Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

move maybe_evict_cache() off default rayon pool #15068

Merged
merged 1 commit into from
Oct 24, 2024

Conversation

msmouse
Copy link
Contributor

@msmouse msmouse commented Oct 23, 2024

Description

This can potentially block the pre_commit_ledger thread while it's holding the buffered_state lock. Before we reduce the scope of that lock, I'm moving this part off the defualt pool so that even if the default pool gets busy (or worse, there are tasks on the threads of the pool trying to lock the buffered_state) this part doesn't get affected.

How Has This Been Tested?

Key Areas to Review

Type of Change

  • New feature

Which Components or Systems Does This Change Impact?

  • Validator Node

Copy link

trunk-io bot commented Oct 23, 2024

⏱️ 2h 31m total CI duration on this PR
Slowest 15 Jobs Cumulative Duration Recent Runs
rust-smoke-tests 1h 27m 🟥🟥🟩
execution-performance / single-node-performance 21m 🟩
test-target-determinator 8m 🟩🟩
rust-cargo-deny 5m 🟩🟩🟩
check-dynamic-deps 5m 🟩🟩🟩🟩
rust-doc-tests 5m 🟩
execution-performance / test-target-determinator 4m 🟩
check 4m 🟩
rust-move-tests 2m 🟩
rust-move-tests 2m 🟩
semgrep/ci 2m 🟩🟩🟩🟩
rust-move-tests 2m 🟩
fetch-last-released-docker-image-tag 1m 🟩
general-lints 1m 🟩🟩🟩
file_change_determinator 32s 🟩🟩🟩

settingsfeedbackdocs ⋅ learn more about trunk.io

@msmouse msmouse added the CICD:build-images when this label is present github actions will start build+push rust images from the PR. label Oct 23, 2024
This can potentially block the pre_commit_ledger thread while it's
holding the buffered_state lock. Before we reduce the scope of that
lock, I'm moving this part off the defualt pool so that even if the
default pool gets busy (or worse, there are tasks on the threads of the
pool trying to lock the buffered_state) this part doesn't get affected.
@msmouse msmouse force-pushed the 1023-alden-evict-cache-off-default-pool branch from 3b5cd91 to a29d236 Compare October 23, 2024 23:36
@msmouse msmouse enabled auto-merge (squash) October 23, 2024 23:36

This comment has been minimized.

This comment has been minimized.

This comment has been minimized.

Copy link
Contributor

✅ Forge suite realistic_env_max_load success on a29d23639745500ecebd263ade708544100c962b

two traffics test: inner traffic : committed: 13917.57 txn/s, latency: 2853.20 ms, (p50: 2700 ms, p70: 2700, p90: 3000 ms, p99: 3300 ms), latency samples: 5291780
two traffics test : committed: 100.00 txn/s, latency: 1495.74 ms, (p50: 1500 ms, p70: 1500, p90: 1600 ms, p99: 1800 ms), latency samples: 1800
Latency breakdown for phase 0: ["MempoolToBlockCreation: max: 2.060, avg: 1.619", "ConsensusProposalToOrdered: max: 0.341, avg: 0.303", "ConsensusOrderedToCommit: max: 0.382, avg: 0.370", "ConsensusProposalToCommit: max: 0.681, avg: 0.673"]
Max non-epoch-change gap was: 0 rounds at version 0 (avg 0.00) [limit 4], 1.10s no progress at version 2071742 (avg 0.21s) [limit 15].
Max epoch-change gap was: 0 rounds at version 0 (avg 0.00) [limit 4], 8.20s no progress at version 2071740 (avg 6.89s) [limit 15].
Test Ok

Copy link
Contributor

✅ Forge suite framework_upgrade success on b29f09f57e898d8d211c8bc3e303f6e50bba2266 ==> a29d23639745500ecebd263ade708544100c962b

