3.2.1-alpha -> M: 3 m:2 p:1 s:-alpha
3.2.1 == 3.2.1-alpha -> 1.
3.2.1 < 3.2.1-alpha -> 1.
3.2.1 > 3.2.1-alpha -> 0.
Alternatively, you can source it from within a script:
. ./semver.sh
local MAJOR=0
local MINOR=0
local PATCH=0
local SPECIAL=""
semverParseInto "1.2.3" MAJOR MINOR PATCH SPECIAL
semverParseInto "3.2.1" MAJOR MINOR PATCH SPECIAL

130
ci/semver_bash/semver.sh Executable file

@@ -0,0 +1,130 @@
#!/usr/bin/env sh
function semverParseInto() {
local RE='[^0-9]*\([0-9]*\)[.]\([0-9]*\)[.]\([0-9]*\)\([0-9A-Za-z-]*\)'
#MAJOR
eval $2=`echo $1 | sed -e "s#$RE#\1#"`
#MINOR
eval $3=`echo $1 | sed -e "s#$RE#\2#"`
#MINOR
eval $4=`echo $1 | sed -e "s#$RE#\3#"`
#SPECIAL
eval $5=`echo $1 | sed -e "s#$RE#\4#"`
}
function semverEQ() {
local MAJOR_A=0
local MINOR_A=0
local PATCH_A=0
local SPECIAL_A=0
local MAJOR_B=0
local MINOR_B=0
local PATCH_B=0
local SPECIAL_B=0
semverParseInto $1 MAJOR_A MINOR_A PATCH_A SPECIAL_A
semverParseInto $2 MAJOR_B MINOR_B PATCH_B SPECIAL_B
if [ $MAJOR_A -ne $MAJOR_B ]; then
return 1
fi
if [ $MINOR_A -ne $MINOR_B ]; then
return 1
fi
if [ $PATCH_A -ne $PATCH_B ]; then
return 1
fi
if [[ "_$SPECIAL_A" != "_$SPECIAL_B" ]]; then
return 1
fi
return 0
}
function semverLT() {
local MAJOR_A=0
local MINOR_A=0
local PATCH_A=0
local SPECIAL_A=0
local MAJOR_B=0
local MINOR_B=0
local PATCH_B=0
local SPECIAL_B=0
semverParseInto $1 MAJOR_A MINOR_A PATCH_A SPECIAL_A
semverParseInto $2 MAJOR_B MINOR_B PATCH_B SPECIAL_B
if [ $MAJOR_A -lt $MAJOR_B ]; then
return 0
fi
if [[ $MAJOR_A -le $MAJOR_B && $MINOR_A -lt $MINOR_B ]]; then
return 0
fi
if [[ $MAJOR_A -le $MAJOR_B && $MINOR_A -le $MINOR_B && $PATCH_A -lt $PATCH_B ]]; then
return 0
fi
if [[ "_$SPECIAL_A" == "_" ]] && [[ "_$SPECIAL_B" == "_" ]] ; then
return 1
fi
if [[ "_$SPECIAL_A" == "_" ]] && [[ "_$SPECIAL_B" != "_" ]] ; then
return 1
fi
if [[ "_$SPECIAL_A" != "_" ]] && [[ "_$SPECIAL_B" == "_" ]] ; then
return 0
fi
if [[ "_$SPECIAL_A" < "_$SPECIAL_B" ]]; then
return 0
fi
return 1
}
function semverGT() {
semverEQ $1 $2
local EQ=$?
semverLT $1 $2
local LT=$?
if [ $EQ -ne 0 ] && [ $LT -ne 0 ]; then
return 0
else
return 1
fi
}
if [ "___semver.sh" == "___`basename $0`" ]; then
MAJOR=0
MINOR=0
PATCH=0
SPECIAL=""
semverParseInto $1 MAJOR MINOR PATCH SPECIAL
echo "$1 -> M: $MAJOR m:$MINOR p:$PATCH s:$SPECIAL"
semverParseInto $2 MAJOR MINOR PATCH SPECIAL
echo "$2 -> M: $MAJOR m:$MINOR p:$PATCH s:$SPECIAL"
semverEQ $1 $2
echo "$1 == $2 -> $?."
semverLT $1 $2
echo "$1 < $2 -> $?."
semverGT $1 $2
echo "$1 > $2 -> $?."
fi

151
ci/semver_bash/semver_test.sh Executable file

@@ -0,0 +1,151 @@
#!/usr/bin/env bash
. ./semver.sh
semverTest() {
local A=R1.3.2
local B=R2.3.2
local C=R1.4.2
local D=R1.3.3
local E=R1.3.2a
local F=R1.3.2b
local G=R1.2.3
local MAJOR=0
local MINOR=0
local PATCH=0
local SPECIAL=""
semverParseInto $A MAJOR MINOR PATCH SPECIAL
echo "$A -> M:$MAJOR m:$MINOR p:$PATCH s:$SPECIAL. Expect M:1 m:3 p:2 s:"
semverParseInto $E MAJOR MINOR PATCH SPECIAL
echo "$E -> M:$MAJOR m:$MINOR p:$PATCH s:$SPECIAL. Expect M:1 m:3 p:2 s:a"
echo "Equality comparisions"
semverEQ $A $A
echo "$A == $A -> $?. Expect 0."
semverLT $A $A
echo "$A < $A -> $?. Expect 1."
semverGT $A $A
echo "$A > $A -> $?. Expect 1."
echo "Major number comparisions"
semverEQ $A $B
echo "$A == $B -> $?. Expect 1."
semverLT $A $B
echo "$A < $B -> $?. Expect 0."
semverGT $A $B
echo "$A > $B -> $?. Expect 1."
semverEQ $B $A
echo "$B == $A -> $?. Expect 1."
semverLT $B $A
echo "$B < $A -> $?. Expect 1."
semverGT $B $A
echo "$B > $A -> $?. Expect 0."
echo "Minor number comparisions"
semverEQ $A $C
echo "$A == $C -> $?. Expect 1."
semverLT $A $C
echo "$A < $C -> $?. Expect 0."
semverGT $A $C
echo "$A > $C -> $?. Expect 1."
semverEQ $C $A
echo "$C == $A -> $?. Expect 1."
semverLT $C $A
echo "$C < $A -> $?. Expect 1."
semverGT $C $A
echo "$C > $A -> $?. Expect 0."
echo "patch number comparisions"
semverEQ $A $D
echo "$A == $D -> $?. Expect 1."
semverLT $A $D
echo "$A < $D -> $?. Expect 0."
semverGT $A $D
echo "$A > $D -> $?. Expect 1."
semverEQ $D $A
echo "$D == $A -> $?. Expect 1."
semverLT $D $A
echo "$D < $A -> $?. Expect 1."
semverGT $D $A
echo "$D > $A -> $?. Expect 0."
echo "special section vs no special comparisions"
semverEQ $A $E
echo "$A == $E -> $?. Expect 1."
semverLT $A $E
echo "$A < $E -> $?. Expect 1."
semverGT $A $E
echo "$A > $E -> $?. Expect 0."
semverEQ $E $A
echo "$E == $A -> $?. Expect 1."
semverLT $E $A
echo "$E < $A -> $?. Expect 0."
semverGT $E $A
echo "$E > $A -> $?. Expect 1."
echo "special section vs special comparisions"
semverEQ $E $F
echo "$E == $F -> $?. Expect 1."
semverLT $E $F
echo "$E < $F -> $?. Expect 0."
semverGT $E $F
echo "$E > $F -> $?. Expect 1."
semverEQ $F $E
echo "$F == $E -> $?. Expect 1."
semverLT $F $E
echo "$F < $E -> $?. Expect 1."
semverGT $F $E
echo "$F > $E -> $?. Expect 0."
echo "Minor and patch number comparisons"
semverEQ $A $G
echo "$A == $G -> $?. Expect 1."
semverLT $A $G
echo "$A < $G -> $?. Expect 1."
semverGT $A $G
echo "$A > $G -> $?. Expect 0."
semverEQ $G $A
echo "$G == $A -> $?. Expect 1."
semverLT $G $A
echo "$G < $A -> $?. Expect 0."
semverGT $G $A
echo "$G > $A -> $?. Expect 1."
}
semverTest

@@ -5,7 +5,13 @@
cd "$(dirname "$0")/.." cd "$(dirname "$0")/.."
set -x set -x
find . -name "*.sh" -not -regex ".*/.cargo/.*" -not -regex ".*/node_modules/.*" -print0 \ find . -name "*.sh" \
-not -regex ".*/ci/semver_bash/.*" \
-not -regex ".*/.cargo/.*" \
-not -regex ".*/node_modules/.*" \
-not -regex ".*/target/.*" \
-print0 \
| xargs -0 \ | xargs -0 \
ci/docker-run.sh koalaman/shellcheck --color=always --external-sources --shell=bash ci/docker-run.sh koalaman/shellcheck --color=always --external-sources --shell=bash
exit 0 exit 0

61
ci/snap.sh Executable file

@@ -0,0 +1,61 @@
#!/bin/bash -e
cd "$(dirname "$0")/.."
DRYRUN=
if [[ -z $BUILDKITE_BRANCH ]] || ./ci/is-pr.sh; then
DRYRUN="echo"
fi
eval "$(ci/channel-info.sh)"
if [[ $BUILDKITE_BRANCH = "$STABLE_CHANNEL" ]]; then
CHANNEL=stable
elif [[ $BUILDKITE_BRANCH = "$EDGE_CHANNEL" ]]; then
CHANNEL=edge
elif [[ $BUILDKITE_BRANCH = "$BETA_CHANNEL" ]]; then
CHANNEL=beta
fi
if [[ -z $CHANNEL ]]; then
echo Unable to determine channel to publish into, exiting.
exit 0
fi
if [[ -z $DRYRUN ]]; then
[[ -n $SNAPCRAFT_CREDENTIALS_KEY ]] || {
echo SNAPCRAFT_CREDENTIALS_KEY not defined
exit 1;
}
(
openssl aes-256-cbc -d \
-in ci/snapcraft.credentials.enc \
-out ci/snapcraft.credentials \
-k "$SNAPCRAFT_CREDENTIALS_KEY"
snapcraft login --with ci/snapcraft.credentials
) || {
rm -f ci/snapcraft.credentials;
exit 1
}
fi
set -x
echo --- checking for multilog
if [[ ! -x /usr/bin/multilog ]]; then
if [[ -z $CI ]]; then
echo "multilog not found, install with: sudo apt-get install -y daemontools"
exit 1
fi
sudo apt-get install -y daemontools
fi
echo --- build: $CHANNEL channel
snapcraft
source ci/upload_ci_artifact.sh
upload_ci_artifact solana_*.snap
echo --- publish: $CHANNEL channel
$DRYRUN snapcraft push solana_*.snap --release $CHANNEL

Binary file not shown.

17
ci/test-bench.sh Executable file

@@ -0,0 +1,17 @@
#!/bin/bash -e
cd "$(dirname "$0")/.."
ci/version-check.sh nightly
export RUST_BACKTRACE=1
_() {
echo "--- $*"
"$@"
}
set -o pipefail
BENCH_FILE=bench_output.log
_ cargo bench --features=unstable --verbose -- -Z unstable-options --format=json | tee $BENCH_FILE
_ cargo run --release --bin solana-upload-perf -- $BENCH_FILE

@@ -1,22 +0,0 @@
#!/bin/bash -e
cd "$(dirname "$0")/.."
LIB=libcuda_verify_ed25519.a
if [[ ! -r $LIB ]]; then
if [[ -z "${libcuda_verify_ed25519_URL:-}" ]]; then
echo "$0 skipped. Unable to locate $LIB"
exit 0
fi
export LD_LIBRARY_PATH=/usr/local/cuda/lib64
export PATH=$PATH:/usr/local/cuda/bin
curl -X GET -o $LIB "$libcuda_verify_ed25519_URL"
fi
# shellcheck disable=SC1090 # <-- shellcheck can't follow ~
source ~/.cargo/env
export RUST_BACKTRACE=1
cargo test --features=cuda
exit 0

@@ -1,9 +0,0 @@
#!/bin/bash -e
cd "$(dirname "$0")/.."
rustc --version
cargo --version
export RUST_BACKTRACE=1
cargo test -- --ignored

45
ci/test-large-network.sh Executable file

@@ -0,0 +1,45 @@
#!/bin/bash -e
here=$(dirname "$0")
cd "$here"/..
if ! ci/version-check.sh stable; then
# This job doesn't run within a container, try once to upgrade tooling on a
# version check failure
rustup install stable
ci/version-check.sh stable
fi
export RUST_BACKTRACE=1
./fetch-perf-libs.sh
export LD_LIBRARY_PATH=$PWD/target/perf-libs:$LD_LIBRARY_PATH
export RUST_LOG=multinode=info
if [[ $(ulimit -n) -lt 65000 ]]; then
echo 'Error: nofiles too small, run "ulimit -n 65000" to continue'
exit 1
fi
if [[ $(sysctl -n net.core.rmem_default) -lt 1610612736 ]]; then
echo 'Error: rmem_default too small, run "sudo sysctl -w net.core.rmem_default=1610612736" to continue'
exit 1
fi
if [[ $(sysctl -n net.core.rmem_max) -lt 1610612736 ]]; then
echo 'Error: rmem_max too small, run "sudo sysctl -w net.core.rmem_max=1610612736" to continue'
exit 1
fi
if [[ $(sysctl -n net.core.wmem_default) -lt 1610612736 ]]; then
echo 'Error: rmem_default too small, run "sudo sysctl -w net.core.wmem_default=1610612736" to continue'
exit 1
fi
if [[ $(sysctl -n net.core.wmem_max) -lt 1610612736 ]]; then
echo 'Error: rmem_max too small, run "sudo sysctl -w net.core.wmem_max=1610612736" to continue'
exit 1
fi
set -x
exec cargo test --release --features=erasure test_multi_node_dynamic_network -- --ignored

@@ -2,13 +2,28 @@
cd "$(dirname "$0")/.." cd "$(dirname "$0")/.."
rustc --version ci/version-check.sh nightly
cargo --version
export RUST_BACKTRACE=1 export RUST_BACKTRACE=1
rustup component add rustfmt-preview
cargo build --verbose --features unstable _() {
cargo test --verbose --features unstable echo "--- $*"
cargo bench --verbose --features unstable "$@"
}
_ cargo build --verbose --features unstable
_ cargo test --verbose --features=unstable
exit 0 exit 0
# Coverage disabled (see issue #433)
_ cargo cov test
_ cargo cov report
echo --- Coverage report:
ls -l target/cov/report/index.html
if [[ -z "$CODECOV_TOKEN" ]]; then
echo CODECOV_TOKEN undefined
else
bash <(curl -s https://codecov.io/bash) -x 'llvm-cov-6.0 gcov'
fi

31
ci/test-stable-perf.sh Executable file

@@ -0,0 +1,31 @@
#!/bin/bash -e
cd "$(dirname "$0")/.."
if ! ci/version-check.sh stable; then
# This job doesn't run within a container, try once to upgrade tooling on a
# version check failure
rustup install stable
ci/version-check.sh stable
fi
export RUST_BACKTRACE=1
export RUSTFLAGS="-D warnings"
./fetch-perf-libs.sh
export LD_LIBRARY_PATH=$PWD/target/perf-libs:/usr/local/cuda/lib64:$LD_LIBRARY_PATH
export PATH=$PATH:/usr/local/cuda/bin
_() {
echo "--- $*"
"$@"
}
_ cargo test --features=cuda,erasure
echo --- ci/localnet-sanity.sh
(
set -x
# Assume |cargo build| has populated target/debug/ successfully.
export PATH=$PWD/target/debug:$PATH
USE_INSTALL=1 ci/localnet-sanity.sh
)

@@ -2,13 +2,25 @@
cd "$(dirname "$0")/.." cd "$(dirname "$0")/.."
rustc --version ci/version-check.sh stable
cargo --version
export RUST_BACKTRACE=1 export RUST_BACKTRACE=1
rustup component add rustfmt-preview export RUSTFLAGS="-D warnings"
cargo fmt -- --write-mode=diff
cargo build --verbose
cargo test --verbose
exit 0 _() {
echo "--- $*"
"$@"
}
_ cargo fmt -- --check
_ cargo build --verbose
_ cargo test --verbose
echo --- ci/localnet-sanity.sh
(
set -x
# Assume |cargo build| has populated target/debug/ successfully.
export PATH=$PWD/target/debug:$PATH
USE_INSTALL=1 ci/localnet-sanity.sh
)
_ ci/audit.sh || true

136
ci/testnet-deploy.sh Executable file

@@ -0,0 +1,136 @@
#!/bin/bash -e
cd "$(dirname "$0")"/..
zone=
leaderAddress=
leaderMachineType=
clientNodeCount=0
validatorNodeCount=10
publicNetwork=false
snapChannel=edge
delete=false
enableGpu=false
usage() {
exitcode=0
if [[ -n "$1" ]]; then
exitcode=1
echo "Error: $*"
fi
cat <<EOF
usage: $0 [name] [zone] [options...]
Deploys a CD testnet
name - name of the network
zone - zone to deploy the network into
options:
-s edge|beta|stable - Deploy the specified Snap release channel
(default: $snapChannel)
-n [number] - Number of validator nodes (default: $validatorNodeCount)
-c [number] - Number of client nodes (default: $clientNodeCount)
-P - Use public network IP addresses (default: $publicNetwork)
-G - Enable GPU, and set count/type of GPUs to use (e.g n1-standard-16 --accelerator count=4,type=nvidia-tesla-k80)
-g - Enable GPU (default: $enableGpu)
-a [address] - Set the leader node's external IP address to this GCE address
-d - Delete the network
Note: the SOLANA_METRICS_CONFIG environment variable is used to configure
metrics
EOF
exit $exitcode
}
netName=$1
zone=$2
[[ -n $netName ]] || usage
[[ -n $zone ]] || usage "Zone not specified"
shift 2
while getopts "h?p:Pn:c:s:gG:a:d" opt; do
case $opt in
h | \?)
usage
;;
P)
publicNetwork=true
;;
n)
validatorNodeCount=$OPTARG
;;
c)
clientNodeCount=$OPTARG
;;
s)
case $OPTARG in
edge|beta|stable)
snapChannel=$OPTARG
;;
*)
usage "Invalid snap channel: $OPTARG"
;;
esac
;;
g)
enableGpu=true
;;
G)
enableGpu=true
leaderMachineType=$OPTARG
;;
a)
leaderAddress=$OPTARG
;;
d)
delete=true
;;
*)
usage "Error: unhandled option: $opt"
;;
esac
done
gce_create_args=(
-a "$leaderAddress"
-c "$clientNodeCount"
-n "$validatorNodeCount"
-p "$netName"
-z "$zone"
)
if $enableGpu; then
if [[ -z $leaderMachineType ]]; then
gce_create_args+=(-g)
else
gce_create_args+=(-G "$leaderMachineType")
fi
fi
if $publicNetwork; then
gce_create_args+=(-P)
fi
set -x
echo --- gce.sh delete
time net/gce.sh delete -z "$zone" -p "$netName"
if $delete; then
exit 0
fi
echo --- gce.sh create
time net/gce.sh create "${gce_create_args[@]}"
net/init-metrics.sh -e
echo --- net.sh start
maybeRejectExtraNodes=
if ! $publicNetwork; then
maybeRejectExtraNodes="-o rejectExtraNodes"
fi
# shellcheck disable=SC2086 # Don't want to double quote maybeRejectExtraNodes
time net/net.sh start -s "$snapChannel" $maybeRejectExtraNodes
exit 0

40
ci/testnet-sanity.sh Executable file

@@ -0,0 +1,40 @@
#!/bin/bash -e
cd "$(dirname "$0")/.."
usage() {
exitcode=0
if [[ -n "$1" ]]; then
exitcode=1
echo "Error: $*"
fi
cat <<EOF
usage: $0 [name] [zone]
Sanity check a CD testnet
name - name of the network
zone - zone of the network
Note: the SOLANA_METRICS_CONFIG environment variable is used to configure
metrics
EOF
exit $exitcode
}
netName=$1
zone=$2
[[ -n $netName ]] || usage ""
[[ -n $zone ]] || usage "Zone not specified"
set -x
echo --- gce.sh config
net/gce.sh config -p "$netName" -z "$zone"
net/init-metrics.sh -e
echo --- net.sh sanity
net/net.sh sanity \
${NO_LEDGER_VERIFY:+-o noLedgerVerify} \
${NO_VALIDATOR_SANITY:+-o noValidatorSanity} \
${REJECT_EXTRA_NODES:+-o rejectExtraNodes} \
exit 0

18
ci/upload_ci_artifact.sh Normal file

@@ -0,0 +1,18 @@
# |source| me
upload_ci_artifact() {
echo "--- artifact: $1"
if [[ -r "$1" ]]; then
ls -l "$1"
if ${BUILDKITE:-false}; then
(
set -x
buildkite-agent artifact upload "$1"
)
fi
else
echo ^^^ +++
echo "$1 not found"
fi
}

35
ci/version-check.sh Executable file

@@ -0,0 +1,35 @@
#!/bin/bash -e
require() {
declare expectedProgram="$1"
declare expectedVersion="$2"
read -r program version _ < <($expectedProgram -V)
declare ok=true
[[ $program = "$expectedProgram" ]] || ok=false
[[ $version =~ $expectedVersion ]] || ok=false
echo "Found $program $version"
if ! $ok; then
echo Error: expected "$expectedProgram $expectedVersion"
exit 1
fi
}
case ${1:-stable} in
nightly)
require rustc 1.30.[0-9]+-nightly
require cargo 1.29.[0-9]+-nightly
;;
stable)
require rustc 1.29.[0-9]+
require cargo 1.29.[0-9]+
;;
*)
echo Error: unknown argument: "$1"
exit 1
;;
esac
exit 0

22
common/Cargo.toml Normal file

@@ -0,0 +1,22 @@
[package]
name = "solana_program_interface"
version = "0.1.0"
authors = [
"Anatoly Yakovenko <anatoly@solana.com>",
"Greg Fitzgerald <greg@solana.com>",
"Stephen Akridge <stephen@solana.com>",
"Michael Vines <mvines@solana.com>",
"Rob Walker <rob@solana.com>",
"Pankaj Garg <pankaj@solana.com>",
"Tyera Eulberg <tyera@solana.com>",
"Jack May <jack@solana.com>",
]
[dependencies]
bincode = "1.0.0"
bs58 = "0.2.0"
generic-array = { version = "0.12.0", default-features = false, features = ["serde"] }
serde = "1.0.27"
serde_derive = "1.0.27"

29
common/src/account.rs Normal file

@@ -0,0 +1,29 @@
use pubkey::Pubkey;
/// An Account with userdata that is stored on chain
#[derive(Serialize, Deserialize, Debug, Clone, Default)]
pub struct Account {
/// tokens in the account
pub tokens: i64,
/// user data
/// A transaction can write to its userdata
pub userdata: Vec<u8>,
/// contract id this contract belongs to
pub program_id: Pubkey,
}
impl Account {
pub fn new(tokens: i64, space: usize, program_id: Pubkey) -> Account {
Account {
tokens,
userdata: vec![0u8; space],
program_id,
}
}
}
#[derive(Debug)]
pub struct KeyedAccount<'a> {
pub key: &'a Pubkey,
pub account: &'a mut Account,
}

7
common/src/lib.rs Normal file

@@ -0,0 +1,7 @@
pub mod account;
pub mod pubkey;
extern crate bincode;
extern crate bs58;
extern crate generic_array;
#[macro_use]
extern crate serde_derive;

31
common/src/pubkey.rs Normal file

@@ -0,0 +1,31 @@
use bs58;
use generic_array::typenum::U32;
use generic_array::GenericArray;
use std::fmt;
#[derive(Serialize, Deserialize, Clone, Copy, Default, Eq, PartialEq, Ord, PartialOrd, Hash)]
pub struct Pubkey(GenericArray<u8, U32>);
impl Pubkey {
pub fn new(pubkey_vec: &[u8]) -> Self {
Pubkey(GenericArray::clone_from_slice(&pubkey_vec))
}
}
impl AsRef<[u8]> for Pubkey {
fn as_ref(&self) -> &[u8] {
&self.0[..]
}
}
impl fmt::Debug for Pubkey {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(f, "{}", bs58::encode(self.0).into_string())
}
}
impl fmt::Display for Pubkey {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(f, "{}", bs58::encode(self.0).into_string())
}
}

@@ -1,65 +0,0 @@
The Historian
===
Create a *Historian* and send it *events* to generate an *event log*, where each *entry*
is tagged with the historian's latest *hash*. Then ensure the order of events was not tampered
with by verifying each entry's hash can be generated from the hash in the previous entry:
![historian](https://user-images.githubusercontent.com/55449/36950845-459bdb58-1fb9-11e8-850e-894586f3729b.png)
```rust
extern crate solana;
use solana::historian::Historian;
use solana::ledger::{Block, Entry, Hash};
use solana::event::{generate_keypair, get_pubkey, sign_claim_data, Event};
use std::thread::sleep;
use std::time::Duration;
use std::sync::mpsc::SendError;
fn create_ledger(hist: &Historian<Hash>) -> Result<(), SendError<Event<Hash>>> {
sleep(Duration::from_millis(15));
let tokens = 42;
let keypair = generate_keypair();
let event0 = Event::new_claim(get_pubkey(&keypair), tokens, sign_claim_data(&tokens, &keypair));
hist.sender.send(event0)?;
sleep(Duration::from_millis(10));
Ok(())
}
fn main() {
let seed = Hash::default();
let hist = Historian::new(&seed, Some(10));
create_ledger(&hist).expect("send error");
drop(hist.sender);
let entries: Vec<Entry<Hash>> = hist.receiver.iter().collect();
for entry in &entries {
println!("{:?}", entry);
}
// Proof-of-History: Verify the historian learned about the events
// in the same order they appear in the vector.
assert!(entries[..].verify(&seed));
}
```
Running the program should produce a ledger similar to:
```rust
Entry { num_hashes: 0, id: [0, ...], event: Tick }
Entry { num_hashes: 3, id: [67, ...], event: Transaction { tokens: 42 } }
Entry { num_hashes: 3, id: [123, ...], event: Tick }
```
Proof-of-History
---
Take note of the last line:
```rust
assert!(entries[..].verify(&seed));
```
[It's a proof!](https://en.wikipedia.org/wiki/CurryHoward_correspondence) For each entry returned by the
historian, we can verify that `id` is the result of applying a sha256 hash to the previous `id`
exactly `num_hashes` times, and then hashing then event data on top of that. Because the event data is
included in the hash, the events cannot be reordered without regenerating all the hashes.

@@ -1,18 +0,0 @@
msc {
client,historian,recorder;
recorder=>historian [ label = "e0 = Entry{id: h0, n: 0, event: Tick}" ] ;
recorder=>recorder [ label = "h1 = hash(h0)" ] ;
recorder=>recorder [ label = "h2 = hash(h1)" ] ;
client=>historian [ label = "Transaction(d0)" ] ;
historian=>recorder [ label = "Transaction(d0)" ] ;
recorder=>recorder [ label = "h3 = hash(h2 + d0)" ] ;
recorder=>historian [ label = "e1 = Entry{id: hash(h3), n: 3, event: Transaction(d0)}" ] ;
recorder=>recorder [ label = "h4 = hash(h3)" ] ;
recorder=>recorder [ label = "h5 = hash(h4)" ] ;
recorder=>recorder [ label = "h6 = hash(h5)" ] ;
recorder=>historian [ label = "e2 = Entry{id: h6, n: 3, event: Tick}" ] ;
client=>historian [ label = "collect()" ] ;
historian=>client [ label = "entries = [e0, e1, e2]" ] ;
client=>client [ label = "entries.verify(h0)" ] ;
}

229
doc/json-rpc.md Normal file

@@ -0,0 +1,229 @@
Solana JSON RPC API
===
Solana nodes accept HTTP requests using the [JSON-RPC 2.0](https://www.jsonrpc.org/specification) specification.
To interact with a Solana node inside a JavaScript application, use the [solana-web3.js](https://github.com/solana-labs/solana-web3.js) library, which gives a convenient interface for the RPC methods.
RPC Endpoint
---
**Default port:** 8899
eg. http://localhost:8899, http://192.168.1.88:8899
Methods
---
* [confirmTransaction](#confirmtransaction)
* [getAddress](#getaddress)
* [getBalance](#getbalance)
* [getAccountInfo](#getaccountinfo)
* [getLastId](#getlastid)
* [getSignatureStatus](#getsignaturestatus)
* [getTransactionCount](#gettransactioncount)
* [requestAirdrop](#requestairdrop)
* [sendTransaction](#sendtransaction)
Request Formatting
---
To make a JSON-RPC request, send an HTTP POST request with a `Content-Type: application/json` header. The JSON request data should contain 4 fields:
* `jsonrpc`, set to `"2.0"`
* `id`, a unique client-generated identifying integer
* `method`, a string containing the method to be invoked
* `params`, a JSON array of ordered parameter values
Example using curl:
```bash
curl -X POST -H "Content-Type: application/json" -d '{"jsonrpc":"2.0", "id":1, "method":"getBalance", "params":["83astBRguLMdt2h5U1Tpdq5tjFoJ6noeGwaY3mDLVcri"]}' 192.168.1.88:8899
```
The response output will be a JSON object with the following fields:
* `jsonrpc`, matching the request specification
* `id`, matching the request identifier
* `result`, requested data or success confirmation
Requests can be sent in batches by sending an array of JSON-RPC request objects as the data for a single POST.
Definitions
---
* Hash: A SHA-256 hash of a chunk of data.
* Pubkey: The public key of a Ed25519 key-pair.
* Signature: An Ed25519 signature of a chunk of data.
* Transaction: A Solana instruction signed by a client key-pair.
JSON RPC API Reference
---
### confirmTransaction
Returns a transaction receipt
##### Parameters:
* `string` - Signature of Transaction to confirm, as base-58 encoded string
##### Results:
* `boolean` - Transaction status, true if Transaction is confirmed
##### Example:
```bash
// Request
curl -X POST -H "Content-Type: application/json" -d '{"jsonrpc":"2.0", "id":1, "method":"confirmTransaction", "params":["5VERv8NMvzbJMEkV8xnrLkEaWRtSz9CosKDYjCJjBRnbJLgp8uirBgmQpjKhoR4tjF3ZpRzrFmBV6UjKdiSZkQUW"]}' http://localhost:8899
// Result
{"jsonrpc":"2.0","result":true,"id":1}
```
---
### getBalance
Returns the balance of the account of provided Pubkey
##### Parameters:
* `string` - Pubkey of account to query, as base-58 encoded string
##### Results:
* `integer` - quantity, as a signed 64-bit integer
##### Example:
```bash
// Request
curl -X POST -H "Content-Type: application/json" -d '{"jsonrpc":"2.0", "id":1, "method":"getBalance", "params":["83astBRguLMdt2h5U1Tpdq5tjFoJ6noeGwaY3mDLVcri"]}' http://localhost:8899
// Result
{"jsonrpc":"2.0","result":0,"id":1}
```
---
### getAccountInfo
Returns all information associated with the account of provided Pubkey
##### Parameters:
* `string` - Pubkey of account to query, as base-58 encoded string
##### Results:
The result field will be a JSON object with the following sub fields:
* `tokens`, number of tokens assigned to this account, as a signed 64-bit integer
* `program_id`, array of 32 bytes representing the program this account has been assigned to
* `userdata`, array of bytes representing any userdata associated with the account
##### Example:
```bash
// Request
curl -X POST -H "Content-Type: application/json" -d '{"jsonrpc":"2.0", "id":1, "method":"getAccountInfo", "params":["FVxxngPx368XvMCoeskdd6U8cZJFsfa1BEtGWqyAxRj4"]}' http://localhost:8899
// Result
{"jsonrpc":"2.0","result":{"program_id":[1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],"tokens":1,"userdata":[3,0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,0,0,20,0,0,0,0,0,0,0,50,48,53,48,45,48,49,45,48,49,84,48,48,58,48,48,58,48,48,90,252,10,7,28,246,140,88,177,98,82,10,227,89,81,18,30,194,101,199,16,11,73,133,20,246,62,114,39,20,113,189,32,50,0,0,0,0,0,0,0,247,15,36,102,167,83,225,42,133,127,82,34,36,224,207,130,109,230,224,188,163,33,213,13,5,117,211,251,65,159,197,51,0,0,0,0,0,0]},"id":1}
```
---
### getLastId
Returns the last entry ID from the ledger
##### Parameters:
None
##### Results:
* `string` - the ID of last entry, a Hash as base-58 encoded string
##### Example:
```bash
// Request
curl -X POST -H "Content-Type: application/json" -d '{"jsonrpc":"2.0","id":1, "method":"getLastId"}' http://localhost:8899
// Result
{"jsonrpc":"2.0","result":"GH7ome3EiwEr7tu9JuTh2dpYWBJK3z69Xm1ZE3MEE6JC","id":1}
```
---
### getSignatureStatus
Returns the status of a given signature. This method is similar to
[confirmTransaction](#confirmtransaction) but provides more resolution for error
events.
##### Parameters:
* `string` - Signature of Transaction to confirm, as base-58 encoded string
##### Results:
* `string` - Transaction status:
* `Confirmed` - Transaction was successful
* `SignatureNotFound` - Unknown transaction
* `ProgramRuntimeError` - An error occurred in the program that processed this Transaction
* `GenericFailure` - Some other error occurred. **Note**: In the future new Transaction statuses may be added to this list. It's safe to assume that all new statuses will be more specific error conditions that previously presented as `GenericFailure`
##### Example:
```bash
// Request
curl -X POST -H "Content-Type: application/json" -d '{"jsonrpc":"2.0", "id":1, "method":"getSignatureStatus", "params":["5VERv8NMvzbJMEkV8xnrLkEaWRtSz9CosKDYjCJjBRnbJLgp8uirBgmQpjKhoR4tjF3ZpRzrFmBV6UjKdiSZkQUW"]}' http://localhost:8899
// Result
{"jsonrpc":"2.0","result":"SignatureNotFound","id":1}
```
---
### getTransactionCount
Returns the current Transaction count from the ledger
##### Parameters:
None
##### Results:
* `integer` - count, as unsigned 64-bit integer
##### Example:
```bash
// Request
curl -X POST -H "Content-Type: application/json" -d '{"jsonrpc":"2.0","id":1, "method":"getTransactionCount"}' http://localhost:8899
// Result
{"jsonrpc":"2.0","result":268,"id":1}
```
---
### requestAirdrop
Requests an airdrop of tokens to a Pubkey
##### Parameters:
* `string` - Pubkey of account to receive tokens, as base-58 encoded string
* `integer` - token quantity, as a signed 64-bit integer
##### Results:
* `string` - Transaction Signature of airdrop, as base-58 encoded string
##### Example:
```bash
// Request
curl -X POST -H "Content-Type: application/json" -d '{"jsonrpc":"2.0","id":1, "method":"requestAirdrop", "params":["83astBRguLMdt2h5U1Tpdq5tjFoJ6noeGwaY3mDLVcri", 50]}' http://localhost:8899
// Result
{"jsonrpc":"2.0","result":"5VERv8NMvzbJMEkV8xnrLkEaWRtSz9CosKDYjCJjBRnbJLgp8uirBgmQpjKhoR4tjF3ZpRzrFmBV6UjKdiSZkQUW","id":1}
```
---
### sendTransaction
Creates new transaction
##### Parameters:
* `array` - array of octets containing a fully-signed Transaction
##### Results:
* `string` - Transaction Signature, as base-58 encoded string
##### Example:
```bash
// Request
curl -X POST -H "Content-Type: application/json" -d '{"jsonrpc":"2.0","id":1, "method":"sendTransaction", "params":[[61, 98, 55, 49, 15, 187, 41, 215, 176, 49, 234, 229, 228, 77, 129, 221, 239, 88, 145, 227, 81, 158, 223, 123, 14, 229, 235, 247, 191, 115, 199, 71, 121, 17, 32, 67, 63, 209, 239, 160, 161, 2, 94, 105, 48, 159, 235, 235, 93, 98, 172, 97, 63, 197, 160, 164, 192, 20, 92, 111, 57, 145, 251, 6, 40, 240, 124, 194, 149, 155, 16, 138, 31, 113, 119, 101, 212, 128, 103, 78, 191, 80, 182, 234, 216, 21, 121, 243, 35, 100, 122, 68, 47, 57, 13, 39, 0, 0, 0, 0, 50, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 50, 0, 0, 0, 0, 0, 0, 0, 40, 240, 124, 194, 149, 155, 16, 138, 31, 113, 119, 101, 212, 128, 103, 78, 191, 80, 182, 234, 216, 21, 121, 243, 35, 100, 122, 68, 47, 57, 11, 12, 106, 49, 74, 226, 201, 16, 161, 192, 28, 84, 124, 97, 190, 201, 171, 186, 6, 18, 70, 142, 89, 185, 176, 154, 115, 61, 26, 163, 77, 1, 88, 98, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]}' http://localhost:8899
// Result
{"jsonrpc":"2.0","result":"2EBVM6cB8vAAD93Ktr6Vd8p67XPbQzCJX47MpReuiCXJAtcjaxpvWpcg9Ege1Nr5Tk3a2GFrByT7WPBjdsTycY9b","id":1}
```
---

47
doc/testnet.md Normal file

@@ -0,0 +1,47 @@
# TestNet debugging info
Currently we have three testnets:
* `testnet` - public beta channel testnet accessible via testnet.solana.com. Runs 24/7
* `testnet-perf` - private beta channel testnet with clients trying to flood the network
with transactions until failure. Runs 24/7
* `testnet-master` - private edge channel testnet with clients trying to flood the network
with transactions until failure. Runs on weekday mornings for a couple hours
## Deploy process
They are deployed with the `ci/testnet-deploy.sh` script. There is a scheduled buildkite job which runs to do the deploy,
look at `testnet-deploy` to see the agent which ran it and the logs. There is also a manual job to do the deploy manually..
Validators are selected based on their machine name and everyone gets the binaries installed from snap.
## Where are the testnet logs?
Attach to the testnet first by running one of:
```bash
$ net/gce.sh config testnet-solana-com
$ net/gce.sh config master-testnet-solana-com
$ net/gce.sh config perf-testnet-solana-com
```
Then run:
```bash
$ net/ssh.sh
```
for log location details
## How do I reset the testnet?
Manually trigger the [testnet-deploy](https://buildkite.com/solana-labs/testnet-deploy/) pipeline
## How can I scale the tx generation rate?
Increase the TX rate by increasing the number of cores on the client machine which is running
`bench-tps` or run multiple clients. Decrease by lowering cores or using the rayon env
variable `RAYON_NUM_THREADS=<xx>`
## How can I test a change on the testnet?
Currently, a merged PR is the only way to test a change on the testnet. But you
can run your own testnet using the scripts in the `net/` directory.
## Adjusting the number of clients or validators on the testnet
Through the [testnet-deploy](https://buildkite.com/solana-labs/testnet-deploy/) settings.

39
fetch-perf-libs.sh Executable file

@@ -0,0 +1,39 @@
#!/bin/bash -e
if [[ $(uname) != Linux ]]; then
echo Performance libraries are only available for Linux
exit 1
fi
if [[ $(uname -m) != x86_64 ]]; then
echo Performance libraries are only available for x86_64 architecture
exit 1
fi
mkdir -p target/perf-libs
(
cd target/perf-libs
(
set -x
curl https://solana-perf.s3.amazonaws.com/v0.9.0/x86_64-unknown-linux-gnu/solana-perf.tgz | tar zxvf -
)
if [[ -r /usr/local/cuda/version.txt && -r cuda-version.txt ]]; then
if ! diff /usr/local/cuda/version.txt cuda-version.txt > /dev/null; then
echo ==============================================
echo Warning: possible CUDA version mismatch
echo
echo "Expected version: $(cat cuda-version.txt)"
echo "Detected version: $(cat /usr/local/cuda/version.txt)"
echo ==============================================
fi
else
echo ==============================================
echo Warning: unable to validate CUDA version
echo ==============================================
fi
echo "Downloaded solana-perf version: $(cat solana-perf-HEAD.txt)"
)
exit 0

@@ -1,16 +1,25 @@
#!/bin/bash -e #!/bin/bash -e
if [[ -z "$1" ]]; then here=$(dirname "$0")
echo "usage: $0 [network path to solana repo on leader machine]" # shellcheck source=multinode-demo/common.sh
source "$here"/common.sh
usage() {
if [[ -n $1 ]]; then
echo "$*"
echo
fi
echo "usage: $0 [extra args]"
echo
echo " Run bench-tps "
echo
echo " extra args: additional arguments are pass along to solana-bench-tps"
echo
exit 1 exit 1
}
if [[ -z $1 ]]; then # default behavior
$solana_bench_tps --identity config-private/client-id.json --network 127.0.0.1:8001 --duration 90
else
$solana_bench_tps "$@"
fi fi
LEADER="$1"
set -x
export RUST_LOG=solana=info
rsync -v -e ssh "$LEADER/leader.json" .
rsync -v -e ssh "$LEADER/mint-demo.json" .
cargo run --release --bin solana-client-demo -- \
-l leader.json < mint-demo.json 2>&1 | tee client.log

