Skip to content

Conversation

@emwang-jump
Copy link
Contributor

No description provided.

@emwang-jump emwang-jump requested a review from lidatong as a code owner October 8, 2025 17:33
@github-actions
Copy link

github-actions bot commented Oct 8, 2025

Performance Measurements ⏳

Suite Baseline New Change
backtest mainnet-368528500-perf per slot 0.097117 s 0.097648 s 0.547%
backtest mainnet-368528500-perf snapshot load 3 s 2.9 s -3.333%
backtest mainnet-368528500-perf total elapsed 97.117216 s 97.648318 s 0.547%
firedancer mem usage with mainnet.toml 1008.12 GiB 1008.12 GiB 0.000%

@emwang-jump emwang-jump force-pushed the emwang/repair-peer-select branch 2 times, most recently from 5a2017c to a6342bc Compare October 13, 2025 16:45
@github-actions
Copy link

Performance Measurements ⏳

Suite Baseline New Change
backtest mainnet-368528500-perf per slot 0.081662 s 0.082372 s 0.869%
backtest mainnet-368528500-perf snapshot load 2.965 s 2.897 s -2.293%
backtest mainnet-368528500-perf total elapsed 81.662023 s 82.371805 s 0.869%
firedancer mem usage with mainnet.toml 1014.12 GiB 1014.12 GiB 0.000%

@emwang-jump emwang-jump force-pushed the emwang/repair-peer-select branch from a6342bc to 1a3fe76 Compare October 13, 2025 16:52
@emwang-jump emwang-jump changed the title repair: coarse peer selection repair: coarse peer selection by latency Oct 13, 2025
@github-actions
Copy link

Performance Measurements ⏳

Suite Baseline New Change
backtest mainnet-368528500-perf per slot 0.081499 s 0.081942 s 0.544%
backtest mainnet-368528500-perf snapshot load 2.91 s 2.902 s -0.275%
backtest mainnet-368528500-perf total elapsed 81.498802 s 81.941686 s 0.543%
firedancer mem usage with mainnet.toml 1014.12 GiB 1014.12 GiB 0.000%

@emwang-jump emwang-jump force-pushed the emwang/repair-peer-select branch from 1a3fe76 to 5fdbc8d Compare October 13, 2025 17:04
@github-actions
Copy link

Performance Measurements ⏳

Suite Baseline New Change
backtest mainnet-368528500-perf per slot 0.081934 s 0.082189 s 0.311%
backtest mainnet-368528500-perf snapshot load 2.928 s 2.912 s -0.546%
backtest mainnet-368528500-perf total elapsed 81.933648 s 82.188535 s 0.311%
firedancer mem usage with mainnet.toml 1014.12 GiB 1014.12 GiB 0.000%

lidatong
lidatong previously approved these changes Oct 14, 2025
Comment on lines 167 to 168
FD_POLICY_LATENCY_WORST, /* do a cycle through worst peers 1/7 times to see if any improvements are made */
FD_POLICY_LATENCY_BEST,
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

fast slow?

Comment on lines 178 to 184
fd_policy_dedup_t dedup; /* dedup cache of already sent requests */
fd_policy_peers_t peers; /* repair peers (strategy & data) */
long tsmax; /* maximum time for an iteration before resetting the DFS to root */
long tsref; /* reference timestamp for resetting DFS */

fd_forest_iter_t iterf; /* forest iterator */
ulong tsreset; /* ms timestamp of last reset of iterf */
fd_forest_iter_t iterf; /* forest iterator */
ulong tsreset; /* ms timestamp of last reset of iterf */
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

seems like extra space?

@emwang-jump emwang-jump added this pull request to the merge queue Oct 14, 2025
@github-actions
Copy link

Performance Measurements ⏳

Suite Baseline New Change
backtest mainnet-368528500-perf per slot 0.081597 s 0.081701 s 0.127%
backtest mainnet-368528500-perf snapshot load 2.938 s 2.949 s 0.374%
backtest mainnet-368528500-perf total elapsed 81.596541 s 81.700965 s 0.128%
firedancer mem usage with mainnet.toml 1014.12 GiB 1014.12 GiB 0.000%

Merged via the queue into main with commit 07c27f8 Oct 14, 2025
12 checks passed
@emwang-jump emwang-jump deleted the emwang/repair-peer-select branch October 14, 2025 19:50
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants