You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@datafusion.apache.org by "alamb (via GitHub)" <gi...@apache.org> on 2026/09/28 14:36:56 UTC
[PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
alamb opened a new pull request, #25832:
URL: https://github.com/apache/datafusion/pull/25832
## Which issue does this PR close?
- Related to https://github.com/apache/arrow-rs/pull/10859
- Related to https://github.com/apache/arrow-rs/issues/10926
**DO NOT MERGE**: this PR exists only to run the DataFusion tests and benchmarks against a pre-release arrow-rs branch.
## Rationale for this change
https://github.com/apache/arrow-rs/pull/10859 changes the parquet push decoder to fuse consecutive `RowFilter` predicates that use the same projection, so a column is decoded once and later predicates see only the surviving rows. I want to (1) confirm the existing DataFusion parquet filter pushdown tests pass against it and (2) measure its effect on ClickBench with `pushdown_filters` enabled, using `datafusion-cli` and the benchmark runner.
## What changes are included in this PR?
Based on `main` (arrow/parquet 60.0.0), plus a `[patch.crates-io]` section pinning all arrow/parquet crates to https://github.com/alamb/arrow-rs/tree/pr-10859-fused-predicates, which is the head of apache/arrow-rs#10859 (commit `257753e2d32f6f398dc6db14f8195b3447dd4abe`) pushed to my fork so the pin stays stable if the PR branch is force pushed.
## What is the testing strategy for this PR?
Tests and benchmarks only:
- The existing parquet filter pushdown tests in `datafusion/core/tests/parquet/filter_pushdown.rs`
- `clickbench_pushdown` via the benchmark bot, and local `datafusion-cli` runs with `datafusion.execution.parquet.pushdown_filters = true`
## Are there any user-facing changes?
No.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872307502
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25832#issuecomment-5872245663)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5872245663-2881-xpsfn 6.12.94+ #1 SMP Wed Aug 19 07:47:20 UTC 2026 aarch64 GNU/Linux`
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
Comparing test-arrow-pr-10859 (702bffb69332edf26d82a6d4d6fff3ce8c825343) to 991fd23 (merge-base) [diff](https://github.com/apache/datafusion/compare/991fd23dd0be0046af5945b8c3905612859b657a..702bffb69332edf26d82a6d4d6fff3ce8c825343)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: "true"
```
</details>
Results will be posted here when complete
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872840275
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25832#issuecomment-5872602693)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing test-arrow-pr-10859 (702bffb69332edf26d82a6d4d6fff3ce8c825343) to 991fd23 (merge-base) [diff](https://github.com/apache/datafusion/compare/991fd23dd0be0046af5945b8c3905612859b657a..702bffb69332edf26d82a6d4d6fff3ce8c825343)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: "true"
```
</details>
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and test-arrow-pr-10859
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β test-arrow-pr-10859 β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.23 ms β 1.23 ms β no change β
β QQuery 1 β 12.22 ms β 12.33 ms β no change β
β QQuery 2 β 36.86 ms β 36.65 ms β no change β
β QQuery 3 β 31.61 ms β 31.54 ms β no change β
β QQuery 4 β 237.15 ms β 236.03 ms β no change β
β QQuery 5 β 271.52 ms β 274.71 ms β no change β
β QQuery 6 β 1.29 ms β 1.31 ms β no change β
β QQuery 7 β 15.96 ms β 15.84 ms β no change β
β QQuery 8 β 327.21 ms β 333.04 ms β no change β
β QQuery 9 β 468.31 ms β 468.36 ms β no change β
β QQuery 10 β 88.32 ms β 88.93 ms β no change β
β QQuery 11 β 100.14 ms β 99.62 ms β no change β
β QQuery 12 β 289.06 ms β 290.72 ms β no change β
β QQuery 13 β 394.73 ms β 392.24 ms β no change β
β QQuery 14 β 308.29 ms β 305.64 ms β no change β
β QQuery 15 β 285.39 ms β 284.86 ms β no change β
β QQuery 16 β 619.39 ms β 622.32 ms β no change β
β QQuery 17 β 621.84 ms β 622.11 ms β no change β
β QQuery 18 β 1260.53 ms β 1299.01 ms β no change β
β QQuery 19 β 29.03 ms β 29.59 ms β no change β
β QQuery 20 β 519.80 ms β 513.86 ms β no change β
β QQuery 21 β 544.86 ms β 544.01 ms β no change β
β QQuery 22 β 787.69 ms β 765.13 ms β no change β
β QQuery 23 β 133.78 ms β 130.77 ms β no change β
β QQuery 24 β 44.61 ms β 41.54 ms β +1.07x faster β
β QQuery 25 β 134.59 ms β 106.51 ms β +1.26x faster β
β QQuery 26 β 47.83 ms β 45.19 ms β +1.06x faster β
β QQuery 27 β 540.25 ms β 539.98 ms β no change β
β QQuery 28 β 2912.45 ms β 2894.96 ms β no change β
β QQuery 29 β 41.64 ms β 41.89 ms β no change β
β QQuery 30 β 310.33 ms β 308.08 ms β no change β
β QQuery 31 β 280.87 ms β 275.30 ms β no change β
β QQuery 32 β 927.59 ms β 966.18 ms β no change β
β QQuery 33 β 1428.83 ms β 1427.07 ms β no change β
β QQuery 34 β 1474.42 ms β 1453.06 ms β no change β
β QQuery 35 β 295.22 ms β 297.24 ms β no change β
β QQuery 36 β 66.20 ms β 69.14 ms β no change β
β QQuery 37 β 35.78 ms β 35.76 ms β no change β
β QQuery 38 β 36.13 ms β 35.68 ms β no change β
β QQuery 39 β 136.58 ms β 131.34 ms β no change β
β QQuery 40 β 17.96 ms β 17.43 ms β no change β
β QQuery 41 β 16.59 ms β 16.31 ms β no change β
β QQuery 42 β 14.14 ms β 14.01 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 16148.22ms β
β Total Time (test-arrow-pr-10859) β 16116.52ms β
β Average Time (HEAD) β 375.54ms β
β Average Time (test-arrow-pr-10859) β 374.80ms β
β Queries Faster β 3 β
β Queries Slower β 0 β
β Queries with No Change β 40 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and test-arrow-pr-10859
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β test-arrow-pr-10859 β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.23 / 4.03 Β±5.48 / 14.99 ms β 1.23 / 4.04 Β±5.51 / 15.06 ms β no change β
β QQuery 1 β 12.22 / 12.39 Β±0.14 / 12.58 ms β 12.33 / 12.51 Β±0.16 / 12.75 ms β no change β
β QQuery 2 β 36.86 / 37.19 Β±0.23 / 37.46 ms β 36.65 / 37.17 Β±0.27 / 37.39 ms β no change β
β QQuery 3 β 31.61 / 32.41 Β±0.74 / 33.74 ms β 31.54 / 31.80 Β±0.20 / 32.12 ms β no change β
β QQuery 4 β 237.15 / 240.03 Β±3.38 / 246.30 ms β 236.03 / 238.89 Β±1.98 / 241.98 ms β no change β
β QQuery 5 β 271.52 / 275.35 Β±2.67 / 279.09 ms β 274.71 / 277.11 Β±2.85 / 282.55 ms β no change β
β QQuery 6 β 1.29 / 1.45 Β±0.24 / 1.93 ms β 1.31 / 1.46 Β±0.24 / 1.94 ms β no change β
β QQuery 7 β 15.96 / 16.17 Β±0.12 / 16.32 ms β 15.84 / 16.00 Β±0.16 / 16.28 ms β no change β
β QQuery 8 β 327.21 / 335.25 Β±4.49 / 341.09 ms β 333.04 / 344.33 Β±6.62 / 350.52 ms β no change β
β QQuery 9 β 468.31 / 472.77 Β±2.63 / 476.15 ms β 468.36 / 478.30 Β±5.17 / 482.20 ms β no change β
β QQuery 10 β 88.32 / 89.22 Β±0.54 / 89.91 ms β 88.93 / 90.22 Β±1.13 / 92.30 ms β no change β
β QQuery 11 β 100.14 / 100.71 Β±0.86 / 102.41 ms β 99.62 / 100.92 Β±0.86 / 102.33 ms β no change β
β QQuery 12 β 289.06 / 295.89 Β±4.05 / 301.49 ms β 290.72 / 298.64 Β±8.64 / 315.34 ms β no change β
β QQuery 13 β 394.73 / 407.51 Β±10.43 / 423.66 ms β 392.24 / 410.11 Β±12.59 / 424.42 ms β no change β
β QQuery 14 β 308.29 / 312.62 Β±3.94 / 319.51 ms β 305.64 / 309.50 Β±3.26 / 314.73 ms β no change β
β QQuery 15 β 285.39 / 292.00 Β±9.83 / 311.50 ms β 284.86 / 294.25 Β±9.31 / 310.88 ms β no change β
β QQuery 16 β 619.39 / 626.88 Β±8.43 / 641.12 ms β 622.32 / 635.74 Β±12.60 / 657.46 ms β no change β
β QQuery 17 β 621.84 / 626.21 Β±5.19 / 634.98 ms β 622.11 / 639.37 Β±11.82 / 658.41 ms β no change β
β QQuery 18 β 1260.53 / 1276.67 Β±17.21 / 1302.32 ms β 1299.01 / 1331.25 Β±22.69 / 1369.77 ms β no change β
β QQuery 19 β 29.03 / 29.83 Β±0.50 / 30.46 ms β 29.59 / 29.81 Β±0.19 / 30.12 ms β no change β
β QQuery 20 β 519.80 / 528.28 Β±9.05 / 545.36 ms β 513.86 / 528.52 Β±12.65 / 549.74 ms β no change β
β QQuery 21 β 544.86 / 564.75 Β±13.48 / 584.11 ms β 544.01 / 556.45 Β±9.53 / 568.18 ms β no change β
β QQuery 22 β 787.69 / 791.62 Β±3.83 / 797.64 ms β 765.13 / 774.95 Β±6.19 / 782.69 ms β no change β
β QQuery 23 β 133.78 / 141.47 Β±6.26 / 151.31 ms β 130.77 / 138.23 Β±8.28 / 153.88 ms β no change β
β QQuery 24 β 44.61 / 47.05 Β±3.91 / 54.81 ms β 41.54 / 41.86 Β±0.40 / 42.65 ms β +1.12x faster β
β QQuery 25 β 134.59 / 138.50 Β±5.52 / 149.36 ms β 106.51 / 112.38 Β±5.35 / 119.47 ms β +1.23x faster β
β QQuery 26 β 47.83 / 48.36 Β±0.49 / 48.96 ms β 45.19 / 45.89 Β±0.46 / 46.36 ms β +1.05x faster β
β QQuery 27 β 540.25 / 554.78 Β±9.56 / 568.35 ms β 539.98 / 556.41 Β±12.84 / 571.56 ms β no change β
β QQuery 28 β 2912.45 / 2951.55 Β±26.57 / 2995.21 ms β 2894.96 / 2926.95 Β±23.89 / 2950.20 ms β no change β
β QQuery 29 β 41.64 / 50.96 Β±14.98 / 80.73 ms β 41.89 / 45.42 Β±5.97 / 57.33 ms β +1.12x faster β
β QQuery 30 β 310.33 / 321.51 Β±8.81 / 333.45 ms β 308.08 / 316.58 Β±7.52 / 327.95 ms β no change β
β QQuery 31 β 280.87 / 291.26 Β±5.95 / 298.78 ms β 275.30 / 281.21 Β±4.42 / 287.55 ms β no change β
β QQuery 32 β 927.59 / 964.88 Β±33.66 / 1028.39 ms β 966.18 / 986.54 Β±20.57 / 1020.00 ms β no change β
β QQuery 33 β 1428.83 / 1466.46 Β±19.75 / 1483.92 ms β 1427.07 / 1477.34 Β±32.19 / 1520.42 ms β no change β
β QQuery 34 β 1474.42 / 1498.94 Β±15.31 / 1515.27 ms β 1453.06 / 1488.53 Β±21.05 / 1519.23 ms β no change β
β QQuery 35 β 295.22 / 328.99 Β±39.77 / 401.46 ms β 297.24 / 355.77 Β±80.99 / 504.68 ms β 1.08x slower β
β QQuery 36 β 66.20 / 67.93 Β±1.50 / 70.26 ms β 69.14 / 73.72 Β±2.56 / 76.72 ms β 1.09x slower β
β QQuery 37 β 35.78 / 39.94 Β±3.23 / 43.44 ms β 35.76 / 38.17 Β±3.08 / 43.87 ms β no change β
β QQuery 38 β 36.13 / 41.26 Β±4.95 / 50.51 ms β 35.68 / 38.80 Β±2.92 / 42.29 ms β +1.06x faster β
β QQuery 39 β 136.58 / 144.92 Β±5.93 / 153.98 ms β 131.34 / 140.66 Β±5.90 / 149.17 ms β no change β
β QQuery 40 β 17.96 / 20.55 Β±3.72 / 27.73 ms β 17.43 / 17.75 Β±0.27 / 18.20 ms β +1.16x faster β
β QQuery 41 β 16.59 / 17.91 Β±1.54 / 20.88 ms β 16.31 / 17.50 Β±2.00 / 21.50 ms β no change β
β QQuery 42 β 14.14 / 14.39 Β±0.19 / 14.62 ms β 14.01 / 14.25 Β±0.13 / 14.36 ms β no change β
βββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 16520.83ms β
β Total Time (test-arrow-pr-10859) β 16555.33ms β
β Average Time (HEAD) β 384.21ms β
β Average Time (test-arrow-pr-10859) β 385.01ms β
β Queries Faster β 6 β
β Queries Slower β 2 β
β Queries with No Change β 35 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 85.0s |
| Peak memory | 12.1 GiB |
| Avg memory | 4.9 GiB |
| CPU user | 842.4s |
| CPU sys | 56.7s |
| Peak spill | 0 B |
**clickbench_partitioned β branch**
| Metric | Value |
|--------|-------|
| Wall time | 85.0s |
| Peak memory | 10.0 GiB |
| Avg memory | 4.6 GiB |
| CPU user | 841.8s |
| CPU sys | 58.8s |
| Peak spill | 0 B |
</details>
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "alamb (via GitHub)" <gi...@apache.org>.
alamb commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872287098
run benchmark clickbench_pushdown
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "alamb (via GitHub)" <gi...@apache.org>.
alamb commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872245143
run clickbench_pushdown
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "alamb (via GitHub)" <gi...@apache.org>.
alamb commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872599192
run benchmark clickbench_pushdown
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872344420
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25832#issuecomment-5872287098)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5872287098-2882-vfvrc 6.12.94+ #1 SMP Wed Aug 19 07:47:20 UTC 2026 aarch64 GNU/Linux`
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
Comparing test-arrow-pr-10859 (702bffb69332edf26d82a6d4d6fff3ce8c825343) to 991fd23 (merge-base) [diff](https://github.com/apache/datafusion/compare/991fd23dd0be0046af5945b8c3905612859b657a..702bffb69332edf26d82a6d4d6fff3ce8c825343)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_pushdown
```
</details>
Results will be posted here when complete
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872660407
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25832#issuecomment-5872599192)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5872599192-2888-4xfk4 6.12.94+ #1 SMP Wed Aug 19 07:47:20 UTC 2026 aarch64 GNU/Linux`
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
Comparing test-arrow-pr-10859 (702bffb69332edf26d82a6d4d6fff3ce8c825343) to 991fd23 (merge-base) [diff](https://github.com/apache/datafusion/compare/991fd23dd0be0046af5945b8c3905612859b657a..702bffb69332edf26d82a6d4d6fff3ce8c825343)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_pushdown
```
</details>
Results will be posted here when complete
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "codecov-commenter (via GitHub)" <gi...@apache.org>.
codecov-commenter commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5873903956
## [Codecov](https://app.codecov.io/gh/apache/datafusion/pull/25832?dropdown=coverage&src=pr&el=h1&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache) Report
:white_check_mark: All modified and coverable lines are covered by tests.
:white_check_mark: Project coverage is 82.55%. Comparing base ([`991fd23`](https://app.codecov.io/gh/apache/datafusion/commit/991fd23dd0be0046af5945b8c3905612859b657a?dropdown=coverage&el=desc&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache)) to head ([`702bffb`](https://app.codecov.io/gh/apache/datafusion/commit/702bffb69332edf26d82a6d4d6fff3ce8c825343?dropdown=coverage&el=desc&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache)).
<details><summary>Additional details and impacted files</summary>
```diff
@@ Coverage Diff @@
## main #25832 +/- ##
==========================================
- Coverage 82.55% 82.55% -0.01%
==========================================
Files 1141 1141
Lines 440209 440209
Branches 440209 440209
==========================================
- Hits 363409 363401 -8
- Misses 54829 54833 +4
- Partials 21971 21975 +4
```
</details>
[:umbrella: View full report in Codecov by Harness](https://app.codecov.io/gh/apache/datafusion/pull/25832?dropdown=coverage&src=pr&el=continue&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache).
:loudspeaker: Have feedback on the report? [Share it here](https://about.codecov.io/codecov-pr-comment-feedback/?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache).
<details><summary> :rocket: New features to boost your workflow: </summary>
- :snowflake: [Test Analytics](https://docs.codecov.com/docs/test-analytics): Detect flaky tests, report on failures, and find test suite problems.
- :package: [JS Bundle Analysis](https://docs.codecov.com/docs/javascript-bundle-analysis): Save yourself from yourself by tracking and limiting bundle sizes in JS merges.
</details>
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872672063
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25832#issuecomment-5872602693)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5872602693-2889-52g4q 6.12.94+ #1 SMP Wed Aug 19 07:47:20 UTC 2026 aarch64 GNU/Linux`
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
Comparing test-arrow-pr-10859 (702bffb69332edf26d82a6d4d6fff3ce8c825343) to 991fd23 (merge-base) [diff](https://github.com/apache/datafusion/compare/991fd23dd0be0046af5945b8c3905612859b657a..702bffb69332edf26d82a6d4d6fff3ce8c825343)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: "true"
```
</details>
Results will be posted here when complete
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "alamb (via GitHub)" <gi...@apache.org>.
alamb commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872602693
run benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: "true"
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872842740
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25832#issuecomment-5872599192)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing test-arrow-pr-10859 (702bffb69332edf26d82a6d4d6fff3ce8c825343) to 991fd23 (merge-base) [diff](https://github.com/apache/datafusion/compare/991fd23dd0be0046af5945b8c3905612859b657a..702bffb69332edf26d82a6d4d6fff3ce8c825343)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_pushdown
```
</details>
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and test-arrow-pr-10859
--------------------
Benchmark clickbench_pushdown.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β test-arrow-pr-10859 β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.26 ms β 1.24 ms β no change β
β QQuery 1 β 12.65 ms β 12.29 ms β no change β
β QQuery 2 β 37.01 ms β 37.22 ms β no change β
β QQuery 3 β 31.95 ms β 31.81 ms β no change β
β QQuery 4 β 235.77 ms β 246.92 ms β no change β
β QQuery 5 β 276.73 ms β 278.56 ms β no change β
β QQuery 6 β 1.28 ms β 1.29 ms β no change β
β QQuery 7 β 16.00 ms β 15.84 ms β no change β
β QQuery 8 β 338.71 ms β 349.02 ms β no change β
β QQuery 9 β 462.47 ms β 483.64 ms β no change β
β QQuery 10 β 89.19 ms β 93.41 ms β no change β
β QQuery 11 β 100.64 ms β 100.88 ms β no change β
β QQuery 12 β 294.20 ms β 288.55 ms β no change β
β QQuery 13 β 423.43 ms β 403.98 ms β no change β
β QQuery 14 β 313.12 ms β 304.93 ms β no change β
β QQuery 15 β 289.58 ms β 288.86 ms β no change β
β QQuery 16 β 623.11 ms β 617.90 ms β no change β
β QQuery 17 β 624.97 ms β 642.97 ms β no change β
β QQuery 18 β 1266.62 ms β 1329.38 ms β no change β
β QQuery 19 β 29.51 ms β 30.39 ms β no change β
β QQuery 20 β 520.15 ms β 527.56 ms β no change β
β QQuery 21 β 559.10 ms β 551.91 ms β no change β
β QQuery 22 β 913.52 ms β 907.57 ms β no change β
β QQuery 23 β 112.98 ms β 116.29 ms β no change β
β QQuery 24 β 36.69 ms β 33.70 ms β +1.09x faster β
β QQuery 25 β 135.12 ms β 106.29 ms β +1.27x faster β
β QQuery 26 β 48.29 ms β 45.93 ms β no change β
β QQuery 27 β 552.84 ms β 552.05 ms β no change β
β QQuery 28 β 2915.34 ms β 2911.44 ms β no change β
β QQuery 29 β 43.05 ms β 41.89 ms β no change β
β QQuery 30 β 311.66 ms β 309.46 ms β no change β
β QQuery 31 β 292.70 ms β 273.72 ms β +1.07x faster β
β QQuery 32 β 991.83 ms β 1015.54 ms β no change β
β QQuery 33 β 1478.86 ms β 1433.56 ms β no change β
β QQuery 34 β 1478.97 ms β 1500.91 ms β no change β
β QQuery 35 β 297.64 ms β 291.71 ms β no change β
β QQuery 36 β 64.15 ms β 66.07 ms β no change β
β QQuery 37 β 35.64 ms β 35.78 ms β no change β
β QQuery 38 β 33.23 ms β 32.86 ms β no change β
β QQuery 39 β 135.47 ms β 133.88 ms β no change β
β QQuery 40 β 17.68 ms β 17.83 ms β no change β
β QQuery 41 β 16.59 ms β 16.74 ms β no change β
β QQuery 42 β 14.53 ms β 14.37 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 16474.25ms β
β Total Time (test-arrow-pr-10859) β 16496.13ms β
β Average Time (HEAD) β 383.12ms β
β Average Time (test-arrow-pr-10859) β 383.63ms β
β Queries Faster β 3 β
β Queries Slower β 0 β
β Queries with No Change β 40 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and test-arrow-pr-10859
--------------------
Benchmark clickbench_pushdown.json
--------------------
βββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β test-arrow-pr-10859 β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.26 / 4.16 Β±5.63 / 15.42 ms β 1.24 / 4.17 Β±5.67 / 15.52 ms β no change β
β QQuery 1 β 12.65 / 12.86 Β±0.24 / 13.33 ms β 12.29 / 12.55 Β±0.22 / 12.81 ms β no change β
β QQuery 2 β 37.01 / 37.56 Β±0.34 / 38.05 ms β 37.22 / 37.52 Β±0.17 / 37.68 ms β no change β
β QQuery 3 β 31.95 / 32.66 Β±0.53 / 33.21 ms β 31.81 / 32.09 Β±0.33 / 32.68 ms β no change β
β QQuery 4 β 235.77 / 243.27 Β±4.67 / 249.55 ms β 246.92 / 248.98 Β±1.56 / 251.22 ms β no change β
β QQuery 5 β 276.73 / 277.99 Β±0.75 / 278.72 ms β 278.56 / 283.55 Β±3.26 / 288.82 ms β no change β
β QQuery 6 β 1.28 / 1.44 Β±0.23 / 1.89 ms β 1.29 / 1.44 Β±0.24 / 1.91 ms β no change β
β QQuery 7 β 16.00 / 17.49 Β±2.13 / 21.72 ms β 15.84 / 18.13 Β±3.51 / 25.12 ms β no change β
β QQuery 8 β 338.71 / 347.36 Β±5.39 / 355.01 ms β 349.02 / 357.28 Β±6.77 / 366.66 ms β no change β
β QQuery 9 β 462.47 / 478.41 Β±11.95 / 488.46 ms β 483.64 / 495.80 Β±8.13 / 505.38 ms β no change β
β QQuery 10 β 89.19 / 93.85 Β±6.76 / 107.18 ms β 93.41 / 96.13 Β±1.91 / 98.97 ms β no change β
β QQuery 11 β 100.64 / 101.64 Β±0.99 / 103.42 ms β 100.88 / 103.17 Β±2.13 / 107.19 ms β no change β
β QQuery 12 β 294.20 / 303.43 Β±6.30 / 311.99 ms β 288.55 / 297.46 Β±8.81 / 312.22 ms β no change β
β QQuery 13 β 423.43 / 438.11 Β±12.19 / 454.85 ms β 403.98 / 416.11 Β±13.40 / 432.70 ms β +1.05x faster β
β QQuery 14 β 313.12 / 317.71 Β±4.27 / 324.78 ms β 304.93 / 316.93 Β±9.87 / 331.74 ms β no change β
β QQuery 15 β 289.58 / 297.31 Β±4.33 / 302.25 ms β 288.86 / 305.73 Β±29.47 / 364.62 ms β no change β
β QQuery 16 β 623.11 / 632.64 Β±6.77 / 640.37 ms β 617.90 / 642.57 Β±14.47 / 659.71 ms β no change β
β QQuery 17 β 624.97 / 638.70 Β±7.82 / 648.89 ms β 642.97 / 659.23 Β±13.87 / 682.55 ms β no change β
β QQuery 18 β 1266.62 / 1301.37 Β±31.62 / 1348.43 ms β 1329.38 / 1349.88 Β±18.63 / 1380.58 ms β no change β
β QQuery 19 β 29.51 / 29.96 Β±0.47 / 30.80 ms β 30.39 / 33.49 Β±5.75 / 44.99 ms β 1.12x slower β
β QQuery 20 β 520.15 / 530.78 Β±9.66 / 548.71 ms β 527.56 / 539.46 Β±6.61 / 546.96 ms β no change β
β QQuery 21 β 559.10 / 566.87 Β±6.55 / 575.22 ms β 551.91 / 564.72 Β±10.77 / 578.78 ms β no change β
β QQuery 22 β 913.52 / 927.21 Β±10.79 / 941.03 ms β 907.57 / 911.49 Β±3.08 / 916.51 ms β no change β
β QQuery 23 β 112.98 / 121.97 Β±6.78 / 129.93 ms β 116.29 / 122.00 Β±8.05 / 137.50 ms β no change β
β QQuery 24 β 36.69 / 37.16 Β±0.50 / 37.78 ms β 33.70 / 34.68 Β±0.78 / 35.81 ms β +1.07x faster β
β QQuery 25 β 135.12 / 137.27 Β±1.65 / 139.06 ms β 106.29 / 109.07 Β±1.91 / 111.43 ms β +1.26x faster β
β QQuery 26 β 48.29 / 51.33 Β±4.68 / 60.53 ms β 45.93 / 49.07 Β±3.62 / 55.86 ms β no change β
β QQuery 27 β 552.84 / 571.64 Β±16.21 / 601.23 ms β 552.05 / 567.39 Β±13.96 / 593.45 ms β no change β
β QQuery 28 β 2915.34 / 2956.83 Β±29.09 / 3005.92 ms β 2911.44 / 2927.29 Β±15.61 / 2948.38 ms β no change β
β QQuery 29 β 43.05 / 70.16 Β±21.71 / 101.25 ms β 41.89 / 56.86 Β±20.17 / 96.10 ms β +1.23x faster β
β QQuery 30 β 311.66 / 324.95 Β±14.03 / 350.74 ms β 309.46 / 317.20 Β±7.45 / 330.97 ms β no change β
β QQuery 31 β 292.70 / 309.97 Β±11.24 / 324.18 ms β 273.72 / 297.17 Β±26.82 / 348.48 ms β no change β
β QQuery 32 β 991.83 / 1007.11 Β±20.46 / 1046.56 ms β 1015.54 / 1037.62 Β±27.40 / 1088.46 ms β no change β
β QQuery 33 β 1478.86 / 1507.38 Β±26.33 / 1541.82 ms β 1433.56 / 1507.74 Β±50.32 / 1574.33 ms β no change β
β QQuery 34 β 1478.97 / 1534.36 Β±46.45 / 1592.01 ms β 1500.91 / 1549.68 Β±51.96 / 1646.44 ms β no change β
β QQuery 35 β 297.64 / 324.68 Β±24.21 / 365.13 ms β 291.71 / 314.77 Β±27.10 / 364.73 ms β no change β
β QQuery 36 β 64.15 / 74.29 Β±6.45 / 82.77 ms β 66.07 / 77.45 Β±9.25 / 91.58 ms β no change β
β QQuery 37 β 35.64 / 37.33 Β±1.22 / 38.74 ms β 35.78 / 39.25 Β±2.81 / 42.61 ms β 1.05x slower β
β QQuery 38 β 33.23 / 36.51 Β±2.19 / 38.89 ms β 32.86 / 34.37 Β±1.46 / 37.00 ms β +1.06x faster β
β QQuery 39 β 135.47 / 141.53 Β±4.80 / 147.44 ms β 133.88 / 140.25 Β±5.79 / 150.87 ms β no change β
β QQuery 40 β 17.68 / 18.11 Β±0.37 / 18.56 ms β 17.83 / 18.25 Β±0.44 / 19.05 ms β no change β
β QQuery 41 β 16.59 / 16.77 Β±0.18 / 17.08 ms β 16.74 / 17.28 Β±0.78 / 18.82 ms β no change β
β QQuery 42 β 14.53 / 14.77 Β±0.18 / 15.08 ms β 14.37 / 14.52 Β±0.14 / 14.77 ms β no change β
βββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 16926.91ms β
β Total Time (test-arrow-pr-10859) β 16959.81ms β
β Average Time (HEAD) β 393.65ms β
β Average Time (test-arrow-pr-10859) β 394.41ms β
β Queries Faster β 5 β
β Queries Slower β 2 β
β Queries with No Change β 36 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_pushdown β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 90.0s |
| Peak memory | 12.5 GiB |
| Avg memory | 4.5 GiB |
| CPU user | 861.9s |
| CPU sys | 59.6s |
| Peak spill | 0 B |
**clickbench_pushdown β branch**
| Metric | Value |
|--------|-------|
| Wall time | 90.0s |
| Peak memory | 13.2 GiB |
| Avg memory | 5.0 GiB |
| CPU user | 861.5s |
| CPU sys | 60.2s |
| Peak spill | 0 B |
</details>
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "alamb (via GitHub)" <gi...@apache.org>.
alamb commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5892121443
Performance and local confirmation here:
- https://github.com/apache/arrow-rs/pull/10859#issuecomment-5892107366
TLDR q25 gets faster with pushdown on, all other peformance is about the same
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "alamb (via GitHub)" <gi...@apache.org>.
alamb closed pull request #25832: [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench
URL: https://github.com/apache/datafusion/pull/25832
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872566172
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25832#issuecomment-5872245663)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing test-arrow-pr-10859 (702bffb69332edf26d82a6d4d6fff3ce8c825343) to 991fd23 (merge-base) [diff](https://github.com/apache/datafusion/compare/991fd23dd0be0046af5945b8c3905612859b657a..702bffb69332edf26d82a6d4d6fff3ce8c825343)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: "true"
```
</details>
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and test-arrow-pr-10859
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β test-arrow-pr-10859 β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.45 ms β 1.24 ms β +1.17x faster β
β QQuery 1 β 12.79 ms β 12.19 ms β no change β
β QQuery 2 β 37.73 ms β 36.84 ms β no change β
β QQuery 3 β 33.45 ms β 31.42 ms β +1.06x faster β
β QQuery 4 β 242.74 ms β 232.78 ms β no change β
β QQuery 5 β 270.46 ms β 268.75 ms β no change β
β QQuery 6 β 1.34 ms β 1.58 ms β 1.18x slower β
β QQuery 7 β 16.65 ms β 17.39 ms β no change β
β QQuery 8 β 332.11 ms β 363.78 ms β 1.10x slower β
β QQuery 9 β 465.05 ms β 468.67 ms β no change β
β QQuery 10 β 88.92 ms β 89.19 ms β no change β
β QQuery 11 β 99.65 ms β 99.91 ms β no change β
β QQuery 12 β 289.25 ms β 286.96 ms β no change β
β QQuery 13 β 409.04 ms β 399.63 ms β no change β
β QQuery 14 β 308.63 ms β 306.23 ms β no change β
β QQuery 15 β 283.84 ms β 285.43 ms β no change β
β QQuery 16 β 632.06 ms β 614.33 ms β no change β
β QQuery 17 β 620.86 ms β 632.07 ms β no change β
β QQuery 18 β 1268.09 ms β 1364.49 ms β 1.08x slower β
β QQuery 19 β 30.94 ms β 29.69 ms β no change β
β QQuery 20 β 511.56 ms β 515.85 ms β no change β
β QQuery 21 β 552.36 ms β 547.51 ms β no change β
β QQuery 22 β 772.31 ms β 777.81 ms β no change β
β QQuery 23 β 126.67 ms β 125.65 ms β no change β
β QQuery 24 β 44.14 ms β 41.45 ms β +1.06x faster β
β QQuery 25 β 133.85 ms β 104.97 ms β +1.28x faster β
β QQuery 26 β 47.40 ms β 45.56 ms β no change β
β QQuery 27 β 538.63 ms β 547.50 ms β no change β
β QQuery 28 β 2932.06 ms β 2891.31 ms β no change β
β QQuery 29 β 41.14 ms β 41.54 ms β no change β
β QQuery 30 β 305.62 ms β 303.91 ms β no change β
β QQuery 31 β 310.12 ms β 280.58 ms β +1.11x faster β
β QQuery 32 β 965.65 ms β 958.72 ms β no change β
β QQuery 33 β 1409.21 ms β 1426.33 ms β no change β
β QQuery 34 β 1448.87 ms β 1479.63 ms β no change β
β QQuery 35 β 295.11 ms β 371.63 ms β 1.26x slower β
β QQuery 36 β 65.52 ms β 78.67 ms β 1.20x slower β
β QQuery 37 β 35.98 ms β 40.27 ms β 1.12x slower β
β QQuery 38 β 35.23 ms β 41.01 ms β 1.16x slower β
β QQuery 39 β 128.58 ms β 161.25 ms β 1.25x slower β
β QQuery 40 β 17.75 ms β 19.62 ms β 1.10x slower β
β QQuery 41 β 15.96 ms β 18.35 ms β 1.15x slower β
β QQuery 42 β 13.98 ms β 16.10 ms β 1.15x slower β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 16192.77ms β
β Total Time (test-arrow-pr-10859) β 16377.81ms β
β Average Time (HEAD) β 376.58ms β
β Average Time (test-arrow-pr-10859) β 380.88ms β
β Queries Faster β 5 β
β Queries Slower β 11 β
β Queries with No Change β 27 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and test-arrow-pr-10859
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³βββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β test-arrow-pr-10859 β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.45 / 4.60 Β±5.89 / 16.36 ms β 1.24 / 4.00 Β±5.45 / 14.90 ms β +1.15x faster β
β QQuery 1 β 12.79 / 13.05 Β±0.19 / 13.29 ms β 12.19 / 12.37 Β±0.25 / 12.86 ms β +1.05x faster β
β QQuery 2 β 37.73 / 38.23 Β±0.36 / 38.66 ms β 36.84 / 37.18 Β±0.26 / 37.64 ms β no change β
β QQuery 3 β 33.45 / 34.01 Β±0.69 / 35.22 ms β 31.42 / 31.86 Β±0.22 / 32.06 ms β +1.07x faster β
β QQuery 4 β 242.74 / 266.03 Β±18.89 / 283.73 ms β 232.78 / 238.69 Β±3.28 / 242.67 ms β +1.11x faster β
β QQuery 5 β 270.46 / 279.12 Β±7.15 / 290.31 ms β 268.75 / 286.90 Β±17.66 / 312.54 ms β no change β
β QQuery 6 β 1.34 / 1.50 Β±0.25 / 2.00 ms β 1.58 / 1.79 Β±0.19 / 2.13 ms β 1.20x slower β
β QQuery 7 β 16.65 / 16.81 Β±0.09 / 16.93 ms β 17.39 / 18.32 Β±1.63 / 21.58 ms β 1.09x slower β
β QQuery 8 β 332.11 / 337.90 Β±7.83 / 353.32 ms β 363.78 / 378.66 Β±10.98 / 391.71 ms β 1.12x slower β
β QQuery 9 β 465.05 / 472.27 Β±4.30 / 477.42 ms β 468.67 / 481.62 Β±8.36 / 492.34 ms β no change β
β QQuery 10 β 88.92 / 89.77 Β±0.93 / 91.44 ms β 89.19 / 93.57 Β±7.08 / 107.70 ms β no change β
β QQuery 11 β 99.65 / 100.29 Β±0.54 / 101.17 ms β 99.91 / 103.59 Β±4.17 / 111.53 ms β no change β
β QQuery 12 β 289.25 / 294.40 Β±5.64 / 303.69 ms β 286.96 / 292.93 Β±4.96 / 299.33 ms β no change β
β QQuery 13 β 409.04 / 426.80 Β±27.22 / 480.85 ms β 399.63 / 425.31 Β±25.69 / 473.81 ms β no change β
β QQuery 14 β 308.63 / 310.85 Β±2.11 / 313.99 ms β 306.23 / 313.54 Β±7.58 / 322.96 ms β no change β
β QQuery 15 β 283.84 / 299.17 Β±16.53 / 327.80 ms β 285.43 / 294.88 Β±7.86 / 306.00 ms β no change β
β QQuery 16 β 632.06 / 664.46 Β±27.88 / 695.93 ms β 614.33 / 633.14 Β±15.42 / 659.54 ms β no change β
β QQuery 17 β 620.86 / 631.35 Β±5.40 / 636.00 ms β 632.07 / 700.19 Β±45.18 / 745.87 ms β 1.11x slower β
β QQuery 18 β 1268.09 / 1303.59 Β±64.25 / 1431.90 ms β 1364.49 / 1477.75 Β±61.18 / 1538.05 ms β 1.13x slower β
β QQuery 19 β 30.94 / 31.45 Β±0.39 / 32.08 ms β 29.69 / 33.30 Β±5.89 / 45.04 ms β 1.06x slower β
β QQuery 20 β 511.56 / 544.55 Β±30.18 / 595.63 ms β 515.85 / 523.87 Β±10.35 / 544.17 ms β no change β
β QQuery 21 β 552.36 / 564.59 Β±10.56 / 580.52 ms β 547.51 / 574.92 Β±28.76 / 625.76 ms β no change β
β QQuery 22 β 772.31 / 780.45 Β±7.92 / 795.12 ms β 777.81 / 804.60 Β±15.78 / 820.98 ms β no change β
β QQuery 23 β 126.67 / 134.63 Β±8.70 / 151.35 ms β 125.65 / 132.18 Β±3.82 / 137.11 ms β no change β
β QQuery 24 β 44.14 / 46.82 Β±3.35 / 52.77 ms β 41.45 / 43.71 Β±2.17 / 47.47 ms β +1.07x faster β
β QQuery 25 β 133.85 / 137.76 Β±4.14 / 144.81 ms β 104.97 / 112.38 Β±9.48 / 130.26 ms β +1.23x faster β
β QQuery 26 β 47.40 / 49.56 Β±2.78 / 54.91 ms β 45.56 / 46.27 Β±0.78 / 47.75 ms β +1.07x faster β
β QQuery 27 β 538.63 / 549.36 Β±6.75 / 558.11 ms β 547.50 / 565.04 Β±30.47 / 625.77 ms β no change β
β QQuery 28 β 2932.06 / 2994.99 Β±56.86 / 3076.17 ms β 2891.31 / 2955.38 Β±53.07 / 3043.47 ms β no change β
β QQuery 29 β 41.14 / 47.46 Β±7.53 / 59.34 ms β 41.54 / 53.48 Β±15.84 / 81.99 ms β 1.13x slower β
β QQuery 30 β 305.62 / 317.68 Β±10.71 / 335.57 ms β 303.91 / 316.40 Β±17.64 / 351.16 ms β no change β
β QQuery 31 β 310.12 / 318.41 Β±7.01 / 329.50 ms β 280.58 / 284.84 Β±3.95 / 291.96 ms β +1.12x faster β
β QQuery 32 β 965.65 / 1012.37 Β±40.65 / 1078.36 ms β 958.72 / 1028.78 Β±64.48 / 1108.21 ms β no change β
β QQuery 33 β 1409.21 / 1483.92 Β±43.97 / 1533.93 ms β 1426.33 / 1480.57 Β±43.66 / 1559.67 ms β no change β
β QQuery 34 β 1448.87 / 1522.23 Β±75.41 / 1666.28 ms β 1479.63 / 1604.81 Β±124.56 / 1757.72 ms β 1.05x slower β
β QQuery 35 β 295.11 / 330.24 Β±60.69 / 451.52 ms β 371.63 / 388.86 Β±19.66 / 427.34 ms β 1.18x slower β
β QQuery 36 β 65.52 / 69.50 Β±6.33 / 82.13 ms β 78.67 / 83.33 Β±4.17 / 90.28 ms β 1.20x slower β
β QQuery 37 β 35.98 / 42.75 Β±6.86 / 53.81 ms β 40.27 / 46.50 Β±9.47 / 65.23 ms β 1.09x slower β
β QQuery 38 β 35.23 / 38.83 Β±4.05 / 46.27 ms β 41.01 / 47.03 Β±4.47 / 54.02 ms β 1.21x slower β
β QQuery 39 β 128.58 / 138.79 Β±8.27 / 148.20 ms β 161.25 / 171.39 Β±13.46 / 197.65 ms β 1.23x slower β
β QQuery 40 β 17.75 / 21.11 Β±3.72 / 26.94 ms β 19.62 / 20.39 Β±0.44 / 20.80 ms β no change β
β QQuery 41 β 15.96 / 17.64 Β±2.96 / 23.55 ms β 18.35 / 19.70 Β±2.32 / 24.33 ms β 1.12x slower β
β QQuery 42 β 13.98 / 15.56 Β±2.86 / 21.27 ms β 16.10 / 16.27 Β±0.17 / 16.59 ms β no change β
βββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 16794.83ms β
β Total Time (test-arrow-pr-10859) β 17180.30ms β
β Average Time (HEAD) β 390.58ms β
β Average Time (test-arrow-pr-10859) β 399.54ms β
β Queries Faster β 8 β
β Queries Slower β 14 β
β Queries with No Change β 21 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 85.0s |
| Peak memory | 12.1 GiB |
| Avg memory | 5.0 GiB |
| CPU user | 860.9s |
| CPU sys | 58.2s |
| Peak spill | 0 B |
**clickbench_partitioned β branch**
| Metric | Value |
|--------|-------|
| Wall time | 90.0s |
| Peak memory | 11.6 GiB |
| Avg memory | 4.6 GiB |
| CPU user | 871.0s |
| CPU sys | 62.2s |
| Peak spill | 0 B |
</details>
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872569354
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25832#issuecomment-5872287098)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing test-arrow-pr-10859 (702bffb69332edf26d82a6d4d6fff3ce8c825343) to 991fd23 (merge-base) [diff](https://github.com/apache/datafusion/compare/991fd23dd0be0046af5945b8c3905612859b657a..702bffb69332edf26d82a6d4d6fff3ce8c825343)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_pushdown
```
</details>
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and test-arrow-pr-10859
--------------------
Benchmark clickbench_pushdown.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β test-arrow-pr-10859 β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.24 ms β 1.31 ms β 1.06x slower β
β QQuery 1 β 13.71 ms β 12.50 ms β +1.10x faster β
β QQuery 2 β 37.38 ms β 37.07 ms β no change β
β QQuery 3 β 31.91 ms β 31.94 ms β no change β
β QQuery 4 β 250.73 ms β 249.06 ms β no change β
β QQuery 5 β 282.37 ms β 285.06 ms β no change β
β QQuery 6 β 1.30 ms β 1.34 ms β no change β
β QQuery 7 β 16.35 ms β 16.46 ms β no change β
β QQuery 8 β 343.43 ms β 351.33 ms β no change β
β QQuery 9 β 499.55 ms β 482.35 ms β no change β
β QQuery 10 β 93.55 ms β 91.02 ms β no change β
β QQuery 11 β 104.41 ms β 101.66 ms β no change β
β QQuery 12 β 301.69 ms β 298.80 ms β no change β
β QQuery 13 β 411.75 ms β 412.81 ms β no change β
β QQuery 14 β 327.78 ms β 321.64 ms β no change β
β QQuery 15 β 303.00 ms β 298.53 ms β no change β
β QQuery 16 β 665.41 ms β 657.84 ms β no change β
β QQuery 17 β 655.41 ms β 650.24 ms β no change β
β QQuery 18 β 1312.25 ms β 1340.18 ms β no change β
β QQuery 19 β 29.86 ms β 30.37 ms β no change β
β QQuery 20 β 524.10 ms β 522.70 ms β no change β
β QQuery 21 β 569.45 ms β 552.26 ms β no change β
β QQuery 22 β 936.04 ms β 920.14 ms β no change β
β QQuery 23 β 118.00 ms β 120.93 ms β no change β
β QQuery 24 β 36.40 ms β 33.22 ms β +1.10x faster β
β QQuery 25 β 139.45 ms β 108.83 ms β +1.28x faster β
β QQuery 26 β 51.34 ms β 46.27 ms β +1.11x faster β
β QQuery 27 β 556.66 ms β 549.96 ms β no change β
β QQuery 28 β 3026.18 ms β 2966.38 ms β no change β
β QQuery 29 β 43.11 ms β 42.52 ms β no change β
β QQuery 30 β 335.81 ms β 323.41 ms β no change β
β QQuery 31 β 295.85 ms β 291.89 ms β no change β
β QQuery 32 β 1042.24 ms β 1015.22 ms β no change β
β QQuery 33 β 1549.70 ms β 1536.44 ms β no change β
β QQuery 34 β 1550.97 ms β 1582.01 ms β no change β
β QQuery 35 β 329.91 ms β 328.18 ms β no change β
β QQuery 36 β 70.00 ms β 68.05 ms β no change β
β QQuery 37 β 37.48 ms β 37.63 ms β no change β
β QQuery 38 β 39.19 ms β 39.20 ms β no change β
β QQuery 39 β 132.52 ms β 143.35 ms β 1.08x slower β
β QQuery 40 β 18.74 ms β 18.17 ms β no change β
β QQuery 41 β 17.85 ms β 17.75 ms β no change β
β QQuery 42 β 15.77 ms β 15.17 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 17119.83ms β
β Total Time (test-arrow-pr-10859) β 16951.20ms β
β Average Time (HEAD) β 398.14ms β
β Average Time (test-arrow-pr-10859) β 394.21ms β
β Queries Faster β 4 β
β Queries Slower β 2 β
β Queries with No Change β 37 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and test-arrow-pr-10859
--------------------
Benchmark clickbench_pushdown.json
--------------------
βββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β test-arrow-pr-10859 β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.24 / 4.20 Β±5.71 / 15.62 ms β 1.31 / 4.28 Β±5.79 / 15.86 ms β no change β
β QQuery 1 β 13.71 / 13.90 Β±0.12 / 14.07 ms β 12.50 / 12.85 Β±0.28 / 13.35 ms β +1.08x faster β
β QQuery 2 β 37.38 / 38.08 Β±0.80 / 39.53 ms β 37.07 / 37.40 Β±0.36 / 38.07 ms β no change β
β QQuery 3 β 31.91 / 32.76 Β±0.95 / 34.60 ms β 31.94 / 32.22 Β±0.29 / 32.74 ms β no change β