202
multinode-demo/common.sh Normal file

@@ -0,0 +1,202 @@
# |source| this file
#
# Common utilities shared by other scripts in this directory
#
# The following directive disable complaints about unused variables in this
# file:
# shellcheck disable=2034
#
rsync=rsync
leader_logger="tee leader.log"
validator_logger="tee validator.log"
drone_logger="tee drone.log"
if [[ $(uname) != Linux ]]; then
# Protect against unsupported configurations to prevent non-obvious errors
# later. Arguably these should be fatal errors but for now prefer tolerance.
if [[ -n $USE_SNAP ]]; then
echo "Warning: Snap is not supported on $(uname)"
USE_SNAP=
fi
if [[ -n $SOLANA_CUDA ]]; then
echo "Warning: CUDA is not supported on $(uname)"
SOLANA_CUDA=
fi
fi
if [[ -d $SNAP ]]; then # Running inside a Linux Snap?
solana_program() {
declare program="$1"
printf "%s/command-%s.wrapper" "$SNAP" "$program"
}
rsync="$SNAP"/bin/rsync
multilog="$SNAP/bin/multilog t s16777215 n200"
leader_logger="$multilog $SNAP_DATA/leader"
validator_logger="$multilog t $SNAP_DATA/validator"
drone_logger="$multilog $SNAP_DATA/drone"
# Create log directories manually to prevent multilog from creating them as
# 0700
mkdir -p "$SNAP_DATA"/{drone,leader,validator}
elif [[ -n $USE_SNAP ]]; then # Use the Linux Snap binaries
solana_program() {
declare program="$1"
printf "solana.%s" "$program"
}
elif [[ -n $USE_INSTALL ]]; then # Assume |cargo install| was run
solana_program() {
declare program="$1"
printf "solana-%s" "$program"
}
# CUDA was/wasn't selected at build time, can't affect CUDA state here
unset SOLANA_CUDA
else
solana_program() {
declare program="$1"
declare features=""
if [[ "$program" =~ ^(.*)-cuda$ ]]; then
program=${BASH_REMATCH[1]}
features="--features=cuda"
fi
if [[ -z $DEBUG ]]; then
maybe_release=--release
fi
printf "cargo run $maybe_release --bin solana-%s %s -- " "$program" "$features"
}
if [[ -n $SOLANA_CUDA ]]; then
# shellcheck disable=2154 # 'here' is referenced but not assigned
if [[ -z $here ]]; then
echo "|here| is not defined"
exit 1
fi
# Locate perf libs downloaded by |./fetch-perf-libs.sh|
LD_LIBRARY_PATH=$(cd "$here" && dirname "$PWD"/target/perf-libs):$LD_LIBRARY_PATH
export LD_LIBRARY_PATH
fi
fi
solana_bench_tps=$(solana_program bench-tps)
solana_wallet=$(solana_program wallet)
solana_drone=$(solana_program drone)
solana_fullnode=$(solana_program fullnode)
solana_fullnode_config=$(solana_program fullnode-config)
solana_fullnode_cuda=$(solana_program fullnode-cuda)
solana_genesis=$(solana_program genesis)
solana_keygen=$(solana_program keygen)
solana_ledger_tool=$(solana_program ledger-tool)
export RUST_LOG=${RUST_LOG:-solana=info} # if RUST_LOG is unset, default to info
export RUST_BACKTRACE=1
# shellcheck source=scripts/configure-metrics.sh
source "$(dirname "${BASH_SOURCE[0]}")"/../scripts/configure-metrics.sh
tune_networking() {
# Skip in CI
[[ -z $CI ]] || return 0
# Reference: https://medium.com/@CameronSparr/increase-os-udp-buffers-to-improve-performance-51d167bb1360
if [[ $(uname) = Linux ]]; then
(
set -x +e
# test the existence of the sysctls before trying to set them
# go ahead and return true and don't exit if these calls fail
sysctl net.core.rmem_max 2>/dev/null 1>/dev/null &&
sudo sysctl -w net.core.rmem_max=67108864 1>/dev/null 2>/dev/null
sysctl net.core.rmem_default 2>/dev/null 1>/dev/null &&
sudo sysctl -w net.core.rmem_default=26214400 1>/dev/null 2>/dev/null
sysctl net.core.wmem_max 2>/dev/null 1>/dev/null &&
sudo sysctl -w net.core.wmem_max=67108864 1>/dev/null 2>/dev/null
sysctl net.core.wmem_default 2>/dev/null 1>/dev/null &&
sudo sysctl -w net.core.wmem_default=26214400 1>/dev/null 2>/dev/null
) || true
fi
if [[ $(uname) = Darwin ]]; then
(
if [[ $(sysctl net.inet.udp.maxdgram | cut -d\ -f2) != 65535 ]]; then
echo "Adjusting maxdgram to allow for large UDP packets, see BLOB_SIZE in src/packet.rs:"
set -x
sudo sysctl net.inet.udp.maxdgram=65535
fi
)
fi
}
SOLANA_CONFIG_DIR=${SNAP_DATA:-$PWD}/config
SOLANA_CONFIG_PRIVATE_DIR=${SNAP_DATA:-$PWD}/config-private
SOLANA_CONFIG_VALIDATOR_DIR=${SNAP_DATA:-$PWD}/config-validator
SOLANA_CONFIG_CLIENT_DIR=${SNAP_USER_DATA:-$PWD}/config-client
rsync_url() { # adds the 'rsync://` prefix to URLs that need it
declare url="$1"
if [[ $url =~ ^.*:.*$ ]]; then
# assume remote-shell transport when colon is present, use $url unmodified
echo "$url"
return 0
fi
if [[ -d $url ]]; then
# assume local directory if $url is a valid directory, use $url unmodified
echo "$url"
return 0
fi
# Default to rsync:// URL
echo "rsync://$url"
}
# called from drone, validator, client
find_leader() {
declare leader leader_address
declare shift=0
if [[ -d $SNAP ]]; then
if [[ -n $1 ]]; then
usage "Error: unexpected parameter: $1"
fi
# Select leader from the Snap configuration
leader_ip=$(snapctl get leader-ip)
if [[ -z $leader_ip ]]; then
leader=testnet.solana.com
leader_ip=$(dig +short "${leader%:*}" | head -n1)
if [[ -z $leader_ip ]]; then
usage "Error: unable to resolve IP address for $leader"
fi
fi
leader=$leader_ip
leader_address=$leader_ip:8001
else
if [[ -z $1 ]]; then
leader=${here}/.. # Default to local tree for rsync
leader_address=127.0.0.1:8001 # Default to local leader
elif [[ -z $2 ]]; then
leader=$1
declare leader_ip
leader_ip=$(dig +short "${leader%:*}" | head -n1)
if [[ -z $leader_ip ]]; then
usage "Error: unable to resolve IP address for $leader"
fi
leader_address=$leader_ip:8001
shift=1
else
leader=$1
leader_address=$2
shift=2
fi
fi
echo "$leader" "$leader_address" "$shift"
}

40
multinode-demo/drone.sh Executable file

@@ -0,0 +1,40 @@
#!/bin/bash
#
# Starts an instance of solana-drone
#
here=$(dirname "$0")
# shellcheck source=multinode-demo/common.sh
source "$here"/common.sh
usage() {
if [[ -n $1 ]]; then
echo "$*"
echo
fi
echo "usage: $0 [network entry point]"
echo
echo " Run an airdrop drone for the specified network"
echo
exit 1
}
read -r _ leader_address shift < <(find_leader "${@:1:1}")
shift "$shift"
[[ -f "$SOLANA_CONFIG_PRIVATE_DIR"/mint.json ]] || {
echo "$SOLANA_CONFIG_PRIVATE_DIR/mint.json not found, create it by running:"
echo
echo " ${here}/setup.sh -t leader"
exit 1
}
set -ex
trap 'kill "$pid" && wait "$pid"' INT TERM
$solana_drone \
--keypair "$SOLANA_CONFIG_PRIVATE_DIR"/mint.json \
--network "$leader_address" \
> >($drone_logger) 2>&1 &
pid=$!
wait "$pid"

@@ -1,4 +1,41 @@
#!/bin/bash #!/bin/bash
export RUST_LOG=solana=info #
sudo sysctl -w net.core.rmem_max=26214400 # Starts a leader node
cargo run --release --bin solana-fullnode -- -l leader.json < genesis.log #
here=$(dirname "$0")
# shellcheck source=multinode-demo/common.sh
source "$here"/common.sh
# shellcheck source=scripts/oom-score-adj.sh
source "$here"/../scripts/oom-score-adj.sh
if [[ -d "$SNAP" ]]; then
# Exit if mode is not yet configured
# (typically the case after the Snap is first installed)
[[ -n "$(snapctl get mode)" ]] || exit 0
fi
[[ -f "$SOLANA_CONFIG_DIR"/leader.json ]] || {
echo "$SOLANA_CONFIG_DIR/leader.json not found, create it by running:"
echo
echo " ${here}/setup.sh"
exit 1
}
if [[ -n "$SOLANA_CUDA" ]]; then
program="$solana_fullnode_cuda"
else
program="$solana_fullnode"
fi
tune_networking
trap 'kill "$pid" && wait "$pid"' INT TERM
$program \
--identity "$SOLANA_CONFIG_DIR"/leader.json \
--ledger "$SOLANA_CONFIG_DIR"/ledger \
> >($leader_logger) 2>&1 &
pid=$!
oom_score_adj "$pid" 1000
wait "$pid"

128
multinode-demo/setup.sh Executable file

@@ -0,0 +1,128 @@
#!/bin/bash
#
# Creates a fullnode configuration
#
here=$(dirname "$0")
# shellcheck source=multinode-demo/common.sh
source "$here"/common.sh
usage () {
exitcode=0
if [[ -n "$1" ]]; then
exitcode=1
echo "Error: $*"
fi
cat <<EOF
usage: $0 [-n num_tokens] [-l] [-p] [-t node_type]
Creates a fullnode configuration
-n num_tokens - Number of tokens to create
-l - Detect network address from local machine configuration, which
may be a private IP address unaccessible on the Intenet (default)
-p - Detect public address using public Internet servers
-t node_type - Create configuration files only for this kind of node. Valid
options are validator or leader. Creates configuration files
for both by default
EOF
exit $exitcode
}
ip_address_arg=-l
num_tokens=1000000000
node_type_leader=true
node_type_validator=true
node_type_client=true
while getopts "h?n:lpt:" opt; do
case $opt in
h|\?)
usage
exit 0
;;
l)
ip_address_arg=-l
;;
p)
ip_address_arg=-p
;;
n)
num_tokens="$OPTARG"
;;
t)
node_type="$OPTARG"
case $OPTARG in
leader)
node_type_leader=true
node_type_validator=false
node_type_client=false
;;
validator)
node_type_leader=false
node_type_validator=true
node_type_client=false
;;
client)
node_type_leader=false
node_type_validator=false
node_type_client=true
;;
*)
usage "Error: unknown node type: $node_type"
;;
esac
;;
*)
usage "Error: unhandled option: $opt"
;;
esac
done
set -e
for i in "$SOLANA_CONFIG_DIR" "$SOLANA_CONFIG_VALIDATOR_DIR" "$SOLANA_CONFIG_PRIVATE_DIR"; do
echo "Cleaning $i"
rm -rvf "$i"
mkdir -p "$i"
done
if $node_type_client; then
client_id_path="$SOLANA_CONFIG_PRIVATE_DIR"/client-id.json
$solana_keygen -o "$client_id_path"
ls -lhR "$SOLANA_CONFIG_PRIVATE_DIR"/
fi
if $node_type_leader; then
leader_address_args=("$ip_address_arg")
leader_id_path="$SOLANA_CONFIG_PRIVATE_DIR"/leader-id.json
mint_path="$SOLANA_CONFIG_PRIVATE_DIR"/mint.json
$solana_keygen -o "$leader_id_path"
echo "Creating $mint_path with $num_tokens tokens"
$solana_keygen -o "$mint_path"
echo "Creating $SOLANA_CONFIG_DIR/ledger"
$solana_genesis --tokens="$num_tokens" --ledger "$SOLANA_CONFIG_DIR"/ledger < "$mint_path"
echo "Creating $SOLANA_CONFIG_DIR/leader.json"
$solana_fullnode_config --keypair="$leader_id_path" "${leader_address_args[@]}" > "$SOLANA_CONFIG_DIR"/leader.json
ls -lhR "$SOLANA_CONFIG_DIR"/
ls -lhR "$SOLANA_CONFIG_PRIVATE_DIR"/
fi
if $node_type_validator; then
validator_address_args=("$ip_address_arg" -b 9000)
validator_id_path="$SOLANA_CONFIG_PRIVATE_DIR"/validator-id.json
$solana_keygen -o "$validator_id_path"
echo "Creating $SOLANA_CONFIG_VALIDATOR_DIR/validator.json"
$solana_fullnode_config --keypair="$validator_id_path" "${validator_address_args[@]}" > "$SOLANA_CONFIG_VALIDATOR_DIR"/validator.json
ls -lhR "$SOLANA_CONFIG_VALIDATOR_DIR"/
fi

8
multinode-demo/validator-x.sh Executable file

@@ -0,0 +1,8 @@
#!/bin/bash
#
# Start a dynamically-configured validator node
#
here=$(dirname "$0")
exec "$here"/validator.sh -x "$@"