Compatibility test results for b29f09f57e898d8d211c8bc3e303f6e50bba2266 ==> a29d23639745500ecebd263ade708544100c962b (PR)
Upgrade the nodes to version: a29d23639745500ecebd263ade708544100c962b
framework_upgrade::framework-upgrade::full-framework-upgrade : committed: 1077.62 txn/s, submitted: 1080.15 txn/s, failed submission: 2.53 txn/s, expired: 2.53 txn/s, latency: 2873.92 ms, (p50: 2600 ms, p70: 2900, p90: 4500 ms, p99: 5900 ms), latency samples: 93800
framework_upgrade::framework-upgrade::full-framework-upgrade : committed: 1131.73 txn/s, submitted: 1134.17 txn/s, failed submission: 2.44 txn/s, expired: 2.44 txn/s, latency: 2660.22 ms, (p50: 2500 ms, p70: 3000, p90: 3900 ms, p99: 5400 ms), latency samples: 101920
5. check swarm health
Compatibility test for b29f09f57e898d8d211c8bc3e303f6e50bba2266 ==> a29d23639745500ecebd263ade708544100c962b passed
Upgrade the remaining nodes to version: a29d23639745500ecebd263ade708544100c962b
framework_upgrade::framework-upgrade::full-framework-upgrade : committed: 1182.88 txn/s, submitted: 1184.86 txn/s, failed submission: 1.98 txn/s, expired: 1.98 txn/s, latency: 2512.53 ms, (p50: 2400 ms, p70: 2700, p90: 3900 ms, p99: 6300 ms), latency samples: 107360
Test Ok

Copy link
Contributor

✅ Forge suite compat success on b29f09f57e898d8d211c8bc3e303f6e50bba2266 ==> a29d23639745500ecebd263ade708544100c962b

Compatibility test results for b29f09f57e898d8d211c8bc3e303f6e50bba2266 ==> a29d23639745500ecebd263ade708544100c962b (PR)
1. Check liveness of validators at old version: b29f09f57e898d8d211c8bc3e303f6e50bba2266
compatibility::simple-validator-upgrade::liveness-check : committed: 17199.17 txn/s, latency: 1973.88 ms, (p50: 2000 ms, p70: 2100, p90: 2200 ms, p99: 2400 ms), latency samples: 553120
2. Upgrading first Validator to new version: a29d23639745500ecebd263ade708544100c962b
compatibility::simple-validator-upgrade::single-validator-upgrading : committed: 5720.50 txn/s, latency: 5007.97 ms, (p50: 5300 ms, p70: 5700, p90: 6500 ms, p99: 6800 ms), latency samples: 114280
compatibility::simple-validator-upgrade::single-validator-upgrade : committed: 5859.37 txn/s, latency: 5533.77 ms, (p50: 5900 ms, p70: 6100, p90: 7300 ms, p99: 7500 ms), latency samples: 202560
3. Upgrading rest of first batch to new version: a29d23639745500ecebd263ade708544100c962b
compatibility::simple-validator-upgrade::half-validator-upgrading : committed: 5885.90 txn/s, latency: 4865.80 ms, (p50: 5600 ms, p70: 5800, p90: 6000 ms, p99: 6100 ms), latency samples: 112040
compatibility::simple-validator-upgrade::half-validator-upgrade : committed: 5512.13 txn/s, latency: 5828.73 ms, (p50: 6000 ms, p70: 6100, p90: 7600 ms, p99: 7900 ms), latency samples: 191660
4. upgrading second batch to new version: a29d23639745500ecebd263ade708544100c962b
compatibility::simple-validator-upgrade::rest-validator-upgrading : committed: 8246.54 txn/s, latency: 3433.18 ms, (p50: 3800 ms, p70: 4000, p90: 4200 ms, p99: 4300 ms), latency samples: 153060
compatibility::simple-validator-upgrade::rest-validator-upgrade : committed: 8512.81 txn/s, latency: 3757.82 ms, (p50: 3900 ms, p70: 4100, p90: 4400 ms, p99: 4600 ms), latency samples: 282800
5. check swarm health
Compatibility test for b29f09f57e898d8d211c8bc3e303f6e50bba2266 ==> a29d23639745500ecebd263ade708544100c962b passed
Test Ok

@msmouse msmouse merged commit c070f4c into main Oct 24, 2024
62 of 92 checks passed
@msmouse msmouse deleted the 1023-alden-evict-cache-off-default-pool branch October 24, 2024 18:48
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CICD:build-images when this label is present github actions will start build+push rust images from the PR.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants