Commit Graph

173 Commits

Author SHA1 Message Date
behzad nouri
3f63ed9a72 removes OrderedIterator and transaction batch iteration order (#16153)
In TransactionBatch,
https://github.com/solana-labs/solana/blob/e50f59844/runtime/src/transaction_batch.rs#L4-L11
lock_results[i] is aligned with transactions[iteration_order[i]]:
https://github.com/solana-labs/solana/blob/e50f59844/runtime/src/bank.rs#L2414-L2424
https://github.com/solana-labs/solana/blob/e50f59844/runtime/src/accounts.rs#L788-L817

However load_and_execute_transactions is iterating over
  lock_results[iteration_order[i]]
https://github.com/solana-labs/solana/blob/e50f59844/runtime/src/bank.rs#L2878-L2889
and then returning i as for the index of the retryable transaction.

If iteratorion_order is [1, 2, 0], and i is 0, then:
  lock_results[iteration_order[i]] = lock_results[1]
which corresponds to
  transactions[iteration_order[1]] = transactions[2]
so neither i = 0, nor iteration_order[i] = 1 gives the correct index for the
corresponding transaction (which is 2).

This commit removes OrderedIterator and transaction batch iteration order
entirely. There is only one place in blockstore processor which the
iteration order is not ordinal:
https://github.com/solana-labs/solana/blob/e50f59844/ledger/src/blockstore_processor.rs#L269-L271
It seems like, instead of using an iteration order, that can shuffle entry
transactions in-place.
2021-03-31 23:59:19 +00:00
behzad nouri
4f82b897bc buffers data shreds to make larger erasure coded sets (#15849)
Broadcast stage batches up to 8 entries:
https://github.com/solana-labs/solana/blob/79280b304/core/src/broadcast_stage/broadcast_utils.rs#L26-L29
which will be serialized into some number of shreds and chunked into FEC
sets of at most 32 shreds each:
https://github.com/solana-labs/solana/blob/79280b304/ledger/src/shred.rs#L576-L597
So depending on the size of entries, FEC sets can be small, which may
aggravate loss rate.
For example 16 FEC sets of 2:2 data/code shreds each have higher loss
rate than one 32:32 set.

This commit broadcasts data shreds immediately, but also buffers them
until it has a batch of 32 data shreds, at which point 32 coding shreds
are generated and broadcasted.
2021-03-23 14:52:38 +00:00
Jeff Washington (jwash)
57ba86c821 eliminate lock on record (#15929)
* eliminate lock on record

* use same error as MaxHeightReached

* clippy

* review feedback

* refactor should_tick code

* pr feedback
2021-03-23 09:10:04 -05:00
carllin
c1ba265dd9 Wallclock BankingStage Throttle (#15731) 2021-03-15 17:11:15 -07:00
sakridge
d09112fa6d PoH batch size calibration (#15717) 2021-03-05 16:01:21 -08:00
sakridge
830be855dc Forward and hold packets (#15634) 2021-03-03 10:23:05 -08:00
carllin
ae96ba3459 Plumb slot update pubsub notifications (#15488) 2021-02-28 23:29:11 -08:00
sakridge
1b59b163dd Add max retransmit and shred insert slot (#15475) 2021-02-23 13:06:33 -08:00
Trent Nelson
7f7370c306 Re-allow clippy::integer_arithmetic at crate-level 2021-02-17 13:55:08 -07:00
carllin
629dcd0f39 Cleanup buffered packets (#15210) 2021-02-12 03:27:37 -08:00
behzad nouri
e1021d9f83 removes redundant epoch stakes cache in retransmit (#14781)
Following d6d76219b, staked nodes computed from vote accounts are
already cached in runtime::Stakes, so the caching in retransmit_stage is
redundant.
2021-01-24 21:15:09 +00:00
sakridge
5c95d8e963 Shred filter (#14030) 2020-12-10 07:54:15 -08:00
sakridge
c5fe076432 Better dupe detection (#13992) 2020-12-09 23:14:31 -08:00
behzad nouri
cbea9ebc34 indexes nodes' contact infos in crds table (#13553)
In several places in gossip code, the entire crds table is scanned only
to filter out nodes' contact infos. Currently on mainnet, crds table is
of size ~70k, while there are only ~470 nodes. So the full table scan is
inefficient. Instead we may maintain an index of only nodes' contact
infos.
2020-11-15 16:38:04 +00:00
sakridge
b4cf968e14 Add back shredding broadcast stats (#13463) 2020-11-09 23:04:27 -08:00
behzad nouri
37c8842bcb scans crds table in parallel for finding old labels (#13073)
From runtime profiles, the majority time of ClusterInfo::handle_purge
https://github.com/solana-labs/solana/blob/0776fa05c/core/src/cluster_info.rs#L1605-L1626
is spent scanning crds table finding old labels:
https://github.com/solana-labs/solana/blob/0776fa05c/core/src/crds.rs#L175-L197

This can be done in parallel given that gossip thread-pool:
https://github.com/solana-labs/solana/blob/0776fa05c/core/src/cluster_info.rs#L1637-L1641
is idle when handle_purge is invoked:
https://github.com/solana-labs/solana/blob/0776fa05c/core/src/cluster_info.rs#L1681
2020-10-23 14:17:37 +00:00
Michael Vines
959880db60 Remove unused pubkey::Pubkey imports 2020-10-21 19:08:13 -07:00
Michael Vines
17c391121a Run codemod --extensions rs Hash::new_rand solana_sdk:#️⃣:new_rand 2020-10-21 19:08:13 -07:00
Michael Vines
7bc073defe Run codemod --extensions rs Pubkey::new_rand solana_sdk::pubkey::new_rand 2020-10-21 19:08:13 -07:00
behzad nouri
537bbde22e builds crds filters in parallel (#12360)
Based on run-time profiles, the majority time of new_pull_requests is
spent building bloom filters, in hashing and bit-vec ops.

This commit builds crds filters in parallel using rayon constructs. The
added benchmark shows ~5x speedup (4-core machine, 8 threads).
2020-09-29 23:06:02 +00:00
Ryo Onodera
cb8661bd49 Persistent tower (#10718)
* Save/restore Tower

* Avoid unwrap()

* Rebase cleanups

* Forcibly pass test

* Correct reconcilation of votes after validator resume

* d b g

* Add more tests

* fsync and fix test

* Add test

* Fix fmt

* Debug

* Fix tests...

* save

* Clarify error message and code cleaning around it

* Move most of code out of tower save hot codepath

* Proper comment for the lack of fsync on tower

* Clean up

* Clean up

* Simpler type alias

* Manage tower-restored ancestor slots without banks

* Add comment

* Extract long code blocks...

* Add comment

* Simplify returned tuple...

* Tweak too aggresive log

* Fix typo...

* Add test

* Update comment

* Improve test to require non-empty stray restored slots

* Measure tower save and dump all tower contents

* Log adjust and add threshold related assertions

* cleanup adjust

* Properly lower stray restored slots priority...

* Rust fmt

* Fix test....

* Clarify comments a bit and add TowerError::TooNew

* Further clean-up arround TowerError

* Truly create ancestors by excluding last vote slot

* Add comment for stray_restored_slots

* Add comment for stray_restored_slots

* Use BTreeSet

* Consider root_slot into post-replay adjustment

* Tweak logging

* Add test for stray_restored_ancestors

* Reorder some code

* Better names for unit tests

* Add frozen_abi to SavedTower

* Fold long lines

* Tweak stray ancestors and too old slot history

* Re-adjust error conditon of too old slot history

* Test normal ancestors is checked before stray ones

* Fix conflict, update tests, adjust behavior a bit

* Fix test

* Address review comments

* Last touch!

* Immediately after creating cleaning pr

* Revert stray slots

* Revert comment...

* Report error as metrics

* Revert not to panic! and ignore unfixable test...

* Normalize lockouts.root_slot more strictly

* Add comments for panic! and more assertions

* Proper initialize root without vote account

* Clarify code and comments based on review feedback

* Fix rebase

* Further simplify based on assured tower root

* Reorder code for more readability

Co-authored-by: Michael Vines <mvines@gmail.com>
2020-09-19 14:03:54 +09:00
behzad nouri
9b866d79fb shards crds values based on their hash prefix (#12187)
filter_crds_values checks every crds filter against every hash value:
https://github.com/solana-labs/solana/blob/ee646aa7/core/src/crds_gossip_pull.rs#L432
which can be inefficient if the filter's bit-mask only matches small
portion of the entire crds table.

This commit shards crds values into separate tables based on shard_bits
first bits of their hash prefix. Given a (mask, mask_bits) filter,
filtering crds can be done by inspecting only relevant shards.

If CrdsFilter.mask_bits <= shard_bits, then precisely only the crds
values which match (mask, mask_bits) bit pattern are traversed.
If CrdsFilter.mask_bits > shard_bits, then approximately only
1/2^shard_bits of crds values are inspected.

Benchmarking on a gce cluster of 20 nodes, I see ~10% improvement in
generate_pull_responses metric, but with larger clusters, crds table and
2^mask_bits are both larger, so the impact should be more significant.
2020-09-17 14:05:16 +00:00
behzad nouri
28f2fa3fd5 uses rust intrinsics to convert hashes to u64 (#12097) 2020-09-09 15:28:17 +00:00
Michael Vines
d15173ad9d Address latest nightly clippy lints, but globally disable stable_sort_primitive 2020-08-17 22:36:10 -07:00
carllin
7e25130529 Send votes from banking stage to vote listener (#11434)
*  Send votes from banking stage to vote listener

Co-authored-by: Carl <carl@solana.com>
2020-08-07 11:21:35 -07:00
carllin
bf18524368 Add hook for getting vote transactions on replay (#11264)
* Add hook for getting vote transactions on replay

Co-authored-by: Carl <carl@solana.com>
2020-07-29 23:17:40 -07:00
carllin
f1699721ef Bench RaptorQ (#10886)
Co-authored-by: Carl <carl@solana.com>
2020-07-02 18:31:32 -07:00
Greg Fitzgerald
1c498369b5 Remove fee-payer guesswork from Message and Transaction (#10776)
* Make Message::new_with_payer the default constructor

* Remove Transaction::new_[un]signed_instructions

These guess the fee-payer instead of stating it explicitly
2020-06-24 14:52:38 -06:00
Greg Fitzgerald
6ee222363e Move BankForks to solana_runtime (#10637)
* Move BankForks to solana_runtime

* Update imports
2020-06-17 15:27:03 +00:00
sakridge
0de6c444d6 Simd poh (#10604)
* Simd poh

* Fix poh verify bench
2020-06-16 23:03:26 -07:00
Greg Fitzgerald
2eb6f498a8 Remove redundant BankForks parameter (#10537) 2020-06-12 11:04:17 -06:00
Kristofer Peterson
e23340d89e Clippy cleanup for all targets and nighly rust (also support 1.44.0) (#10445)
* address warnings from 'rustup run beta cargo clippy --workspace'

minor refactoring in:
- cli/src/cli.rs
- cli/src/offline/blockhash_query.rs
- logger/src/lib.rs
- runtime/src/accounts_db.rs

expect some performance improvement AccountsDB::clean_accounts()

* address warnings from 'rustup run beta cargo clippy --workspace --tests'

* address warnings from 'rustup run nightly cargo clippy --workspace --all-targets'

* rustfmt

* fix warning stragglers

* properly fix clippy warnings test_vote_subscribe()
replace ref-to-arc with ref parameters where arc not cloned

* Remove lock around JsonRpcRequestProcessor (#10417)

automerge

* make ancestors parameter optional to avoid forcing construction of empty hash maps

Co-authored-by: Greg Fitzgerald <greg@solana.com>
2020-06-09 09:38:14 +09:00
sakridge
2cf719ac2c Cache tvu peers for broadcast (#10373) 2020-06-03 08:24:05 -07:00
carllin
97f2bcff69 master: Add nonce to shreds repairs, add shred data size to header (#10109)
* Add nonce to shreds/repairs

* Add data shred size to header

Co-authored-by: Carl <carl@solana.com>
2020-05-19 12:38:18 -07:00
Jack May
eb1acaf927 Remove archiver and storage program (#9992)
automerge
2020-05-14 18:22:47 -07:00
Greg Fitzgerald
76b1c2baf0 One less alloc per transaction (#9705)
* One less alloc per transaction

* Fix benches

* Fix compiler warnings in bench build

* Fix move build

* Fix bench
2020-04-24 13:03:46 -06:00
carllin
bab3502260 Push down cluster_info lock (#9594)
* Push down cluster_info lock

* Rework budget decrement

Co-authored-by: Carl <carl@solana.com>
2020-04-21 12:54:45 -07:00
sakridge
69f1e487b3 Reduce cluster-info metrics. (#9465) 2020-04-14 21:21:58 -07:00
sakridge
4677cdb4c2 Optimize broadcast cluster_info critical section (#9327) 2020-04-06 17:36:22 -07:00
carllin
7b68628e6c Remove write lock (#9311)
* Remove write lock

Co-authored-by: Carl <carl@solana.com>
2020-04-05 15:18:45 -07:00
Jack May
d61191db40 fix bench warnings (#9277) 2020-04-02 21:56:38 -07:00
Michael Vines
18c1f0dfe9 Remove stub core/src/genesis_utils.rs (#8999) 2020-03-21 10:54:40 -07:00
anatoly yakovenko
9cedeb0a8d Pull streamer out into its own module. (#8917)
automerge
2020-03-17 23:30:23 -07:00
Tyera Eulberg
ab361a8073 Rename KeypairUtil to Signer (#8360)
automerge
2020-02-20 13:28:55 -08:00
Greg Fitzgerald
b5dba77056 Rename blocktree to blockstore (#7757)
automerge
2020-01-13 13:13:52 -08:00
anatoly yakovenko
97589f77f8 Pipeline broadcast socket transmit and blocktree record (#7481)
automerge
2019-12-16 17:11:18 -08:00
Pankaj Garg
d94041e98d Allow coding shred index to be different than data shred index (#7438)
* Allow coding shred index to be different than data shred index

* move fec_set_index to shred's common header

* fix bench
2019-12-12 16:50:29 -08:00
Tyera Eulberg
97ca6858b7 Write transaction status and fee into persistent store (#7030)
* Pass blocktree into execute_batch, if persist_transaction_status

* Add validator arg to enable persistent transaction status store

* Pass blocktree into banking_stage, if persist_transaction_status

* Add validator params to bash scripts

* Expose actual transaction statuses outside Bank; add tests

* Fix benches

* Offload transaction status writes to a separate thread

* Enable persistent transaction status along with rpc service

* nudge

* Review comments
2019-11-20 16:43:10 -07:00
Jack May
d2ed921bc6 Cleanup nightly warnings (#7055) 2019-11-19 20:15:37 -08:00
Sagar Dhawan
6bfe0fca1f Add a version field to shreds (#7023)
* Add a version field to shreds

* Clippy

* Fix Chacha Golden

* Fix shredder bench compile

* Fix blocktree bench compile
2019-11-18 18:05:02 -08:00