@@ -1,21 +1,100 @@
#!/bin/bash -e #!/bin/bash
#
# Start a validator node
#
here=$(dirname "$0")
# shellcheck source=multinode-demo/common.sh
source "$here"/common.sh
if [[ -z "$1" ]]; then # shellcheck source=scripts/oom-score-adj.sh
echo "usage: $0 [network path to solana repo on leader machine]" source "$here"/../scripts/oom-score-adj.sh
exit 1
if [[ -d "$SNAP" ]]; then
# Exit if mode is not yet configured
# (typically the case after the Snap is first installed)
[[ -n "$(snapctl get mode)" ]] || exit 0
fi fi
LEADER="$1" usage() {
if [[ -n $1 ]]; then
echo "$*"
echo
fi
echo "usage: $0 [-x] [rsync network path to leader] [network entry point]"
echo
echo " Start a validator on the specified network"
echo
echo " -x: runs a new, dynamically-configured validator"
echo
exit 1
}
set -x if [[ $1 = -h ]]; then
usage
fi
rsync -v -e ssh "$LEADER/mint-demo.json" . if [[ $1 == -x ]]; then
rsync -v -e ssh "$LEADER/leader.json" . self_setup=1
rsync -v -e ssh "$LEADER/genesis.log" . shift
else
self_setup=0
fi
export RUST_LOG=solana=info if [[ -n $3 ]]; then
usage
fi
sudo sysctl -w net.core.rmem_max=26214400 read -r leader leader_address shift < <(find_leader "${@:1:2}")
shift "$shift"
cargo run --release --features=cuda --bin solana-fullnode -- \ if [[ -n $SOLANA_CUDA ]]; then
-l validator.json -v leader.json -b 9000 -d < genesis.log program=$solana_fullnode_cuda
else
program=$solana_fullnode
fi
if ((!self_setup)); then
[[ -f $SOLANA_CONFIG_VALIDATOR_DIR/validator.json ]] || {
echo "$SOLANA_CONFIG_VALIDATOR_DIR/validator.json not found, create it by running:"
echo
echo " ${here}/setup.sh"
exit 1
}
validator_json_path=$SOLANA_CONFIG_VALIDATOR_DIR/validator.json
SOLANA_LEADER_CONFIG_DIR=$SOLANA_CONFIG_VALIDATOR_DIR/leader-config
else
mkdir -p "$SOLANA_CONFIG_PRIVATE_DIR"
validator_id_path=$SOLANA_CONFIG_PRIVATE_DIR/validator-id-x$$.json
$solana_keygen -o "$validator_id_path"
mkdir -p "$SOLANA_CONFIG_VALIDATOR_DIR"
validator_json_path=$SOLANA_CONFIG_VALIDATOR_DIR/validator-x$$.json
port=9000
(((port += ($$ % 1000)) && (port == 9000) && port++))
$solana_fullnode_config --keypair="$validator_id_path" -l -b "$port" > "$validator_json_path"
SOLANA_LEADER_CONFIG_DIR=$SOLANA_CONFIG_VALIDATOR_DIR/leader-config-x$$
fi
rsync_leader_url=$(rsync_url "$leader")
tune_networking
set -ex
$rsync -vPr "$rsync_leader_url"/config/ "$SOLANA_LEADER_CONFIG_DIR"
[[ -d $SOLANA_LEADER_CONFIG_DIR/ledger ]] || {
echo "Unable to retrieve ledger from $rsync_leader_url"
exit 1
}
trap 'kill "$pid" && wait "$pid"' INT TERM
$program \
--identity "$validator_json_path" \
--network "$leader_address" \
--ledger "$SOLANA_LEADER_CONFIG_DIR"/ledger \
> >($validator_logger) 2>&1 &
pid=$!
oom_score_adj "$pid" 1000
wait "$pid"

2
net/.gitignore vendored Normal file

@@ -0,0 +1,2 @@
/config/
/log/

100
net/README.md Normal file

@@ -0,0 +1,100 @@
# Network Management
This directory contains scripts useful for working with a test network. It's
intended to be both dev and CD friendly.
### User Account Prerequisites
GCP and AWS are supported.
#### GCP
First authenticate with
```bash
$ gcloud auth login
```
#### AWS
Obtain your credentials from the AWS IAM Console and configure the AWS CLI with
```bash
$ aws configure
```
More information on AWS CLI configuration can be found [here](https://docs.aws.amazon.com/cli/latest/userguide/cli-chap-getting-started.html#cli-quick-configuration)
### Metrics configuration
Ensure that `$(whoami)` is the name of an InfluxDB user account with enough
access to create a new InfluxDB database. Ask mvines@ for help if needed.
## Quick Start
NOTE: This example uses GCP. If you are using AWS, replace `./gce.sh` with
`./ec2.sh` in the commands.
```bash
$ cd net/
$ ./gce.sh create -n 5 -c 1 #<-- Create a GCE testnet with 5 validators, 1 client (billing starts here)
$ ./init-metrics.sh $(whoami) #<-- Configure a metrics database for the testnet
$ ./net.sh start #<-- Deploy the network from the local workspace
$ ./ssh.sh #<-- Details on how to ssh into any testnet node
$ ./gce.sh delete #<-- Dispose of the network (billing stops here)
```
## Tips
### Running the network over public IP addresses
By default private IP addresses are used with all instances in the same
availability zone to avoid GCE network engress charges. However to run the
network over public IP addresses:
```bash
$ ./gce.sh create -P ...
```
or
```bash
$ ./ec2.sh create -P ...
```
### Deploying a Snap-based network
To deploy the latest pre-built `edge` channel Snap (ie, latest from the `master`
branch), once the testnet has been created run:
```bash
$ ./net.sh start -s edge
```
### Enabling CUDA
First ensure the network instances are created with GPU enabled:
```bash
$ ./gce.sh create -g ...
```
or
```bash
$ ./ec2.sh create -g ...
```
If deploying a Snap-based network nothing further is required, as GPU presence
is detected at runtime and the CUDA build is auto selected.
If deploying a locally-built network, first run `./fetch-perf-libs.sh` then
ensure the `cuda` feature is specified at network start:
```bash
$ ./net.sh start -f "cuda,erasure"
```
### How to interact with a CD testnet deployed by ci/testnet-deploy.sh
**AWS-Specific Extra Setup**: Follow the steps in `scripts/add-solana-user-authorized_keys.sh`,
then redeploy the testnet before continuing in this section.
Taking **master-testnet-solana-com** as an example, configure your workspace for
the testnet using:
```bash
$ ./gce.sh config -p master-testnet-solana-com
```
or
```bash
$ ./ec2.sh config -p master-testnet-solana-com
```
Then run the following for details on how to ssh into any testnet node
```bash
$ ./ssh.sh
```

58
net/common.sh Normal file

@@ -0,0 +1,58 @@
# |source| this file
#
# Common utilities shared by other scripts in this directory
#
# The following directive disable complaints about unused variables in this
# file:
# shellcheck disable=2034
#
netDir=$(
cd "$(dirname "${BASH_SOURCE[0]}")" || exit
echo "$PWD"
)
netConfigDir="$netDir"/config
netLogDir="$netDir"/log
mkdir -p "$netConfigDir" "$netLogDir"
# shellcheck source=scripts/configure-metrics.sh
source "$(dirname "${BASH_SOURCE[0]}")"/../scripts/configure-metrics.sh
configFile="$netConfigDir/config"
entrypointIp=
publicNetwork=
leaderIp=
netBasename=
sshPrivateKey=
clientIpList=()
sshOptions=()
validatorIpList=()
buildSshOptions() {
sshOptions=(
-o "BatchMode=yes"
-o "StrictHostKeyChecking=no"
-o "UserKnownHostsFile=/dev/null"
-o "User=solana"
-o "IdentityFile=$sshPrivateKey"
-o "LogLevel=ERROR"
-F /dev/null
)
}
loadConfigFile() {
[[ -r $configFile ]] || usage "Config file unreadable: $configFile"
# shellcheck source=/dev/null
source "$configFile"
[[ -n "$entrypointIp" ]] || usage "Config file invalid, entrypointIp unspecified: $configFile"
[[ -n "$publicNetwork" ]] || usage "Config file invalid, publicNetwork unspecified: $configFile"
[[ -n "$leaderIp" ]] || usage "Config file invalid, leaderIp unspecified: $configFile"
[[ -n "$netBasename" ]] || usage "Config file invalid, netBasename unspecified: $configFile"
[[ -n $sshPrivateKey ]] || usage "Config file invalid, sshPrivateKey unspecified: $configFile"
[[ ${#validatorIpList[@]} -gt 0 ]] || usage "Config file invalid, validatorIpList unspecified: $configFile"
buildSshOptions
configureMetrics
}

1
net/ec2.sh Symbolic link

@@ -0,0 +1 @@
gce.sh

413
net/gce.sh Executable file

@@ -0,0 +1,413 @@
#!/bin/bash -e
here=$(dirname "$0")
# shellcheck source=net/common.sh
source "$here"/common.sh
cloudProvider=$(basename "$0" .sh)
bootDiskType=""
case $cloudProvider in
gce)
# shellcheck source=net/scripts/gce-provider.sh
source "$here"/scripts/gce-provider.sh
imageName="ubuntu-16-04-cuda-9-2-new"
cpuLeaderMachineType=n1-standard-16
gpuLeaderMachineType="$cpuLeaderMachineType --accelerator count=4,type=nvidia-tesla-k80"
leaderMachineType=$cpuLeaderMachineType
validatorMachineType=n1-standard-4
clientMachineType=n1-standard-16
;;
ec2)
# shellcheck source=net/scripts/ec2-provider.sh
source "$here"/scripts/ec2-provider.sh
imageName="ami-0466e26ccc0e752c1"
cpuLeaderMachineType=m4.4xlarge
gpuLeaderMachineType=p2.xlarge
leaderMachineType=$cpuLeaderMachineType
validatorMachineType=m4.xlarge
clientMachineType=m4.4xlarge
;;
*)
echo "Error: Unknown cloud provider: $cloudProvider"
;;
esac
prefix=testnet-dev-${USER//[^A-Za-z0-9]/}
validatorNodeCount=5
clientNodeCount=1
leaderBootDiskSizeInGb=1000
validatorBootDiskSizeInGb=$leaderBootDiskSizeInGb
clientBootDiskSizeInGb=75
publicNetwork=false
enableGpu=false
leaderAddress=
usage() {
exitcode=0
if [[ -n "$1" ]]; then
exitcode=1
echo "Error: $*"
fi
cat <<EOF
usage: $0 [create|config|delete] [common options] [command-specific options]
Manage testnet instances
create - create a new testnet (implies 'config')
config - configure the testnet and write a config file describing it
delete - delete the testnet
common options:
-p [prefix] - Optional common prefix for instance names to avoid
collisions (default: $prefix)
create-specific options:
-n [number] - Number of validator nodes (default: $validatorNodeCount)
-c [number] - Number of client nodes (default: $clientNodeCount)
-P - Use public network IP addresses (default: $publicNetwork)
-z [zone] - Zone for the nodes (default: $zone)
-g - Enable GPU (default: $enableGpu)
-G - Enable GPU, and set count/type of GPUs to use (e.g $cpuLeaderMachineType --accelerator count=4,type=nvidia-tesla-k80)
-a [address] - Set the leader node's external IP address to this value.
For GCE, [address] is the "name" of the desired External
IP Address.
For EC2, [address] is the "allocation ID" of the desired
Elastic IP.
-d [disk-type] - Specify a boot disk type (default None) Use pd-ssd to get ssd on GCE.
config-specific options:
none
delete-specific options:
none
EOF
exit $exitcode
}
command=$1
[[ -n $command ]] || usage
shift
[[ $command = create || $command = config || $command = delete ]] || usage "Invalid command: $command"
while getopts "h?p:Pn:c:z:gG:a:d:" opt; do
case $opt in
h | \?)
usage
;;
p)
[[ ${OPTARG//[^A-Za-z0-9-]/} == "$OPTARG" ]] || usage "Invalid prefix: \"$OPTARG\", alphanumeric only"
prefix=$OPTARG
;;
P)
publicNetwork=true
;;
n)
validatorNodeCount=$OPTARG
;;
c)
clientNodeCount=$OPTARG
;;
z)
cloud_SetZone "$OPTARG"
;;
g)
enableGpu=true
leaderMachineType="$gpuLeaderMachineType"
;;
G)
enableGpu=true
leaderMachineType="$OPTARG"
;;
a)
leaderAddress=$OPTARG
;;
d)
bootDiskType=$OPTARG
;;
*)
usage "Error: unhandled option: $opt"
;;
esac
done
shift $((OPTIND - 1))
[[ -z $1 ]] || usage "Unexpected argument: $1"
sshPrivateKey="$netConfigDir/id_$prefix"
# cloud_ForEachInstance [cmd] [extra args to cmd]
#
# Execute a command for each element in the `instances` array
#
# cmd - The command to execute on each instance
# The command will receive arguments followed by any
# additionl arguments supplied to cloud_ForEachInstance:
# name - name of the instance
# publicIp - The public IP address of this instance
# privateIp - The priate IP address of this instance
# count - Monotonically increasing count for each
# invocation of cmd, starting at 1
# ... - Extra args to cmd..
#
#
cloud_ForEachInstance() {
declare cmd="$1"
shift
[[ -n $cmd ]] || { echo cloud_ForEachInstance: cmd not specified; exit 1; }
declare count=1
for info in "${instances[@]}"; do
declare name publicIp privateIp
IFS=: read -r name publicIp privateIp < <(echo "$info")
eval "$cmd" "$name" "$publicIp" "$privateIp" "$count" "$@"
count=$((count + 1))
done
}
prepareInstancesAndWriteConfigFile() {
$metricsWriteDatapoint "testnet-deploy net-config-begin=1"
cat >> "$configFile" <<EOF
# autogenerated at $(date)
netBasename=$prefix
publicNetwork=$publicNetwork
sshPrivateKey=$sshPrivateKey
EOF
buildSshOptions
recordInstanceIp() {
declare name="$1"
declare publicIp="$2"
declare privateIp="$3"
declare arrayName="$5"
echo "$arrayName+=($publicIp) # $name" >> "$configFile"
if [[ $arrayName = "leaderIp" ]]; then
if $publicNetwork; then
echo "entrypointIp=$publicIp" >> "$configFile"
else
echo "entrypointIp=$privateIp" >> "$configFile"
fi
fi
}
waitForStartupComplete() {
declare name="$1"
declare publicIp="$2"
echo "Waiting for $name to finish booting..."
(
for i in $(seq 1 30); do
if (set -x; ssh "${sshOptions[@]}" "$publicIp" "test -f /.instance-startup-complete"); then
break
fi
sleep 2
echo "Retry $i..."
done
)
echo "$name has booted."
}
echo "Looking for leader instance..."
cloud_FindInstance "$prefix-leader"
[[ ${#instances[@]} -eq 1 ]] || {
echo "Unable to find leader"
exit 1
}
(
declare leaderName
declare leaderIp
IFS=: read -r leaderName leaderIp _ < <(echo "${instances[0]}")
# Try to ping the machine first.
timeout 60s bash -c "set -o pipefail; until ping -c 3 $leaderIp | tr - _; do echo .; done"
if [[ ! -r $sshPrivateKey ]]; then
echo "Fetching $sshPrivateKey from $leaderName"
# Try to scp in a couple times, sshd may not yet be up even though the
# machine can be pinged...
set -x -o pipefail
for i in $(seq 1 30); do
if cloud_FetchFile "$leaderName" "$leaderIp" /solana-id_ecdsa "$sshPrivateKey"; then
break
fi
sleep 1
echo "Retry $i..."
done
chmod 400 "$sshPrivateKey"
ls -l "$sshPrivateKey"
fi
)
echo "leaderIp=()" >> "$configFile"
cloud_ForEachInstance recordInstanceIp leaderIp
cloud_ForEachInstance waitForStartupComplete
echo "Looking for validator instances..."
cloud_FindInstances "$prefix-validator"
[[ ${#instances[@]} -gt 0 ]] || {
echo "Unable to find validators"
exit 1
}
echo "validatorIpList=()" >> "$configFile"
cloud_ForEachInstance recordInstanceIp validatorIpList
cloud_ForEachInstance waitForStartupComplete
echo "clientIpList=()" >> "$configFile"
echo "Looking for client instances..."
cloud_FindInstances "$prefix-client"
[[ ${#instances[@]} -eq 0 ]] || {
cloud_ForEachInstance recordInstanceIp clientIpList
cloud_ForEachInstance waitForStartupComplete
}
echo "Wrote $configFile"
$metricsWriteDatapoint "testnet-deploy net-config-complete=1"
}
delete() {
$metricsWriteDatapoint "testnet-deploy net-delete-begin=1"
# Delete the leader node first to prevent unusual metrics on the dashboard
# during shutdown.
# TODO: It would be better to fully cut-off metrics reporting before any
# instances are deleted.
for filter in "$prefix-leader" "$prefix-"; do
echo "Searching for instances: $filter"
cloud_FindInstances "$filter"
if [[ ${#instances[@]} -eq 0 ]]; then
echo "No instances found matching '$filter'"
else
cloud_DeleteInstances true
fi
done
rm -f "$configFile"
$metricsWriteDatapoint "testnet-deploy net-delete-complete=1"
}
case $command in
delete)
delete
;;
create)
[[ -n $validatorNodeCount ]] || usage "Need number of nodes"
if [[ $validatorNodeCount -le 0 ]]; then
usage "One or more validator nodes is required"
fi
delete
$metricsWriteDatapoint "testnet-deploy net-create-begin=1"
rm -rf "$sshPrivateKey"{,.pub}
# Note: using rsa because |aws ec2 import-key-pair| seems to fail for ecdsa
ssh-keygen -t rsa -N '' -f "$sshPrivateKey"
printNetworkInfo() {
cat <<EOF
========================================================================================
Network composition:
Leader = $leaderMachineType (GPU=$enableGpu)
Validators = $validatorNodeCount x $validatorMachineType
Client(s) = $clientNodeCount x $clientMachineType
========================================================================================
EOF
}
printNetworkInfo
declare startupScript="$netConfigDir"/instance-startup-script.sh
cat > "$startupScript" <<EOF
#!/bin/bash -ex
# autogenerated at $(date)
cat > /etc/motd <<EOM
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
This instance has not been fully configured.
See startup script log messages in /var/log/syslog for status:
$ sudo cat /var/log/syslog | egrep \\(startup-script\\|cloud-init\)
To block until setup is complete, run:
$ until [[ -f /.instance-startup-complete ]]; do sleep 1; done
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
EOM
# Place the generated private key at /solana-id_ecdsa so it's retrievable by anybody
# who is able to log into this machine
cat > /solana-id_ecdsa <<EOK
$(cat "$sshPrivateKey")
EOK
cat > /solana-id_ecdsa.pub <<EOK
$(cat "$sshPrivateKey.pub")
EOK
chmod 444 /solana-id_ecdsa
USER=\$(id -un)
$(
cd "$here"/scripts/
cat \
disable-background-upgrades.sh \
create-solana-user.sh \
add-solana-user-authorized_keys.sh \
install-earlyoom.sh \
install-libssl-compatability.sh \
install-rsync.sh \
)
cat > /etc/motd <<EOM
$(printNetworkInfo)
EOM
touch /.instance-startup-complete
EOF
cloud_CreateInstances "$prefix" "$prefix-leader" 1 \
"$imageName" "$leaderMachineType" "$leaderBootDiskSizeInGb" \
"$startupScript" "$leaderAddress" "$bootDiskType"
cloud_CreateInstances "$prefix" "$prefix-validator" "$validatorNodeCount" \
"$imageName" "$validatorMachineType" "$validatorBootDiskSizeInGb" \
"$startupScript" "" "$bootDiskType"
if [[ $clientNodeCount -gt 0 ]]; then
cloud_CreateInstances "$prefix" "$prefix-client" "$clientNodeCount" \
"$imageName" "$clientMachineType" "$clientBootDiskSizeInGb" \
"$startupScript" "" "$bootDiskType"
fi
$metricsWriteDatapoint "testnet-deploy net-create-complete=1"
prepareInstancesAndWriteConfigFile
;;
config)
prepareInstancesAndWriteConfigFile
;;
*)
usage "Unknown command: $command"
esac

80
net/init-metrics.sh Executable file

@@ -0,0 +1,80 @@
#!/bin/bash -e
here=$(dirname "$0")
# shellcheck source=net/common.sh
source "$here"/common.sh
usage() {
exitcode=0
if [[ -n "$1" ]]; then
exitcode=1
echo "Error: $*"
fi
cat <<EOF
usage: $0 [-e] [-d] [username]
Creates a testnet dev metrics database
username InfluxDB user with access to create a new database
-d Delete the database instead of creating it
-e Assume database already exists and SOLANA_METRICS_CONFIG is
defined in the environment already
EOF
exit $exitcode
}
loadConfigFile
useEnv=false
delete=false
while getopts "hde" opt; do
case $opt in
h|\?)
usage
exit 0
;;
d)
delete=true
;;
e)
useEnv=true
;;
*)
usage "Error: unhandled option: $opt"
;;
esac
done
shift $((OPTIND - 1))
if $useEnv; then
[[ -n $SOLANA_METRICS_CONFIG ]] ||
usage "Error: SOLANA_METRICS_CONFIG is not defined in the environment"
else
username=$1
[[ -n "$username" ]] || usage "username not specified"
read -rs -p "InfluxDB password for $username: " password
[[ -n $password ]] || { echo "Password not specified"; exit 1; }
echo
query() {
echo "$*"
curl -XPOST \
"https://metrics.solana.com:8086/query?u=${username}&p=${password}" \
--data-urlencode "q=$*"
}
query "DROP DATABASE \"$netBasename\""
! $delete || exit 0
query "CREATE DATABASE \"$netBasename\""
query "ALTER RETENTION POLICY autogen ON \"$netBasename\" DURATION 7d"
query "GRANT READ ON \"$netBasename\" TO \"ro\""
query "GRANT WRITE ON \"$netBasename\" TO \"scratch_writer\""
SOLANA_METRICS_CONFIG="db=$netBasename,u=scratch_writer,p=topsecret"
fi
echo "export SOLANA_METRICS_CONFIG=\"$SOLANA_METRICS_CONFIG\"" >> "$configFile"
exit 0

359
net/net.sh Executable file

@@ -0,0 +1,359 @@
#!/bin/bash -e
here=$(dirname "$0")
SOLANA_ROOT="$(cd "$here"/..; pwd)"
# shellcheck source=net/common.sh
source "$here"/common.sh
usage() {
exitcode=0
if [[ -n "$1" ]]; then
exitcode=1
echo "Error: $*"
fi
cat <<EOF
usage: $0 [start|stop|restart|sanity] [command-specific options]
Operate a configured testnet
start - Start the network
sanity - Sanity check the network
stop - Stop the network
restart - Shortcut for stop then start
start-specific options:
-S [snapFilename] - Deploy the specified Snap file
-s edge|beta|stable - Deploy the latest Snap on the specified Snap release channel
-f [cargoFeatures] - List of |cargo --feaures=| to activate
(ignored if -s or -S is specified)
Note: if RUST_LOG is set in the environment it will be propogated into the
network nodes.
sanity/start-specific options:
-o noLedgerVerify - Skip ledger verification
-o noValidatorSanity - Skip validator sanity
-o rejectExtraNodes - Require the exact number of nodes
stop-specific options:
none
EOF
exit $exitcode
}
snapChannel=
snapFilename=
deployMethod=local
sanityExtraArgs=
cargoFeatures=
command=$1
[[ -n $command ]] || usage
shift
while getopts "h?S:s:o:f:" opt; do
case $opt in
h | \?)
usage
;;
S)
snapFilename=$OPTARG
[[ -f $snapFilename ]] || usage "Snap not readable: $snapFilename"
deployMethod=snap
;;
s)
case $OPTARG in
edge|beta|stable)
snapChannel=$OPTARG
deployMethod=snap
;;
*)
usage "Invalid snap channel: $OPTARG"
;;
esac
;;
f)
cargoFeatures=$OPTARG
;;
o)
case $OPTARG in
noLedgerVerify|noValidatorSanity|rejectExtraNodes)
sanityExtraArgs="$sanityExtraArgs -o $OPTARG"
;;
*)
echo "Error: unknown option: $OPTARG"
exit 1
;;
esac
;;
*)
usage "Error: unhandled option: $opt"
;;
esac
done
loadConfigFile
expectedNodeCount=$((${#validatorIpList[@]} + 1))
build() {
declare MAYBE_DOCKER=
if [[ $(uname) != Linux ]]; then
MAYBE_DOCKER="ci/docker-run.sh solanalabs/rust"
fi
SECONDS=0
(
cd "$SOLANA_ROOT"
echo "--- Build started at $(date)"
set -x
rm -rf farf
$MAYBE_DOCKER cargo install --features="$cargoFeatures" --root farf
)
echo "Build took $SECONDS seconds"
}
startCommon() {
declare ipAddress=$1
test -d "$SOLANA_ROOT"
ssh "${sshOptions[@]}" "$ipAddress" "mkdir -p ~/solana ~/.cargo/bin"
rsync -vPrc -e "ssh ${sshOptions[*]}" \
"$SOLANA_ROOT"/{fetch-perf-libs.sh,scripts,net,multinode-demo} \
"$ipAddress":~/solana/
}
startLeader() {
declare ipAddress=$1
declare logFile="$2"
echo "--- Starting leader: $leaderIp"
echo "start log: $logFile"
# Deploy local binaries to leader. Validators and clients later fetch the
# binaries from the leader.
(
set -x
startCommon "$ipAddress" || exit 1
case $deployMethod in
snap)
rsync -vPrc -e "ssh ${sshOptions[*]}" "$snapFilename" "$ipAddress:~/solana/solana.snap"
;;
local)
rsync -vPrc -e "ssh ${sshOptions[*]}" "$SOLANA_ROOT"/farf/bin/* "$ipAddress:~/.cargo/bin/"
;;
*)
usage "Internal error: invalid deployMethod: $deployMethod"
;;
esac
ssh "${sshOptions[@]}" -n "$ipAddress" \
"./solana/net/remote/remote-node.sh $deployMethod leader $publicNetwork $entrypointIp $expectedNodeCount \"$RUST_LOG\""
) >> "$logFile" 2>&1 || {
cat "$logFile"
echo "^^^ +++"
exit 1
}
}
startValidator() {
declare ipAddress=$1
declare logFile="$netLogDir/validator-$ipAddress.log"
echo "--- Starting validator: $leaderIp"
echo "start log: $logFile"
(
set -x
startCommon "$ipAddress"
ssh "${sshOptions[@]}" -n "$ipAddress" \
"./solana/net/remote/remote-node.sh $deployMethod validator $publicNetwork $entrypointIp $expectedNodeCount \"$RUST_LOG\""
) >> "$logFile" 2>&1 &
declare pid=$!
ln -sfT "validator-$ipAddress.log" "$netLogDir/validator-$pid.log"
pids+=("$pid")
}
startClient() {
declare ipAddress=$1
declare logFile="$2"
echo "--- Starting client: $ipAddress"
echo "start log: $logFile"
(
set -x
startCommon "$ipAddress"
ssh "${sshOptions[@]}" -f "$ipAddress" \
"./solana/net/remote/remote-client.sh $deployMethod $entrypointIp $expectedNodeCount \"$RUST_LOG\""
) >> "$logFile" 2>&1 || {
cat "$logFile"
echo "^^^ +++"
exit 1
}
}
sanity() {
declare expectedNodeCount=$((${#validatorIpList[@]} + 1))
declare ok=true
echo "--- Sanity"
$metricsWriteDatapoint "testnet-deploy net-sanity-begin=1"
(
set -x
# shellcheck disable=SC2029 # remote-client.sh args are expanded on client side intentionally
ssh "${sshOptions[@]}" "$leaderIp" \
"./solana/net/remote/remote-sanity.sh $sanityExtraArgs"
) || ok=false
$metricsWriteDatapoint "testnet-deploy net-sanity-complete=1"
$ok || exit 1
}
start() {
case $deployMethod in
snap)
if [[ -n $snapChannel ]]; then
rm -f "$SOLANA_ROOT"/solana_*.snap
if [[ $(uname) != Linux ]]; then
(
set -x
SOLANA_DOCKER_RUN_NOSETUID=1 "$SOLANA_ROOT"/ci/docker-run.sh ubuntu:18.04 bash -c "
set -ex;
apt-get -qq update;
apt-get -qq -y install snapd;
snap download --channel=$snapChannel solana;
"
)
else
(
cd "$SOLANA_ROOT"
snap download --channel="$snapChannel" solana
)
fi
snapFilename="$(echo "$SOLANA_ROOT"/solana_*.snap)"
[[ -r $snapFilename ]] || {
echo "Error: Snap not readable: $snapFilename"
exit 1
}
fi
;;
local)
build
;;
*)
usage "Internal error: invalid deployMethod: $deployMethod"
;;
esac
echo "Deployment started at $(date)"
$metricsWriteDatapoint "testnet-deploy net-start-begin=1"
SECONDS=0
declare leaderDeployTime=
startLeader "$leaderIp" "$netLogDir/leader-$leaderIp.log"
leaderDeployTime=$SECONDS
$metricsWriteDatapoint "testnet-deploy net-leader-started=1"
SECONDS=0
pids=()
loopCount=0
for ipAddress in "${validatorIpList[@]}"; do
startValidator "$ipAddress"
# Staggering validator startup time. If too many validators
# bootup simultaneously, leader node gets more rsync requests
# from the validators than it can handle.
((loopCount++ % 2 == 0)) && sleep 2
done
for pid in "${pids[@]}"; do
declare ok=true
wait "$pid" || ok=false
if ! $ok; then
cat "$netLogDir/validator-$pid.log"
echo ^^^ +++
exit 1
fi
done
$metricsWriteDatapoint "testnet-deploy net-validators-started=1"
validatorDeployTime=$SECONDS
sanity
SECONDS=0
for ipAddress in "${clientIpList[@]}"; do
startClient "$ipAddress" "$netLogDir/client-$ipAddress.log"
done
clientDeployTime=$SECONDS
$metricsWriteDatapoint "testnet-deploy net-start-complete=1"
if [[ $deployMethod = "snap" ]]; then
declare networkVersion=unknown
IFS=\ read -r _ networkVersion _ < <(
ssh "${sshOptions[@]}" "$leaderIp" \
"snap info solana | grep \"^installed:\""
)
networkVersion=${networkVersion/0+git./}
$metricsWriteDatapoint "testnet-deploy version=\"$networkVersion\""
fi
echo
echo "+++ Deployment Successful"
echo "Leader deployment took $leaderDeployTime seconds"
echo "Validator deployment (${#validatorIpList[@]} instances) took $validatorDeployTime seconds"
echo "Client deployment (${#clientIpList[@]} instances) took $clientDeployTime seconds"
echo "Network start logs in $netLogDir:"
ls -l "$netLogDir"
}
stopNode() {
local ipAddress=$1
echo "--- Stopping node: $ipAddress"
(
set -x
ssh "${sshOptions[@]}" "$ipAddress" "
set -x
if snap list solana; then
sudo snap set solana mode=
sudo snap remove solana
fi
! tmux list-sessions || tmux kill-session
for pattern in solana- remote- oom-monitor net-stats; do
pkill -9 \$pattern
done
"
) || true
}
stop() {
SECONDS=0
$metricsWriteDatapoint "testnet-deploy net-stop-begin=1"
stopNode "$leaderIp"
for ipAddress in "${validatorIpList[@]}" "${clientIpList[@]}"; do
stopNode "$ipAddress"
done
$metricsWriteDatapoint "testnet-deploy net-stop-complete=1"
echo "Stopping nodes took $SECONDS seconds"
}
case $command in
restart)
stop
start
;;
start)
start
;;
sanity)
sanity
;;
stop)
stop
;;
*)
echo "Internal error: Unknown command: $command"
exit 1
esac

