Revert "Lodestar gossip queues to wrap processRpcMessage() (#3554)"

[twoeths]

This reverts commit fcbc459.

Motivation

This reverts #3554. For our gossip queues, we wrapped processRpcMessage() instead of validate() because we found that getMsgId() is not cheap (it also uncompresses the message), it may be called and the message is still dropped in the end.

The down side is that we have different copies of same message in our queue, this leads to queue.jobs.shift() functions is called too much, see #3732

With new gossipsub v0.12, we use fast_msg_id function so it resolves the original concerns of #3554

Description

• Revert Lodestar gossip queues to wrap processRpcMessage() #3554: gossip queues to wrap validate() instead of processMessageRpc()
• So that only distinct gossip message is included in our gossip queues, duplicate messages are filtered before that time in processMessageRpc()
• If this is approved, this should be merged before Migrate to libp2p > 0.35.8 and gossipsub 0.12.2 #3661 to avoid the conflict

Closes #3691 , #3749

[codecov]

Codecov Report

Merging #3745 (8d54a70) into master (1c86d16) will increase coverage by 0.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #3745      +/-   ##
==========================================
+ Coverage   37.07%   37.08%   +0.01%     
==========================================
  Files         319      319              
  Lines        8715     8709       -6     
  Branches     1356     1354       -2     
==========================================
- Hits         3231     3230       -1     
+ Misses       5342     5337       -5     
  Partials      142      142              

[github-actions]

Performance Report

✔️ no performance regression detected

Full benchmark results

Benchmark suite	Current: feacff8	Previous: 1c86d16	Ratio
BeaconState.hashTreeRoot - No change	486.00 ns/op	546.00 ns/op	0.89
BeaconState.hashTreeRoot - 1 full validator	116.67 us/op	125.90 us/op	0.93
BeaconState.hashTreeRoot - 32 full validator	1.7103 ms/op	1.8860 ms/op	0.91
BeaconState.hashTreeRoot - 512 full validator	23.066 ms/op	24.359 ms/op	0.95
BeaconState.hashTreeRoot - 1 validator.effectiveBalance	110.02 us/op	123.37 us/op	0.89
BeaconState.hashTreeRoot - 32 validator.effectiveBalance	1.8577 ms/op	2.0019 ms/op	0.93
BeaconState.hashTreeRoot - 512 validator.effectiveBalance	25.046 ms/op	26.002 ms/op	0.96
BeaconState.hashTreeRoot - 1 balances	81.318 us/op	90.503 us/op	0.90
BeaconState.hashTreeRoot - 32 balances	685.29 us/op	757.82 us/op	0.90
BeaconState.hashTreeRoot - 512 balances	7.2300 ms/op	7.2279 ms/op	1.00
BeaconState.hashTreeRoot - 250000 balances	131.08 ms/op	130.03 ms/op	1.01
processSlot - 1 slots	43.384 us/op	44.616 us/op	0.97
processSlot - 32 slots	2.6401 ms/op	2.8076 ms/op	0.94
getCommitteeAssignments - req 1 vs - 250000 vc	4.7307 ms/op	5.3146 ms/op	0.89
getCommitteeAssignments - req 100 vs - 250000 vc	6.6284 ms/op	7.3899 ms/op	0.90
getCommitteeAssignments - req 1000 vs - 250000 vc	7.1411 ms/op	7.9322 ms/op	0.90
computeProposers - vc 250000	18.963 ms/op	19.698 ms/op	0.96
computeEpochShuffling - vc 250000	167.51 ms/op	182.62 ms/op	0.92
getNextSyncCommittee - vc 250000	308.63 ms/op	325.03 ms/op	0.95
altair processAttestation - 250000 vs - 7PWei normalcase	34.277 ms/op	34.540 ms/op	0.99
altair processAttestation - 250000 vs - 7PWei worstcase	36.069 ms/op	31.908 ms/op	1.13
altair processAttestation - setStatus - 1/6 committees join	11.416 ms/op	9.9794 ms/op	1.14
altair processAttestation - setStatus - 1/3 committees join	23.094 ms/op	21.166 ms/op	1.09
altair processAttestation - setStatus - 1/2 committees join	34.080 ms/op	33.818 ms/op	1.01
altair processAttestation - setStatus - 2/3 committees join	45.705 ms/op	43.317 ms/op	1.06
altair processAttestation - setStatus - 4/5 committees join	55.302 ms/op	48.608 ms/op	1.14
altair processAttestation - setStatus - 100% committees join	68.240 ms/op	62.949 ms/op	1.08
altair processAttestation - updateEpochParticipants - 1/6 committees join	11.594 ms/op	12.305 ms/op	0.94
altair processAttestation - updateEpochParticipants - 1/3 committees join	25.074 ms/op	21.080 ms/op	1.19
altair processAttestation - updateEpochParticipants - 1/2 committees join	17.088 ms/op	23.382 ms/op	0.73
altair processAttestation - updateEpochParticipants - 2/3 committees join	23.708 ms/op	21.451 ms/op	1.11
altair processAttestation - updateEpochParticipants - 4/5 committees join	21.277 ms/op	26.521 ms/op	0.80
altair processAttestation - updateEpochParticipants - 100% committees join	23.208 ms/op	23.520 ms/op	0.99
altair processAttestation - updateAllStatus	19.807 ms/op	18.991 ms/op	1.04
altair processBlock - 250000 vs - 7PWei normalcase	30.242 ms/op	37.809 ms/op	0.80
altair processBlock - 250000 vs - 7PWei worstcase	93.041 ms/op	101.13 ms/op	0.92
altair processEpoch - mainnet_e81889	930.93 ms/op	1.0309 s/op	0.90
mainnet_e81889 - altair beforeProcessEpoch	323.17 ms/op	257.55 ms/op	1.25
mainnet_e81889 - altair processJustificationAndFinalization	58.743 us/op	109.47 us/op	0.54
mainnet_e81889 - altair processInactivityUpdates	16.442 ms/op	18.026 ms/op	0.91
mainnet_e81889 - altair processRewardsAndPenalties	199.80 ms/op	249.45 ms/op	0.80
mainnet_e81889 - altair processRegistryUpdates	7.7510 us/op	11.891 us/op	0.65
mainnet_e81889 - altair processSlashings	1.8550 us/op	2.1860 us/op	0.85
mainnet_e81889 - altair processEth1DataReset	1.7270 us/op	1.9930 us/op	0.87
mainnet_e81889 - altair processEffectiveBalanceUpdates	5.7848 ms/op	6.9254 ms/op	0.84
mainnet_e81889 - altair processSlashingsReset	11.214 us/op	17.165 us/op	0.65
mainnet_e81889 - altair processRandaoMixesReset	17.491 us/op	23.890 us/op	0.73
mainnet_e81889 - altair processHistoricalRootsUpdate	2.1360 us/op	2.6860 us/op	0.80
mainnet_e81889 - altair processParticipationFlagUpdates	66.393 ms/op	80.880 ms/op	0.82
mainnet_e81889 - altair processSyncCommitteeUpdates	2.0800 us/op	3.2010 us/op	0.65
mainnet_e81889 - altair afterProcessEpoch	225.43 ms/op	212.87 ms/op	1.06
altair processInactivityUpdates - 250000 normalcase	69.588 ms/op	69.503 ms/op	1.00
altair processInactivityUpdates - 250000 worstcase	79.965 ms/op	65.883 ms/op	1.21
altair processParticipationFlagUpdates - 250000 anycase	64.464 ms/op	58.356 ms/op	1.10
altair processRewardsAndPenalties - 250000 normalcase	183.71 ms/op	203.00 ms/op	0.90
altair processRewardsAndPenalties - 250000 worstcase	204.98 ms/op	199.81 ms/op	1.03
altair processSyncCommitteeUpdates - 250000	326.18 ms/op	362.16 ms/op	0.90
Tree 40 250000 create	609.44 ms/op	790.11 ms/op	0.77
Tree 40 250000 get(125000)	297.76 ns/op	677.71 ns/op	0.44
Tree 40 250000 set(125000)	1.9314 us/op	1.7446 us/op	1.11
Tree 40 250000 toArray()	37.232 ms/op	42.081 ms/op	0.88
Tree 40 250000 iterate all - toArray() + loop	36.854 ms/op	35.520 ms/op	1.04
Tree 40 250000 iterate all - get(i)	107.97 ms/op	116.18 ms/op	0.93
MutableVector 250000 create	22.168 ms/op	23.278 ms/op	0.95
MutableVector 250000 get(125000)	11.528 ns/op	13.023 ns/op	0.89
MutableVector 250000 set(125000)	466.27 ns/op	460.02 ns/op	1.01
MutableVector 250000 toArray()	8.0419 ms/op	8.1762 ms/op	0.98
MutableVector 250000 iterate all - toArray() + loop	9.0236 ms/op	8.4657 ms/op	1.07
MutableVector 250000 iterate all - get(i)	3.7715 ms/op	3.7827 ms/op	1.00
Array 250000 create	5.6827 ms/op	4.7328 ms/op	1.20
Array 250000 clone - spread	2.0524 ms/op	2.1680 ms/op	0.95
Array 250000 get(125000)	0.96900 ns/op	1.0250 ns/op	0.95
Array 250000 set(125000)	0.95600 ns/op	1.0270 ns/op	0.93
Array 250000 iterate all - loop	148.07 us/op	167.79 us/op	0.88
effectiveBalanceIncrements clone Uint8Array 300000	70.677 us/op	60.551 us/op	1.17
effectiveBalanceIncrements clone MutableVector 300000	625.00 ns/op	517.00 ns/op	1.21
effectiveBalanceIncrements rw all Uint8Array 300000	266.27 us/op	301.74 us/op	0.88
effectiveBalanceIncrements rw all MutableVector 300000	172.74 ms/op	145.61 ms/op	1.19
aggregationBits - 2048 els - readonlyValues	187.27 us/op	170.52 us/op	1.10
aggregationBits - 2048 els - zipIndexesInBitList	34.495 us/op	30.606 us/op	1.13
regular array get 100000 times	67.405 us/op	67.398 us/op	1.00
wrappedArray get 100000 times	67.462 us/op	67.395 us/op	1.00
arrayWithProxy get 100000 times	28.437 ms/op	36.661 ms/op	0.78
ssz.Root.equals	1.0690 us/op	1.1490 us/op	0.93
ssz.Root.equals with valueOf()	1.2860 us/op	1.1690 us/op	1.10
byteArrayEquals with valueOf()	1.2400 us/op	1.1220 us/op	1.11
phase0 processBlock - 250000 vs - 7PWei normalcase	7.5519 ms/op	7.2172 ms/op	1.05
phase0 processBlock - 250000 vs - 7PWei worstcase	69.402 ms/op	68.358 ms/op	1.02
phase0 afterProcessEpoch - 250000 vs - 7PWei	189.64 ms/op	200.05 ms/op	0.95
phase0 beforeProcessEpoch - 250000 vs - 7PWei	528.33 ms/op	511.15 ms/op	1.03
phase0 processEpoch - mainnet_e58758	794.19 ms/op	859.91 ms/op	0.92
mainnet_e58758 - phase0 beforeProcessEpoch	418.74 ms/op	441.03 ms/op	0.95
mainnet_e58758 - phase0 processJustificationAndFinalization	76.853 us/op	115.77 us/op	0.66
mainnet_e58758 - phase0 processRewardsAndPenalties	145.71 ms/op	125.21 ms/op	1.16
mainnet_e58758 - phase0 processRegistryUpdates	35.705 us/op	68.795 us/op	0.52
mainnet_e58758 - phase0 processSlashings	1.9750 us/op	2.8130 us/op	0.70
mainnet_e58758 - phase0 processEth1DataReset	1.9500 us/op	2.3990 us/op	0.81
mainnet_e58758 - phase0 processEffectiveBalanceUpdates	4.8115 ms/op	5.8373 ms/op	0.82
mainnet_e58758 - phase0 processSlashingsReset	12.960 us/op	15.125 us/op	0.86
mainnet_e58758 - phase0 processRandaoMixesReset	11.693 us/op	23.558 us/op	0.50
mainnet_e58758 - phase0 processHistoricalRootsUpdate	1.7950 us/op	2.8540 us/op	0.63
mainnet_e58758 - phase0 processParticipationRecordUpdates	11.202 us/op	17.035 us/op	0.66
mainnet_e58758 - phase0 afterProcessEpoch	164.40 ms/op	176.26 ms/op	0.93
phase0 processEffectiveBalanceUpdates - 250000 normalcase	5.4806 ms/op	6.6660 ms/op	0.82
phase0 processEffectiveBalanceUpdates - 250000 worstcase 0.5	5.7693 ms/op	6.8130 ms/op	0.85
phase0 processRegistryUpdates - 250000 normalcase	50.211 us/op	74.282 us/op	0.68
phase0 processRegistryUpdates - 250000 badcase_full_deposits	2.5833 ms/op	2.9238 ms/op	0.88
phase0 processRegistryUpdates - 250000 worstcase 0.5	1.6100 s/op	1.6282 s/op	0.99
phase0 getAttestationDeltas - 250000 normalcase	60.868 ms/op	67.509 ms/op	0.90
phase0 getAttestationDeltas - 250000 worstcase	61.505 ms/op	68.122 ms/op	0.90
phase0 processSlashings - 250000 worstcase	30.910 ms/op	30.077 ms/op	1.03
shuffle list - 16384 els	12.968 ms/op	13.429 ms/op	0.97
shuffle list - 250000 els	185.54 ms/op	179.68 ms/op	1.03
getEffectiveBalanceIncrementsZeroInactive - 250000 vs - 7PWei	424.66 us/op	453.89 us/op	0.94
pass gossip attestations to forkchoice per slot	17.545 ms/op	14.784 ms/op	1.19
computeDeltas	2.9573 ms/op	3.3758 ms/op	0.88
computeProposerBoostScoreFromBalances	443.90 us/op	503.31 us/op	0.88
getPubkeys - index2pubkey - req 1000 vs - 250000 vc	1.8539 ms/op	1.9772 ms/op	0.94
getPubkeys - validatorsArr - req 1000 vs - 250000 vc	632.34 us/op	669.39 us/op	0.94
BLS verify - blst-native	1.6427 ms/op	1.8633 ms/op	0.88
BLS verifyMultipleSignatures 3 - blst-native	3.3743 ms/op	3.8235 ms/op	0.88
BLS verifyMultipleSignatures 8 - blst-native	7.2622 ms/op	8.2465 ms/op	0.88
BLS verifyMultipleSignatures 32 - blst-native	26.354 ms/op	29.929 ms/op	0.88
BLS aggregatePubkeys 32 - blst-native	35.401 us/op	39.909 us/op	0.89
BLS aggregatePubkeys 128 - blst-native	136.14 us/op	152.77 us/op	0.89
getAttestationsForBlock	55.893 ms/op	59.652 ms/op	0.94
CheckpointStateCache - add get delete	17.151 us/op	16.853 us/op	1.02
validate gossip signedAggregateAndProof - struct	4.4686 ms/op	4.4586 ms/op	1.00
validate gossip signedAggregateAndProof - treeBacked	4.4114 ms/op	4.4120 ms/op	1.00
validate gossip attestation - struct	2.0834 ms/op	2.0967 ms/op	0.99
validate gossip attestation - treeBacked	2.1138 ms/op	2.0987 ms/op	1.01
pickEth1Vote - no votes	8.1104 ms/op	8.0227 ms/op	1.01
pickEth1Vote - max votes	46.278 ms/op	45.283 ms/op	1.02
pickEth1Vote - Eth1Data hashTreeRoot value x2048	22.517 ms/op	24.352 ms/op	0.92
pickEth1Vote - Eth1Data hashTreeRoot tree x2048	8.7832 ms/op	8.9805 ms/op	0.98
pickEth1Vote - Eth1Data fastSerialize value x2048	4.4951 ms/op	4.8580 ms/op	0.93
pickEth1Vote - Eth1Data fastSerialize tree x2048	20.711 ms/op	21.357 ms/op	0.97
bytes32 toHexString	1.4210 us/op	1.5320 us/op	0.93
bytes32 Buffer.toString(hex)	580.00 ns/op	685.00 ns/op	0.85
bytes32 Buffer.toString(hex) from Uint8Array	727.00 ns/op	838.00 ns/op	0.87
bytes32 Buffer.toString(hex) + 0x	592.00 ns/op	681.00 ns/op	0.87
Object access 1 prop	0.27300 ns/op	0.30200 ns/op	0.90
Map access 1 prop	0.24700 ns/op	0.27800 ns/op	0.89
Object get x1000	15.864 ns/op	18.140 ns/op	0.87
Map get x1000	0.92800 ns/op	0.99800 ns/op	0.93
Object set x1000	100.02 ns/op	104.86 ns/op	0.95
Map set x1000	54.831 ns/op	60.723 ns/op	0.90
Return object 10000 times	0.32970 ns/op	0.37160 ns/op	0.89
Throw Error 10000 times	5.0600 us/op	5.8865 us/op	0.86
enrSubnets - fastDeserialize 64 bits	1.2510 us/op	1.1610 us/op	1.08
enrSubnets - ssz BitVector 64 bits	14.839 us/op	15.937 us/op	0.93
enrSubnets - fastDeserialize 4 bits	384.00 ns/op	430.00 ns/op	0.89
enrSubnets - ssz BitVector 4 bits	2.6320 us/op	2.8800 us/op	0.91
RateTracker 1000000 limit, 1 obj count per request	158.07 ns/op	171.43 ns/op	0.92
RateTracker 1000000 limit, 2 obj count per request	117.31 ns/op	126.87 ns/op	0.92
RateTracker 1000000 limit, 4 obj count per request	97.380 ns/op	104.57 ns/op	0.93
RateTracker 1000000 limit, 8 obj count per request	87.661 ns/op	93.291 ns/op	0.94
RateTracker with prune	3.4800 us/op	3.3920 us/op	1.03

by benchmarkbot/action