Repository navigation
CI: build benchmarks binaries once for all targets - #9121
Conversation
Signed-off-by: Mikhail Kot <mikhail@spiraldb.com>
| build: | ||
| timeout-minutes: 60 | ||
| env: | ||
| VORTEX_EXPERIMENTAL_PATCHED_ARRAY: "1" |
There was a problem hiding this comment.
I think these are runtime only, should be OK to remove here
AdamGS
left a comment
There was a problem hiding this comment.
Overall LGTM, one small comment
Merging this PR will improve performance by 13.11%
Performance Changes
Tip Curious why this is faster? Comment Comparing Footnotes
|
Polar Signals Profiling ResultsLatest Run
Powered by Polar Signals Cloud |
Benchmarks: FineWeb NVMe 📖Verdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed / ns (0.973x ➖, 0↑ 0↓)
datafusion / parquet / ns (0.984x ➖, 0↑ 0↓)
duckdb / vortex-file-compressed / ns (0.921x ➖, 3↑ 0↓)
duckdb / parquet / ns (1.007x ➖, 0↑ 0↓)
File Size Changes (2 files changed, -46.3% overall, 0↑ 2↓)
Totals:
|
Benchmarks: TPC-H SF=1 on NVME 📖Verdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed / ns (1.000x ➖, 0↑ 0↓)
datafusion / parquet / ns (1.013x ➖, 1↑ 1↓)
duckdb / vortex-file-compressed / ns (1.014x ➖, 0↑ 2↓)
duckdb / parquet / ns (0.999x ➖, 0↑ 0↓)
File Size Changes (9 files changed, -43.9% overall, 0↑ 9↓)
Totals:
|
|
this works 🥳 |
Benchmarks: Clickbench Sorted on NVME 📖Verdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed / ns (1.033x ➖, 0↑ 2↓)
datafusion / parquet / ns (0.987x ➖, 1↑ 1↓)
duckdb / vortex-file-compressed / ns (0.995x ➖, 0↑ 0↓)
duckdb / parquet / ns (0.980x ➖, 0↑ 0↓)
File Size Changes (201 files changed, -42.7% overall, 53↑ 148↓)
Totals:
|
Benchmarks: TPC-H SF=10 on NVME 📖Verdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed / ns (0.990x ➖, 0↑ 0↓)
datafusion / parquet / ns (0.994x ➖, 0↑ 0↓)
duckdb / vortex-file-compressed / ns (1.009x ➖, 0↑ 0↓)
duckdb / parquet / ns (1.005x ➖, 0↑ 0↓)
File Size Changes (9 files changed, -44.0% overall, 0↑ 9↓)
Totals:
|
Benchmarks: Statistical and Population Genetics 📖Verdict: No clear signal (low confidence) How to read Verdict and Engines
duckdb / vortex-file-compressed / ns (1.058x ➖, 1↑ 3↓)
duckdb / parquet / ns (1.015x ➖, 0↑ 0↓)
File Size Changes (2 files changed, -32.3% overall, 0↑ 2↓)
Totals:
|
Benchmarks: PolarSignals Profiling 📖Vortex (geomean): 0.988x ➖ datafusion / vortex-file-compressed / ns (0.988x ➖, 0↑ 0↓)
No file size changes detected. |
Benchmarks: FineWeb S3 📖Verdict: No clear signal (environment too noisy confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed / ns (1.294x ➖, 0↑ 3↓)
datafusion / parquet / ns (0.909x ➖, 1↑ 0↓)
duckdb / vortex-file-compressed / ns (1.008x ➖, 0↑ 0↓)
duckdb / parquet / ns (1.590x ❌, 1↑ 3↓)
|
Benchmarks: TPC-DS SF=1 on NVME 📖Verdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed / ns (1.022x ➖, 2↑ 2↓)
datafusion / parquet / ns (1.005x ➖, 0↑ 3↓)
duckdb / vortex-file-compressed / ns (0.981x ➖, 10↑ 4↓)
duckdb / parquet / ns (1.014x ➖, 3↑ 10↓)
File Size Changes (25 files changed, -43.5% overall, 0↑ 25↓)
Totals:
|
Benchmarks: Clickbench on NVME 📖Verdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed / ns (1.000x ➖, 2↑ 0↓)
datafusion / parquet / ns (1.001x ➖, 0↑ 1↓)
duckdb / vortex-file-compressed / ns (1.011x ➖, 3↑ 6↓)
duckdb / parquet / ns (0.994x ➖, 0↑ 2↓)
File Size Changes (101 files changed, -39.2% overall, 0↑ 101↓)
Totals:
|
Benchmarks: TPC-H SF=1 on S3 📖Verdict: No clear signal (low confidence) How to read Verdict and Engines
datafusion / vortex-file-compressed / ns (1.156x ➖, 0↑ 3↓)
datafusion / parquet / ns (1.000x ➖, 0↑ 0↓)
duckdb / vortex-file-compressed / ns (1.054x ➖, 0↑ 1↓)
duckdb / parquet / ns (0.992x ➖, 0↑ 0↓)
|
We've migrated to large metal runners for our benchmarks, and now build time and debuginfo upload time dominates download time, so it makes sense to avoid re-building.
This also is beneficial because large runners have lower availability, so we want to use each for less time.
Continuation of #8010