1
net/remote/README.md Normal file

@@ -0,0 +1 @@
Scripts that run on the remote testnet nodes

83
net/remote/remote-client.sh Executable file

@@ -0,0 +1,83 @@
#!/bin/bash -e
cd "$(dirname "$0")"/../..
echo "$(date) | $0 $*" > client.log
deployMethod="$1"
entrypointIp="$2"
numNodes="$3"
RUST_LOG="$4"
export RUST_LOG=${RUST_LOG:-solana=info} # if RUST_LOG is unset, default to info
missing() {
echo "Error: $1 not specified"
exit 1
}
[[ -n $deployMethod ]] || missing deployMethod
[[ -n $entrypointIp ]] || missing entrypointIp
[[ -n $numNodes ]] || missing numNodes
source net/common.sh
loadConfigFile
threadCount=$(nproc)
if [[ $threadCount -gt 4 ]]; then
threadCount=4
fi
case $deployMethod in
snap)
net/scripts/rsync-retry.sh -vPrc "$entrypointIp:~/solana/solana.snap" .
sudo snap install solana.snap --devmode --dangerous
solana_bench_tps=/snap/bin/solana.bench-tps
solana_keygen=/snap/bin/solana.keygen
;;
local)
PATH="$HOME"/.cargo/bin:"$PATH"
export USE_INSTALL=1
export SOLANA_DEFAULT_METRICS_RATE=1
net/scripts/rsync-retry.sh -vPrc "$entrypointIp:~/.cargo/bin/solana*" ~/.cargo/bin/
solana_bench_tps=solana-bench-tps
solana_keygen=solana-keygen
;;
*)
echo "Unknown deployment method: $deployMethod"
exit 1
esac
scripts/oom-monitor.sh > oom-monitor.log 2>&1 &
scripts/net-stats.sh > net-stats.log 2>&1 &
! tmux list-sessions || tmux kill-session
clientCommand="\
$solana_bench_tps \
--network $entrypointIp:8001 \
--identity client.json \
--num-nodes $numNodes \
--duration 600 \
--sustained \
--threads $threadCount \
"
keygenCommand="$solana_keygen -o client.json"
tmux new -s solana-bench-tps -d "
[[ -r client.json ]] || {
echo '$ $keygenCommand' | tee -a client.log
$keygenCommand >> client.log 2>&1
}
while true; do
echo === Client start: \$(date) | tee -a client.log
$metricsWriteDatapoint 'testnet-deploy client-begin=1'
echo '$ $clientCommand' | tee -a client.log
$clientCommand >> client.log 2>&1
$metricsWriteDatapoint 'testnet-deploy client-complete=1'
done
"
sleep 1
tmux capture-pane -t solana-bench-tps -p -S -100

113
net/remote/remote-node.sh Executable file

@@ -0,0 +1,113 @@
#!/bin/bash -e
cd "$(dirname "$0")"/../..
deployMethod="$1"
nodeType="$2"
publicNetwork="$3"
entrypointIp="$4"
numNodes="$5"
RUST_LOG="$6"
missing() {
echo "Error: $1 not specified"
exit 1
}
[[ -n $deployMethod ]] || missing deployMethod
[[ -n $nodeType ]] || missing nodeType
[[ -n $publicNetwork ]] || missing publicNetwork
[[ -n $entrypointIp ]] || missing entrypointIp
[[ -n $numNodes ]] || missing numNodes
cat > deployConfig <<EOF
deployMethod="$deployMethod"
entrypointIp="$entrypointIp"
numNodes="$numNodes"
EOF
source net/common.sh
loadConfigFile
if [[ $publicNetwork = true ]]; then
setupArgs="-p"
else
setupArgs="-l"
fi
case $deployMethod in
snap)
SECONDS=0
[[ $nodeType = leader ]] ||
net/scripts/rsync-retry.sh -vPrc "$entrypointIp:~/solana/solana.snap" .
sudo snap install solana.snap --devmode --dangerous
commonNodeConfig="\
leader-ip=$entrypointIp \
default-metrics-rate=1 \
metrics-config=$SOLANA_METRICS_CONFIG \
rust-log=$RUST_LOG \
setup-args=$setupArgs \
"
if [[ -e /dev/nvidia0 ]]; then
commonNodeConfig="$commonNodeConfig enable-cuda=1"
fi
if [[ $nodeType = leader ]]; then
nodeConfig="mode=leader+drone $commonNodeConfig"
ln -sf -T /var/snap/solana/current/leader/current leader.log
ln -sf -T /var/snap/solana/current/drone/current drone.log
else
nodeConfig="mode=validator $commonNodeConfig"
ln -sf -T /var/snap/solana/current/validator/current validator.log
fi
logmarker="solana deploy $(date)/$RANDOM"
logger "$logmarker"
# shellcheck disable=SC2086 # Don't want to double quote "$nodeConfig"
sudo snap set solana $nodeConfig
snap info solana
sudo snap get solana
echo Slight delay to get more syslog output
sleep 2
sudo grep -Pzo "$logmarker(.|\\n)*" /var/log/syslog
echo "Succeeded in ${SECONDS} seconds"
;;
local)
PATH="$HOME"/.cargo/bin:"$PATH"
export USE_INSTALL=1
export RUST_LOG
export SOLANA_DEFAULT_METRICS_RATE=1
./fetch-perf-libs.sh
export LD_LIBRARY_PATH="$PWD/target/perf-libs:$LD_LIBRARY_PATH"
scripts/oom-monitor.sh > oom-monitor.log 2>&1 &
scripts/net-stats.sh > net-stats.log 2>&1 &
case $nodeType in
leader)
./multinode-demo/setup.sh -t leader $setupArgs
./multinode-demo/drone.sh > drone.log 2>&1 &
./multinode-demo/leader.sh > leader.log 2>&1 &
;;
validator)
net/scripts/rsync-retry.sh -vPrc "$entrypointIp:~/.cargo/bin/solana*" ~/.cargo/bin/
./multinode-demo/setup.sh -t validator $setupArgs
./multinode-demo/validator.sh "$entrypointIp":~/solana "$entrypointIp:8001" >validator.log 2>&1 &
;;
*)
echo "Error: unknown node type: $nodeType"
exit 1
;;
esac
;;
*)
echo "Unknown deployment method: $deployMethod"
exit 1
esac

153
net/remote/remote-sanity.sh Executable file

@@ -0,0 +1,153 @@
#!/bin/bash -e
#
# This script is to be run on the leader node
#
cd "$(dirname "$0")"/../..
deployMethod=
entrypointIp=
numNodes=
[[ -r deployConfig ]] || {
echo deployConfig missing
exit 1
}
# shellcheck source=/dev/null # deployConfig is written by remote-node.sh
source deployConfig
missing() {
echo "Error: $1 not specified"
exit 1
}
[[ -n $deployMethod ]] || missing deployMethod
[[ -n $entrypointIp ]] || missing entrypointIp
[[ -n $numNodes ]] || missing numNodes
ledgerVerify=true
validatorSanity=true
rejectExtraNodes=false
while [[ $1 = -o ]]; do
opt="$2"
shift 2
case $opt in
noLedgerVerify)
ledgerVerify=false
;;
noValidatorSanity)
validatorSanity=false
;;
rejectExtraNodes)
rejectExtraNodes=true
;;
*)
echo "Error: unknown option: $opt"
exit 1
;;
esac
done
source net/common.sh
loadConfigFile
case $deployMethod in
snap)
PATH="/snap/bin:$PATH"
export USE_SNAP=1
entrypointRsyncUrl="$entrypointIp"
solana_bench_tps=solana.bench-tps
solana_ledger_tool=solana.ledger-tool
solana_keygen=solana.keygen
ledger=/var/snap/solana/current/config/ledger
client_id=~/snap/solana/current/config/client-id.json
;;
local)
PATH="$HOME"/.cargo/bin:"$PATH"
export USE_INSTALL=1
entrypointRsyncUrl="$entrypointIp:~/solana"
solana_bench_tps=solana-bench-tps
solana_ledger_tool=solana-ledger-tool
solana_keygen=solana-keygen
ledger=config/ledger
client_id=config/client-id.json
;;
*)
echo "Unknown deployment method: $deployMethod"
exit 1
esac
echo "--- $entrypointIp: wallet sanity"
(
set -x
scripts/wallet-sanity.sh "$entrypointIp:8001"
)
echo "+++ $entrypointIp: node count ($numNodes expected)"
(
set -x
$solana_keygen -o "$client_id"
maybeRejectExtraNodes=
if $rejectExtraNodes; then
maybeRejectExtraNodes="--reject-extra-nodes"
fi
$solana_bench_tps \
--network "$entrypointIp:8001" \
--identity "$client_id" \
--num-nodes "$numNodes" \
$maybeRejectExtraNodes \
--converge-only
)
echo "--- $entrypointIp: verify ledger"
if $ledgerVerify; then
if [[ -d $ledger ]]; then
(
set -x
rm -rf /var/tmp/ledger-verify
du -hs "$ledger"
time cp -r "$ledger" /var/tmp/ledger-verify
time $solana_ledger_tool --ledger /var/tmp/ledger-verify verify
)
else
echo "^^^ +++"
echo "Ledger verify skipped: directory does not exist: $ledger"
fi
else
echo "^^^ +++"
echo "Note: ledger verify disabled"
fi
echo "--- $entrypointIp: validator sanity"
if $validatorSanity; then
(
set -ex -o pipefail
./multinode-demo/setup.sh -t validator
timeout 10s ./multinode-demo/validator.sh "$entrypointRsyncUrl" "$entrypointIp:8001" 2>&1 | tee validator.log
) || {
exitcode=$?
[[ $exitcode -eq 124 ]] || exit $exitcode
}
wc -l validator.log
if grep -C100 panic validator.log; then
echo "^^^ +++"
echo "Panic observed"
exit 1
else
echo "Validator log looks ok"
fi
else
echo "^^^ +++"
echo "Note: validator sanity disabled"
fi
echo --- Pass

@@ -0,0 +1,20 @@
#!/bin/bash -ex
[[ $(uname) = Linux ]] || exit 1
[[ $USER = root ]] || exit 1
[[ -d /home/solana/.ssh ]] || exit 1
# /solana-authorized_keys contains the public keys for users that should
# automatically be granted access to ALL testnets.
#
# To add an entry into this list:
# 1. Run: ssh-keygen -t ecdsa -N '' -f ~/.ssh/id-solana-testnet
# 2. Inline ~/.ssh/id-solana-testnet.pub below
cat > /solana-authorized_keys <<EOF
ecdsa-sha2-nistp256 AAAAE2VjZHNhLXNoYTItbmlzdHAyNTYAAAAIbmlzdHAyNTYAAABBBFBNwLw0i+rI312gWshojFlNw9NV7WfaKeeUsYADqOvM2o4yrO2pPw+sgW8W+/rPpVyH7zU9WVRgTME8NgFV1Vc=
EOF
sudo -u solana bash -c "
cat /solana-authorized_keys >> /home/solana/.ssh/authorized_keys
"

@@ -0,0 +1,27 @@
#!/bin/bash -ex
[[ $(uname) = Linux ]] || exit 1
[[ $USER = root ]] || exit 1
adduser solana --gecos "" --disabled-password --quiet
adduser solana sudo
echo "solana ALL=(ALL) NOPASSWD:ALL" >> /etc/sudoers
id solana
[[ -r /solana-id_ecdsa ]] || exit 1
[[ -r /solana-id_ecdsa.pub ]] || exit 1
sudo -u solana bash -c "
mkdir -p /home/solana/.ssh/
cd /home/solana/.ssh/
cp /solana-id_ecdsa.pub authorized_keys
umask 377
cp /solana-id_ecdsa id_ecdsa
echo \"
Host *
BatchMode yes
IdentityFile ~/.ssh/id_ecdsa
StrictHostKeyChecking no
\" > config
"

@@ -0,0 +1,20 @@
#!/bin/bash -ex
#
# Prevent background upgrades that block |apt-get|
#
# TODO: This approach is pretty uncompromising. An alternative solution that
# doesn't involve deleting system files would be welcome.
[[ $(uname) = Linux ]] || exit 1
[[ $USER = root ]] || exit 1
rm -rf /usr/lib/apt/apt.systemd.daily
rm -rf /usr/bin/unattended-upgrade
killall apt.systemd.daily || true
killall unattended-upgrade || true
while fuser /var/lib/dpkg/lock; do
echo Waiting for lock release...
sleep 1
done

240
net/scripts/ec2-provider.sh Normal file

@@ -0,0 +1,240 @@
# |source| this file
#
# Utilities for working with EC2 instances
#
zone=
region=
cloud_SetZone() {
zone="$1"
# AWS region is zone with the last character removed
region="${zone:0:$((${#zone} - 1))}"
}
# Set the default zone
cloud_SetZone "us-east-1b"
# sshPrivateKey should be globally defined whenever this function is called.
#
# TODO: Remove usage of the sshPrivateKey global
__cloud_SshPrivateKeyCheck() {
# shellcheck disable=SC2154
if [[ -z $sshPrivateKey ]]; then
echo Error: sshPrivateKey not defined
exit 1
fi
if [[ ! -r $sshPrivateKey ]]; then
echo "Error: file is not readable: $sshPrivateKey"
exit 1
fi
}
#
# __cloud_FindInstances
#
# Find instances with name matching the specified pattern.
#
# For each matching instance, an entry in the `instances` array will be added with the
# following information about the instance:
# "name:public IP:private IP"
#
# filter - The instances to filter on
#
# examples:
# $ __cloud_FindInstances "exact-machine-name"
# $ __cloud_FindInstances "all-machines-with-a-common-machine-prefix*"
#
__cloud_FindInstances() {
declare filter="$1"
instances=()
declare name publicIp privateIp
while read -r name publicIp privateIp; do
printf "%-30s | publicIp=%-16s privateIp=%s\n" "$name" "$publicIp" "$privateIp"
instances+=("$name:$publicIp:$privateIp")
done < <(aws ec2 describe-instances \
--region "$region" \
--filters \
"Name=tag:name,Values=$filter" \
"Name=instance-state-name,Values=pending,running" \
--query "Reservations[].Instances[].[InstanceId,PublicIpAddress,PrivateIpAddress]" \
--output text
)
}
#
# cloud_FindInstances [namePrefix]
#
# Find instances with names matching the specified prefix
#
# For each matching instance, an entry in the `instances` array will be added with the
# following information about the instance:
# "name:public IP:private IP"
#
# namePrefix - The instance name prefix to look for
#
# examples:
# $ cloud_FindInstances all-machines-with-a-common-machine-prefix
#
cloud_FindInstances() {
declare namePrefix="$1"
__cloud_FindInstances "$namePrefix*"
}
#
# cloud_FindInstance [name]
#
# Find an instance with a name matching the exact pattern.
#
# For each matching instance, an entry in the `instances` array will be added with the
# following information about the instance:
# "name:public IP:private IP"
#
# name - The instance name to look for
#
# examples:
# $ cloud_FindInstance exact-machine-name
#
cloud_FindInstance() {
declare name="$1"
__cloud_FindInstances "$name"
}
#
# cloud_CreateInstances [networkName] [namePrefix] [numNodes] [imageName]
# [machineType] [bootDiskSize] [startupScript] [address]
#
# Creates one more identical instances.
#
# networkName - unique name of this testnet
# namePrefix - unique string to prefix all the instance names with
# numNodes - number of instances to create
# imageName - Disk image for the instances
# machineType - GCE machine type
# bootDiskSize - Optional size of the boot disk in GB
# startupScript - Optional startup script to execute when the instance boots
# address - Optional name of the GCE static IP address to attach to the
# instance. Requires that |numNodes| = 1 and that addressName
# has been provisioned in the GCE region that is hosting |zone|
#
# Tip: use cloud_FindInstances to locate the instances once this function
# returns
cloud_CreateInstances() {
declare networkName="$1"
declare namePrefix="$2"
declare numNodes="$3"
declare imageName="$4"
declare machineType="$5"
declare optionalBootDiskSize="$6"
declare optionalStartupScript="$7"
declare optionalAddress="$8"
__cloud_SshPrivateKeyCheck
(
set -x
aws ec2 delete-key-pair --region "$region" --key-name "$networkName"
aws ec2 import-key-pair --region "$region" --key-name "$networkName" \
--public-key-material file://"${sshPrivateKey}".pub
)
declare -a args
args=(
--key-name "$networkName"
--count "$numNodes"
--region "$region"
--placement "AvailabilityZone=$zone"
--security-groups testnet
--image-id "$imageName"
--instance-type "$machineType"
--tag-specifications "ResourceType=instance,Tags=[{Key=name,Value=$namePrefix}]"
)
if [[ -n $optionalBootDiskSize ]]; then
args+=(
--block-device-mapping "[{\"DeviceName\": \"/dev/sda1\", \"Ebs\": { \"VolumeSize\": $optionalBootDiskSize }}]"
)
fi
if [[ -n $optionalStartupScript ]]; then
args+=(
--user-data "file://$optionalStartupScript"
)
fi
if [[ -n $optionalAddress ]]; then
[[ $numNodes = 1 ]] || {
echo "Error: address may not be supplied when provisioning multiple nodes: $optionalAddress"
exit 1
}
fi
(
set -x
aws ec2 run-instances "${args[@]}"
)
if [[ -n $optionalAddress ]]; then
cloud_FindInstance "$namePrefix"
if [[ ${#instances[@]} -ne 1 ]]; then
echo "Failed to find newly created instance: $namePrefix"
fi
declare instanceId
IFS=: read -r instanceId _ < <(echo "${instances[0]}")
(
set -x
# TODO: Poll that the instance has moved to the 'running' state instead of
# blindly sleeping for 30 seconds...
sleep 30
aws ec2 associate-address \
--instance-id "$instanceId" \
--region "$region" \
--allocation-id "$optionalAddress"
)
fi
}
#
# cloud_DeleteInstances
#
# Deletes all the instances listed in the `instances` array
#
cloud_DeleteInstances() {
if [[ ${#instances[0]} -eq 0 ]]; then
echo No instances to delete
return
fi
declare names=("${instances[@]/:*/}")
(
set -x
aws ec2 terminate-instances --region "$region" --instance-ids "${names[@]}"
)
}
#
# cloud_FetchFile [instanceName] [publicIp] [remoteFile] [localFile]
#
# Fetch a file from the given instance. This function uses a cloud-specific
# mechanism to fetch the file
#
cloud_FetchFile() {
# shellcheck disable=SC2034 # instanceName is unused
declare instanceName="$1"
declare publicIp="$2"
declare remoteFile="$3"
declare localFile="$4"
__cloud_SshPrivateKeyCheck
(
set -x
scp \
-o "StrictHostKeyChecking=no" \
-o "UserKnownHostsFile=/dev/null" \
-o "User=solana" \
-o "IdentityFile=$sshPrivateKey" \
-o "LogLevel=ERROR" \
-F /dev/null \
"solana@$publicIp:$remoteFile" "$localFile"
)
}

201
net/scripts/gce-provider.sh Normal file

@@ -0,0 +1,201 @@
# |source| this file
#
# Utilities for working with GCE instances
#
# Default zone
zone="us-west1-b"
cloud_SetZone() {
zone="$1"
}
#
# __cloud_FindInstances
#
# Find instances matching the specified pattern.
#
# For each matching instance, an entry in the `instances` array will be added with the
# following information about the instance:
# "name:zone:public IP:private IP"
#
# filter - The instances to filter on
#
# examples:
# $ __cloud_FindInstances "name=exact-machine-name"
# $ __cloud_FindInstances "name~^all-machines-with-a-common-machine-prefix"
#
__cloud_FindInstances() {
declare filter="$1"
instances=()
declare name zone publicIp privateIp status
while read -r name publicIp privateIp status; do
printf "%-30s | publicIp=%-16s privateIp=%s staus=%s\n" "$name" "$publicIp" "$privateIp" "$status"
instances+=("$name:$publicIp:$privateIp")
done < <(gcloud compute instances list \
--filter "$filter" \
--format 'value(name,networkInterfaces[0].accessConfigs[0].natIP,networkInterfaces[0].networkIP,status)')
}
#
# cloud_FindInstances [namePrefix]
#
# Find instances with names matching the specified prefix
#
# For each matching instance, an entry in the `instances` array will be added with the
# following information about the instance:
# "name:public IP:private IP"
#
# namePrefix - The instance name prefix to look for
#
# examples:
# $ cloud_FindInstances all-machines-with-a-common-machine-prefix
#
cloud_FindInstances() {
declare namePrefix="$1"
__cloud_FindInstances "name~^$namePrefix"
}
#
# cloud_FindInstance [name]
#
# Find an instance with a name matching the exact pattern.
#
# For each matching instance, an entry in the `instances` array will be added with the
# following information about the instance:
# "name:public IP:private IP"
#
# name - The instance name to look for
#
# examples:
# $ cloud_FindInstance exact-machine-name
#
cloud_FindInstance() {
declare name="$1"
__cloud_FindInstances "name=$name"
}
#
# cloud_CreateInstances [networkName] [namePrefix] [numNodes] [imageName]
# [machineType] [bootDiskSize] [enableGpu]
# [startupScript] [address]
#
# Creates one more identical instances.
#
# networkName - unique name of this testnet
# namePrefix - unique string to prefix all the instance names with
# numNodes - number of instances to create
# imageName - Disk image for the instances
# machineType - GCE machine type. Note that this may also include an
# `--accelerator=` or other |gcloud compute instances create|
# options
# bootDiskSize - Optional size of the boot disk in GB
# enableGpu - Optionally enable GPU, use the value "true" to enable
# eg, request 4 K80 GPUs with "count=4,type=nvidia-tesla-k80"
# startupScript - Optional startup script to execute when the instance boots
# address - Optional name of the GCE static IP address to attach to the
# instance. Requires that |numNodes| = 1 and that addressName
# has been provisioned in the GCE region that is hosting `$zone`
#
# Tip: use cloud_FindInstances to locate the instances once this function
# returns
cloud_CreateInstances() {
declare networkName="$1"
declare namePrefix="$2"
declare numNodes="$3"
declare imageName="$4"
declare machineType="$5"
declare optionalBootDiskSize="$6"
declare optionalStartupScript="$7"
declare optionalAddress="$8"
declare optionalBootDiskType="$9"
declare nodes
if [[ $numNodes = 1 ]]; then
nodes=("$namePrefix")
else
read -ra nodes <<<$(seq -f "${namePrefix}%0${#numNodes}g" 1 "$numNodes")
fi
declare -a args
args=(
--zone "$zone"
--tags testnet
--metadata "testnet=$networkName"
--image "$imageName"
--maintenance-policy TERMINATE
--no-restart-on-failure
)
# shellcheck disable=SC2206 # Do not want to quote $machineType as it may contain extra args
args+=(--machine-type $machineType)
if [[ -n $optionalBootDiskSize ]]; then
args+=(
--boot-disk-size "${optionalBootDiskSize}GB"
)
fi
if [[ -n $optionalStartupScript ]]; then
args+=(
--metadata-from-file "startup-script=$optionalStartupScript"
)
fi
if [[ -n $optionalBootDiskType ]]; then
args+=(
--boot-disk-type "${optionalBootDiskType}"
)
fi
if [[ -n $optionalAddress ]]; then
[[ $numNodes = 1 ]] || {
echo "Error: address may not be supplied when provisioning multiple nodes: $optionalAddress"
exit 1
}
args+=(
--address "$optionalAddress"
)
fi
(
set -x
gcloud beta compute instances create "${nodes[@]}" "${args[@]}"
)
}
#
# cloud_DeleteInstances
#
# Deletes all the instances listed in the `instances` array
#
cloud_DeleteInstances() {
if [[ ${#instances[0]} -eq 0 ]]; then
echo No instances to delete
return
fi
declare names=("${instances[@]/:*/}")
(
set -x
gcloud beta compute instances delete --zone "$zone" --quiet "${names[@]}"
)
}
#
# cloud_FetchFile [instanceName] [publicIp] [remoteFile] [localFile]
#
# Fetch a file from the given instance. This function uses a cloud-specific
# mechanism to fetch the file
#
cloud_FetchFile() {
declare instanceName="$1"
# shellcheck disable=SC2034 # publicIp is unused
declare publicIp="$2"
declare remoteFile="$3"
declare localFile="$4"
(
set -x
gcloud compute scp --zone "$zone" "$instanceName:$remoteFile" "$localFile"
)
}

25
net/scripts/install-docker.sh Executable file

@@ -0,0 +1,25 @@
#!/bin/bash -ex
[[ $(uname) = Linux ]] || exit 1
[[ $USER = root ]] || exit 1
apt-get update
apt-get install -y \
apt-transport-https \
ca-certificates \
curl \
software-properties-common \
curl -fsSL https://download.docker.com/linux/ubuntu/gpg | apt-key add -
add-apt-repository \
"deb [arch=amd64] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable"
apt-get update
apt-get install -y docker-ce
docker run hello-world
# Grant the solana user access to docker
if id solana; then
addgroup solana docker
fi

Some files were not shown because too many files have changed in this diff Show More