Halve random-access benchmark CI time: 5 s measurements, parallel and cached data generation by joseph-isaacs · Pull Request #11244 · vortex-data/vortex

joseph-isaacs · 2026-11-04T20:34:09Z

Summary

The local random-access benchmark job took ~23 min: 18.3 min of timed measurements and 4.1 min of data generation, ~80% of which was writing LitData. This PR cuts the job to roughly 12 min without changing what is measured.

Changes

5 s time limit per measurement (was the 10 s default). scripts/random-access-split.py passes --time-limit 5. Each number is a median per-lookup latency, so the time limit only sets the sample count. Even at 5 s the fast formats take thousands of samples.
Parallel data generation. --prepare-data runs one process per dataset concurrently. Each dataset's output is printed in its own log group.
Cached Parquet and LitData data. develop-bench.yml and pr-bench-runner.yml restore Parquet and LitData with actions/cache. Vortex, Arrow IPC and Lance are written every run, so the formats under test always come from the current code and lockfile. The key covers the dataset generators (vortex-bench/src/datasets/**, vortex-bench/src/random_access/**), the parquet/arrow-*/rand* versions in Cargo.lock, the LitData script, litdata.rs and the pinned litdata version. Local and S3 jobs use separate keys.

Results on this PR

All three runs use the same code; only the time limit or cache state differs.

Run	Time limit	Cache	Data generation	Benchmark step	Bench job
develop (before)	10 s	n/a	4.1 min	18.3 min	23.0 min
5 s run	5 s	miss (saved)	2.5 min	10.0 min	13.2 min
10 s control	10 s	hit	0.9 min	18.3 min	19.9 min

With 5 s and a cache hit the job should take ~11.5 min.

Do the numbers change? No. Hot geomeans vs develop:

Format	5 s	10 s control
Vortex	1.005x	1.000x
Parquet	1.013x	1.009x
Lance	1.006x	1.005x
LitData	1.011x	1.000x
Arrow IPC	1.077x	1.046x

Both runs report "No clear signal". Arrow IPC nested-lists/nested-structs read ~10–20% slower than develop at both 5 s and 10 s, so that shift is not caused by the time limit. It is either run-to-run variance between bench machines or something about PR runs; worth a separate look if it persists.

Checks: yamllint --strict on both workflows and git diff --check. The workflow changes were exercised by the benchmark runs above, covering both a cache miss and a cache hit.

🤖 Generated with Claude Code

https://claude.ai/code/session_01UeniB9mYWJs2z49XPgvbPi

…e legacy taxi runs Each of the 80 per-combination invocations measured for 10 s, and every taxi invocation repeated the legacy fixed-index taxi measurement, which the merge then discarded. Measure for 5 s and run the legacy taxi measurement only with the correlated pattern, which covers both open modes once. Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk> Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UeniB9mYWJs2z49XPgvbPi Signed-off-by: Claude <noreply@anthropic.com>

Revert the benchmark run changes from the previous commit, so measurements keep the 10 s time limit and the legacy taxi runs. Data generation now runs one process per dataset concurrently. LitData, about 80% of generation time and independent of Vortex, is generated in a separate stage and cached with actions/cache, keyed on the Parquet files it is converted from plus the conversion script and litdata version. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UeniB9mYWJs2z49XPgvbPi Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>

Only Vortex, Arrow IPC and Lance are written on every run. Parquet and LitData are restored from actions/cache, keyed on the dataset generators, the parquet/arrow/rand crate versions in Cargo.lock, and the LitData conversion script and version. Local and S3 jobs use separate keys. This replaces the separate LitData generation stage: every format is now generated in one parallel pass that skips files restored from the cache. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UeniB9mYWJs2z49XPgvbPi Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>

Experiment: halve the per-measurement time limit from 10 s to 5 s to check whether the benchmark results change. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UeniB9mYWJs2z49XPgvbPi Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>

github-actions · 2026-11-04T20:43:06Z

Polar Signals Profiling Results

Latest Run

Status	Commit	Job	Attempt	Link
🟢 Done	`11b301f`	random-access-bench	1	Explore Profiling Data

Previous Runs (1)

Status	Commit	Job	Attempt	Link
🟢 Done	`a92a117`	random-access-bench	1	Explore Profiling Data

Powered by Polar Signals Cloud

github-actions · 2026-11-04T20:55:56Z

Benchmarks: Random Access 📖

Commits: PR 11b301fc vs base 9d2da798
Verdict: No clear signal (low confidence)
Attributed Vortex impact: -0.9%
Engines: random-access No clear signal (-0.9%, low confidence)
Vortex (geomean): hot 1.000x ➖ · cold 1.051x ➖
Parquet (geomean): hot 1.009x ➖ · cold 1.046x ➖
Shifts: Parquet (control) +0.9% · Median polish +0.1%

How to read Verdict and Engines

Verdict: Overall PR-level signal after subtracting baseline drift estimated from Parquet control rows. It can be Likely improvement, Likely regression, or No clear signal.
Engines: Per-engine attribution. DataFusion is compared against DataFusion/Parquet controls; DuckDB is compared against DuckDB/Parquet controls. This answers whether each engine improved or regressed independently.
Confidence: Based on directional consistency, share of rows above the noise floor, and control-run noise.
Hot vs cold: Every measurement is run several times. The first run is reported as the cold run, and the median of the runs after it is reported as the hot run. The verdict and significance use hot runs; each target's geomean reports hot and cold beside each other where the individual runs were recorded and hot alone where they were not, the cold column shows first-run cost per row, and hot/cold is how much of each run the warm path saves. Rows whose results predate per-run reporting show only one value, taken from the value the runner reported.
Table cells: Each cell reads PR / base / %diff. The hot and cold columns mark a change as 🔴 slower or 🟢 faster once it clears this suite's threshold; anything smaller is noise here and is left unmarked, as is hot/cold, because a shift in the warm-up ratio is not a win or a loss by itself.

vortex / arrow-ipc / ns (1.046x ➖, 1↑ 7↓)

name	hot ns (PR / base / %diff)	cold ns (PR / base / %diff)	hot/cold (PR / base / %diff)
random-access/feature-vectors/correlated/arrow-ipc-cached	3153975 / 3141154 / +0.4%	3214844 / 3277667 / -1.9%	0.98 / 0.96 / +2.4%
random-access/feature-vectors/correlated/arrow-ipc-reopen	3269134 / 3205640 / +2.0%	12622452 / 9552681 / +32.1% 🔴	0.26 / 0.34 / -22.8%
random-access/feature-vectors/uniform/arrow-ipc-cached	51684094 / 55345047 / -6.6%	51373082 / 55410467 / -7.3%	1.01 / 1.00 / +0.7%
random-access/feature-vectors/uniform/arrow-ipc-reopen	34402829 / 55696350 / -38.2% 🟢	203474127 / 130458053 / +56.0% 🔴	0.17 / 0.43 / -60.4%
random-access/nested-lists/correlated/arrow-ipc-cached	65846 / 60121 / +9.5%	72284 / 57722 / +25.2% 🔴	0.91 / 1.04 / -12.5%
random-access/nested-lists/correlated/arrow-ipc-reopen	106123 / 94296 / +12.5% 🔴	3789675 / 3679155 / +3.0%	0.03 / 0.03 / +9.3%
random-access/nested-lists/uniform/arrow-ipc-cached	1217269 / 1027408 / +18.5% 🔴	1234774 / 1024864 / +20.5% 🔴	0.99 / 1.00 / -1.7%
random-access/nested-lists/uniform/arrow-ipc-reopen	1265064 / 1080478 / +17.1% 🔴	26175564 / 28891585 / -9.4%	0.05 / 0.04 / +29.2%
random-access/nested-structs/correlated/arrow-ipc-cached	53654 / 47519 / +12.9% 🔴	52385 / 48715 / +7.5%	1.02 / 0.98 / +5.0%
random-access/nested-structs/correlated/arrow-ipc-reopen	94034 / 82851 / +13.5% 🔴	3639712 / 3367051 / +8.1%	0.03 / 0.02 / +5.0%
random-access/nested-structs/uniform/arrow-ipc-cached	873708 / 776104 / +12.6% 🔴	893406 / 770824 / +15.9% 🔴	0.98 / 1.01 / -2.9%
random-access/nested-structs/uniform/arrow-ipc-reopen	926142.5 / 829577 / +11.6% 🔴	19105115 / 25889153 / -26.2% 🟢	0.05 / 0.03 / +51.3%
random-access/taxi/arrow-ipc-cached	72495 / 67412 / +7.5%	73420 / 68479 / +7.2%	0.99 / 0.98 / +0.3%
random-access/taxi/arrow-ipc-reopen	166962 / 157762 / +5.8%	3490074 / 3115786 / +12.0% 🔴	0.05 / 0.05 / -5.5%
random-access/taxi/correlated/arrow-ipc-cached	121236 / 119468 / +1.5%	135727 / 116530 / +16.5% 🔴	0.89 / 1.03 / -12.9%
random-access/taxi/correlated/arrow-ipc-reopen	222958 / 210416 / +6.0%	1565129 / 1772715 / -11.7% 🟢	0.14 / 0.12 / +20.0%
random-access/taxi/uniform/arrow-ipc-cached	2893268 / 2739994 / +5.6%	2864969 / 2729635 / +5.0%	1.01 / 1.00 / +0.6%
random-access/taxi/uniform/arrow-ipc-reopen	3059619 / 2864807 / +6.8%	37473396 / 32351016 / +15.8% 🔴	0.08 / 0.09 / -7.8%

random-access / vortex-file-compressed / ns (1.000x ➖, 0↑ 0↓)

name	hot ns (PR / base / %diff)	cold ns (PR / base / %diff)	hot/cold (PR / base / %diff)
random-access/feature-vectors/correlated/vortex-cached	110767 / 109292 / +1.3%	102828 / 103131 / -0.3%	1.08 / 1.06 / +1.6%
random-access/feature-vectors/correlated/vortex-reopen	1942736 / 1855534 / +4.7%	4204333 / 3704662 / +13.5% 🔴	0.46 / 0.50 / -7.7%
random-access/feature-vectors/uniform/vortex-cached	1064804 / 1066422 / -0.2%	1061814 / 1118666 / -5.1%	1.00 / 0.95 / +5.2%
random-access/feature-vectors/uniform/vortex-reopen	3080990 / 3045703 / +1.2%	11909703 / 11523122 / +3.4%	0.26 / 0.26 / -2.1%
random-access/nested-lists/correlated/vortex-cached	96971 / 95884 / +1.1%	104405 / 97781 / +6.8%	0.93 / 0.98 / -5.3%
random-access/nested-lists/correlated/vortex-reopen	368416 / 347105 / +6.1%	2348506 / 2075737 / +13.1% 🔴	0.16 / 0.17 / -6.2%
random-access/nested-lists/uniform/vortex-cached	479054 / 496571 / -3.5%	475623 / 493356 / -3.6%	1.01 / 1.01 / +0.1%
random-access/nested-lists/uniform/vortex-reopen	953321 / 918116.5 / +3.8%	9286204 / 8046617 / +15.4% 🔴	0.10 / 0.11 / -10.0%
random-access/nested-structs/correlated/vortex-cached	125743 / 125400 / +0.3%	127155 / 128538 / -1.1%	0.99 / 0.98 / +1.4%
random-access/nested-structs/correlated/vortex-reopen	425874 / 419909 / +1.4%	4052366 / 3404620 / +19.0% 🔴	0.11 / 0.12 / -14.8%
random-access/nested-structs/uniform/vortex-cached	291563 / 294890 / -1.1%	286301 / 292251 / -2.0%	1.02 / 1.01 / +0.9%
random-access/nested-structs/uniform/vortex-reopen	669325.5 / 687787 / -2.7%	5156191 / 4595512 / +12.2% 🔴	0.13 / 0.15 / -13.3%
random-access/taxi/correlated/vortex-cached	325955 / 334400 / -2.5%	324111 / 333060 / -2.7%	1.01 / 1.00 / +0.2%
random-access/taxi/correlated/vortex-reopen	1236161 / 1245518 / -0.8%	2608805 / 2473185 / +5.5%	0.47 / 0.50 / -5.9%
random-access/taxi/uniform/vortex-cached	1458476 / 1493556 / -2.3%	1466145 / 1485363 / -1.3%	0.99 / 1.01 / -1.1%
random-access/taxi/uniform/vortex-reopen	3658072 / 3713758 / -1.5%	7122952 / 6134172 / +16.1% 🔴	0.51 / 0.61 / -15.2%
random-access/taxi/vortex-cached	266413 / 272222 / -2.1%	283162 / 266531 / +6.2%	0.94 / 1.02 / -7.9%
random-access/taxi/vortex-reopen	1042272 / 1064199 / -2.1%	3483058 / 3428621 / +1.6%	0.30 / 0.31 / -3.6%

random-access / parquet / ns (1.009x ➖, 0↑ 0↓)

name	hot ns (PR / base / %diff)	cold ns (PR / base / %diff)	hot/cold (PR / base / %diff)
random-access/feature-vectors/correlated/parquet-cached	10240527 / 10097577 / +1.4%	10207238 / 10098602 / +1.1%	1.00 / 1.00 / +0.3%
random-access/feature-vectors/correlated/parquet-reopen	10530512 / 10416595 / +1.1%	17983440 / 17133082 / +5.0%	0.59 / 0.61 / -3.7%
random-access/feature-vectors/uniform/parquet-cached	136263148 / 131444674 / +3.7%	136554444 / 131410339 / +3.9%	1.00 / 1.00 / -0.2%
random-access/feature-vectors/uniform/parquet-reopen	134915942 / 131533008 / +2.6%	194550431 / 183409316 / +6.1%	0.69 / 0.72 / -3.3%
random-access/nested-lists/correlated/parquet-cached	4617452 / 4541310 / +1.7%	4676669 / 4562315 / +2.5%	0.99 / 1.00 / -0.8%
random-access/nested-lists/correlated/parquet-reopen	4763451 / 4607986 / +3.4%	11455431 / 10061734 / +13.9% 🔴	0.42 / 0.46 / -9.2%
random-access/nested-lists/uniform/parquet-cached	37356482 / 40126231 / -6.9%	37345773 / 40256058 / -7.2%	1.00 / 1.00 / +0.4%
random-access/nested-lists/uniform/parquet-reopen	37525123 / 37116322 / +1.1%	60957499 / 51400078 / +18.6% 🔴	0.62 / 0.72 / -14.8%
random-access/nested-structs/correlated/parquet-cached	9011816 / 9008864 / +0.0%	9021193 / 9008288 / +0.1%	1.00 / 1.00 / -0.1%
random-access/nested-structs/correlated/parquet-reopen	9292398 / 9105971 / +2.0%	22017919 / 20847595 / +5.6%	0.42 / 0.44 / -3.4%
random-access/nested-structs/uniform/parquet-cached	36667118 / 36773362 / -0.3%	36710839 / 36944371 / -0.6%	1.00 / 1.00 / +0.3%
random-access/nested-structs/uniform/parquet-reopen	36776932 / 36752645 / +0.1%	45298766 / 41666473 / +8.7%	0.81 / 0.88 / -8.0%
random-access/taxi/correlated/parquet-cached	23719659 / 23310706 / +1.8%	23650947 / 23365919 / +1.2%	1.00 / 1.00 / +0.5%
random-access/taxi/correlated/parquet-reopen	23924524 / 23704938 / +0.9%	28496318 / 25284168 / +12.7% 🔴	0.84 / 0.94 / -10.5%
random-access/taxi/parquet-cached	16883216 / 16481702 / +2.4%	16933102 / 16487371 / +2.7%	1.00 / 1.00 / -0.3%
random-access/taxi/parquet-reopen	16809261 / 16758385 / +0.3%	22878187 / 21674615 / +5.6%	0.73 / 0.77 / -5.0%
random-access/taxi/uniform/parquet-cached	37726408 / 37405826 / +0.9%	37876176 / 37222967 / +1.8%	1.00 / 1.00 / -0.9%
random-access/taxi/uniform/parquet-reopen	37798708 / 37453725 / +0.9%	40893430 / 39319729 / +4.0%	0.92 / 0.95 / -3.0%

random-access / lance / ns (1.005x ➖, 0↑ 0↓)

name	hot ns (PR / base / %diff)	cold ns (PR / base / %diff)	hot/cold (PR / base / %diff)
random-access/feature-vectors/correlated/lance-cached	150719 / 148878 / +1.2%	146990 / 148654 / -1.1%	1.03 / 1.00 / +2.4%
random-access/feature-vectors/correlated/lance-reopen	939938.5 / 929724 / +1.1%	6869400 / 6513966 / +5.5%	0.14 / 0.14 / -4.1%
random-access/feature-vectors/uniform/lance-cached	988804 / 974537 / +1.5%	991569 / 968942 / +2.3%	1.00 / 1.01 / -0.9%
random-access/feature-vectors/uniform/lance-reopen	1740885 / 1739973 / +0.1%	9452229 / 10394055 / -9.1%	0.18 / 0.17 / +10.0%
random-access/nested-lists/correlated/lance-cached	111216 / 110363 / +0.8%	117039 / 102603 / +14.1% 🔴	0.95 / 1.08 / -11.7%
random-access/nested-lists/correlated/lance-reopen	430134 / 414603.5 / +3.7%	6037830 / 5356824 / +12.7% 🔴	0.07 / 0.08 / -8.0%
random-access/nested-lists/uniform/lance-cached	809100 / 823992 / -1.8%	803857 / 811116 / -0.9%	1.01 / 1.02 / -0.9%
random-access/nested-lists/uniform/lance-reopen	1326269 / 1317005 / +0.7%	9391172 / 9919327 / -5.3%	0.14 / 0.13 / +6.4%
random-access/nested-structs/correlated/lance-cached	210735 / 208854 / +0.9%	209365 / 206631 / +1.3%	1.01 / 1.01 / -0.4%
random-access/nested-structs/correlated/lance-reopen	485982 / 471948.5 / +3.0%	5961112 / 5327380 / +11.9% 🔴	0.08 / 0.09 / -8.0%
random-access/nested-structs/uniform/lance-cached	2181479 / 2237599 / -2.5%	2145475 / 2257059 / -4.9%	1.02 / 0.99 / +2.6%
random-access/nested-structs/uniform/lance-reopen	2444974 / 2454705 / -0.4%	14983779 / 15828415 / -5.3%	0.16 / 0.16 / +5.2%
random-access/taxi/correlated/lance-cached	618507.5 / 616872 / +0.3%	615470 / 615443 / +0.0%	1.00 / 1.00 / +0.3%
random-access/taxi/correlated/lance-reopen	1604756 / 1596446 / +0.5%	2796508 / 3772993 / -25.9% 🟢	0.57 / 0.42 / +35.6%
random-access/taxi/lance-cached	424151 / 427238 / -0.7%	438789 / 421715 / +4.0%	0.97 / 1.01 / -4.6%
random-access/taxi/lance-reopen	1231900 / 1217668 / +1.2%	7975211 / 8578815 / -7.0%	0.15 / 0.14 / +8.8%
random-access/taxi/uniform/lance-cached	6749720 / 6814754 / -1.0%	6729796 / 6850314 / -1.8%	1.00 / 0.99 / +0.8%
random-access/taxi/uniform/lance-reopen	7877895 / 7797222 / +1.0%	25114237 / 28855869 / -13.0% 🟢	0.31 / 0.27 / +16.1%

random-access / litdata / ns (1.000x ➖, 0↑ 0↓)

name	hot ns (PR / base / %diff)	cold ns (PR / base / %diff)	hot/cold (PR / base / %diff)
random-access/feature-vectors/correlated/litdata-cached	726239 / 740510 / -1.9%	734834 / 742064 / -1.0%	0.99 / 1.00 / -1.0%
random-access/feature-vectors/correlated/litdata-reopen	3077867 / 3071051 / +0.2%	3780394 / 3778857 / +0.0%	0.81 / 0.81 / +0.2%
random-access/feature-vectors/uniform/litdata-cached	1121227 / 1101830 / +1.8%	1107340 / 1113495 / -0.6%	1.01 / 0.99 / +2.3%
random-access/feature-vectors/uniform/litdata-reopen	16153329 / 16005438 / +0.9%	32693432 / 31345475 / +4.3%	0.49 / 0.51 / -3.2%
random-access/nested-lists/correlated/litdata-cached	582322 / 576246 / +1.1%	583744 / 594489 / -1.8%	1.00 / 0.97 / +2.9%
random-access/nested-lists/correlated/litdata-reopen	17486713 / 17443986 / +0.2%	25262376 / 25252601 / +0.0%	0.69 / 0.69 / +0.2%
random-access/nested-lists/uniform/litdata-cached	606893 / 610551 / -0.6%	611420 / 610413 / +0.2%	0.99 / 1.00 / -0.8%
random-access/nested-lists/uniform/litdata-reopen	17146695 / 17188866 / -0.2%	25530626 / 25485895 / +0.2%	0.67 / 0.67 / -0.4%
random-access/nested-structs/correlated/litdata-cached	652678 / 652867 / -0.0%	652580 / 653203 / -0.1%	1.00 / 1.00 / +0.1%
random-access/nested-structs/correlated/litdata-reopen	18832181 / 18888657 / -0.3%	21973191 / 22031030 / -0.3%	0.86 / 0.86 / -0.0%
random-access/nested-structs/uniform/litdata-cached	711266 / 705765 / +0.8%	712211 / 711487 / +0.1%	1.00 / 0.99 / +0.7%
random-access/nested-structs/uniform/litdata-reopen	16234588 / 16448113 / -1.3%	24760859 / 24650735 / +0.4%	0.66 / 0.67 / -1.7%
random-access/taxi/correlated/litdata-cached	1142344 / 1148818 / -0.6%	1143117 / 1146438 / -0.3%	1.00 / 1.00 / -0.3%
random-access/taxi/correlated/litdata-reopen	19991708 / 19977238 / +0.1%	29543954 / 29408538 / +0.5%	0.68 / 0.68 / -0.4%
random-access/taxi/litdata-cached	79960 / 78598 / +1.7%	84058 / 80795 / +4.0%	0.95 / 0.97 / -2.2%
random-access/taxi/litdata-reopen	10620898 / 10546015 / +0.7%	14394735 / 14781539 / -2.6%	0.74 / 0.71 / +3.4%
random-access/taxi/uniform/litdata-cached	1291892 / 1291445 / +0.0%	1299416 / 1300325 / -0.1%	0.99 / 0.99 / +0.1%
random-access/taxi/uniform/litdata-reopen	55014687 / 56197474 / -2.1%	83349852 / 83950345 / -0.7%	0.66 / 0.67 / -1.4%

Temporarily undo the 5 s experiment so the PR benchmark measures this branch at 10 s, to separate run-to-run noise from the effect of the time limit. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UeniB9mYWJs2z49XPgvbPi Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>

A 10 s control run on this branch moved by the same amount against develop as the 5 s run, so halving the time limit does not change the results; it only halves the benchmark step. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UeniB9mYWJs2z49XPgvbPi Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>

joseph-isaacs · 2026-11-04T22:16:49Z

CI failures on this PR so far, none caused by its diff, which only touches the random-access benchmark script and two benchmark workflows:

Python (wheel build) and check-java-publish-build in Publish Dry Runs failed in the mlugg/setup-zig step with ##[error]fetch failed, while downloading Zig and before anything was built. Both passed on re-run.
Python CUDA (test) in run 41923639578 never started: the runner could not launch because AWS reported InsufficientInstanceCapacity for g6/g5 GPU instances in the us-east availability zones it tried. Failed job re-run once.

Generated by Claude Code

joseph-isaacs and others added 4 commits November 4, 2026 19:18

joseph-isaacs added the action/bench-random-access Run only the random-access benchmark on this PR label Nov 4, 2026 — with Claude

github-actions Bot removed the action/bench-random-access Run only the random-access benchmark on this PR label Nov 4, 2026

joseph-isaacs added the action/bench-random-access Run only the random-access benchmark on this PR label Nov 4, 2026 — with Claude

github-actions Bot removed the action/bench-random-access Run only the random-access benchmark on this PR label Nov 4, 2026

joseph-isaacs changed the title ~~Speed up random-access benchmark data generation (and trial a 5 s time limit)~~ Halve random-access benchmark CI time: 5 s measurements, parallel and cached data generation Nov 4, 2026

joseph-isaacs marked this pull request as ready for review November 4, 2026 22:15

joseph-isaacs requested review from connortsui20 and robert3005 November 4, 2026 22:15

joseph-isaacs added the changelog/ci label Nov 4, 2026 — with Claude

robert3005 approved these changes Nov 4, 2026

View reviewed changes

robert3005 merged commit 6e2f427 into develop Nov 5, 2026
179 of 183 checks passed

robert3005 deleted the ji/cool-hypatia-lsuwfm branch November 5, 2026 00:11

Navigation Menu

Repository navigation

Halve random-access benchmark CI time: 5 s measurements, parallel and cached data generation - #11244

Halve random-access benchmark CI time: 5 s measurements, parallel and cached data generation#11244
robert3005 merged 6 commits into
developfrom
ji/cool-hypatia-lsuwfm

joseph-isaacs commented Nov 4, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Nov 4, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Nov 4, 2026 •

edited

Loading

Uh oh!

joseph-isaacs commented Nov 4, 2026 •

edited

Loading

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

joseph-isaacs commented Nov 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Changes

Results on this PR

Uh oh!

github-actions Bot commented Nov 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Polar Signals Profiling Results

Latest Run

Uh oh!

github-actions Bot commented Nov 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Benchmarks: Random Access 📖

Uh oh!

joseph-isaacs commented Nov 4, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

joseph-isaacs commented Nov 4, 2026 •

edited

Loading

github-actions Bot commented Nov 4, 2026 •

edited

Loading

github-actions Bot commented Nov 4, 2026 •

edited

Loading

joseph-isaacs commented Nov 4, 2026 •

edited

Loading