β QQuery 4 β 250.73 / 258.15 Β±4.07 / 261.67 ms β 249.06 / 255.59 Β±3.36 / 258.39 ms β no change β
β QQuery 5 β 282.37 / 286.19 Β±2.18 / 289.04 ms β 285.06 / 288.82 Β±2.36 / 291.95 ms β no change β
β QQuery 6 β 1.30 / 1.46 Β±0.26 / 1.97 ms β 1.34 / 1.49 Β±0.24 / 1.97 ms β no change β
β QQuery 7 β 16.35 / 16.54 Β±0.13 / 16.66 ms β 16.46 / 16.62 Β±0.09 / 16.71 ms β no change β
β QQuery 8 β 343.43 / 352.49 Β±6.90 / 361.35 ms β 351.33 / 356.42 Β±4.65 / 363.11 ms β no change β
β QQuery 9 β 499.55 / 512.52 Β±14.98 / 541.27 ms β 482.35 / 503.93 Β±17.22 / 532.54 ms β no change β
β QQuery 10 β 93.55 / 98.89 Β±8.74 / 116.31 ms β 91.02 / 92.83 Β±1.74 / 96.09 ms β +1.07x faster β
β QQuery 11 β 104.41 / 107.25 Β±3.12 / 112.68 ms β 101.66 / 102.77 Β±0.95 / 104.28 ms β no change β
β QQuery 12 β 301.69 / 313.95 Β±9.89 / 331.32 ms β 298.80 / 305.94 Β±5.51 / 315.85 ms β no change β
β QQuery 13 β 411.75 / 430.65 Β±14.84 / 451.55 ms β 412.81 / 424.58 Β±9.31 / 440.06 ms β no change β
β QQuery 14 β 327.78 / 331.13 Β±4.39 / 339.75 ms β 321.64 / 327.69 Β±6.30 / 335.43 ms β no change β
β QQuery 15 β 303.00 / 311.03 Β±9.18 / 328.63 ms β 298.53 / 309.48 Β±13.39 / 335.22 ms β no change β
β QQuery 16 β 665.41 / 677.60 Β±14.62 / 704.29 ms β 657.84 / 666.30 Β±5.39 / 673.30 ms β no change β
β QQuery 17 β 655.41 / 671.82 Β±13.03 / 689.78 ms β 650.24 / 663.86 Β±12.00 / 684.94 ms β no change β
β QQuery 18 β 1312.25 / 1363.65 Β±27.47 / 1388.50 ms β 1340.18 / 1388.46 Β±28.44 / 1416.98 ms β no change β
β QQuery 19 β 29.86 / 35.82 Β±6.62 / 45.92 ms β 30.37 / 34.25 Β±7.24 / 48.73 ms β no change β
β QQuery 20 β 524.10 / 534.57 Β±10.66 / 554.36 ms β 522.70 / 534.72 Β±7.12 / 541.63 ms β no change β
β QQuery 21 β 569.45 / 581.77 Β±12.16 / 603.17 ms β 552.26 / 570.34 Β±9.77 / 580.01 ms β no change β
β QQuery 22 β 936.04 / 945.85 Β±9.14 / 962.58 ms β 920.14 / 926.65 Β±8.46 / 943.29 ms β no change β
β QQuery 23 β 118.00 / 124.90 Β±5.60 / 133.04 ms β 120.93 / 127.85 Β±7.21 / 139.64 ms β no change β
β QQuery 24 β 36.40 / 38.04 Β±2.51 / 43.00 ms β 33.22 / 40.13 Β±9.97 / 59.93 ms β 1.06x slower β
β QQuery 25 β 139.45 / 142.69 Β±2.11 / 145.85 ms β 108.83 / 112.80 Β±6.38 / 125.52 ms β +1.26x faster β
β QQuery 26 β 51.34 / 57.57 Β±9.10 / 75.09 ms β 46.27 / 49.38 Β±3.99 / 57.23 ms β +1.17x faster β
β QQuery 27 β 556.66 / 569.16 Β±8.64 / 579.97 ms β 549.96 / 564.60 Β±11.07 / 579.58 ms β no change β
β QQuery 28 β 3026.18 / 3051.49 Β±16.39 / 3074.00 ms β 2966.38 / 2996.26 Β±17.49 / 3018.50 ms β no change β
β QQuery 29 β 43.11 / 45.51 Β±2.73 / 50.85 ms β 42.52 / 52.98 Β±14.44 / 79.85 ms β 1.16x slower β
β QQuery 30 β 335.81 / 340.13 Β±4.82 / 346.59 ms β 323.41 / 342.18 Β±20.39 / 376.32 ms β no change β
β QQuery 31 β 295.85 / 309.46 Β±7.96 / 320.04 ms β 291.89 / 309.76 Β±14.62 / 331.10 ms β no change β
β QQuery 32 β 1042.24 / 1047.94 Β±10.33 / 1068.56 ms β 1015.22 / 1047.38 Β±18.06 / 1068.50 ms β no change β
β QQuery 33 β 1549.70 / 1600.32 Β±41.55 / 1656.64 ms β 1536.44 / 1598.53 Β±49.59 / 1676.05 ms β no change β
β QQuery 34 β 1550.97 / 1585.90 Β±20.44 / 1613.06 ms β 1582.01 / 1600.16 Β±21.02 / 1636.97 ms β no change β
β QQuery 35 β 329.91 / 354.86 Β±30.73 / 415.37 ms β 328.18 / 358.75 Β±43.92 / 443.03 ms β no change β
β QQuery 36 β 70.00 / 75.62 Β±7.55 / 90.46 ms β 68.05 / 78.01 Β±7.63 / 86.98 ms β no change β
β QQuery 37 β 37.48 / 43.42 Β±7.18 / 56.66 ms β 37.63 / 39.29 Β±1.85 / 42.88 ms β +1.10x faster β
β QQuery 38 β 39.19 / 44.34 Β±3.94 / 49.38 ms β 39.20 / 44.86 Β±3.05 / 48.18 ms β no change β
β QQuery 39 β 132.52 / 151.65 Β±12.27 / 170.34 ms β 143.35 / 152.02 Β±7.60 / 164.47 ms β no change β
β QQuery 40 β 18.74 / 22.48 Β±4.94 / 31.70 ms β 18.17 / 20.12 Β±3.07 / 26.24 ms β +1.12x faster β
β QQuery 41 β 17.85 / 20.05 Β±3.49 / 26.92 ms β 17.75 / 19.82 Β±3.23 / 26.18 ms β no change β
β QQuery 42 β 15.77 / 16.19 Β±0.29 / 16.61 ms β 15.17 / 17.45 Β±4.02 / 25.48 ms β 1.08x slower β
βββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 17561.97ms β
β Total Time (test-arrow-pr-10859) β 17429.82ms β
β Average Time (HEAD) β 408.42ms β
β Average Time (test-arrow-pr-10859) β 405.34ms β
β Queries Faster β 6 β
β Queries Slower β 3 β
β Queries with No Change β 34 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_pushdown β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 90.0s |
| Peak memory | 11.4 GiB |
| Avg memory | 4.8 GiB |
| CPU user | 893.0s |
| CPU sys | 64.8s |
| Peak spill | 0 B |
**clickbench_pushdown β branch**
| Metric | Value |
|--------|-------|
| Wall time | 90.0s |
| Peak memory | 12.2 GiB |
| Avg memory | 5.0 GiB |
| CPU user | 885.0s |
| CPU sys | 63.6s |
| Peak spill | 0 B |
</details>
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "alamb (via GitHub)" <gi...@apache.org>.
alamb commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872245663
run benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: "true"
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] [DO NOT MERGE] Test apache/arrow-rs#10859 (fuse consecutive same-projection filters) with pushdown tests and ClickBench [datafusion]
Posted by "alamb (via GitHub)" <gi...@apache.org>.
alamb commented on PR #25832:
URL: https://github.com/apache/datafusion/pull/25832#issuecomment-5872281522
run clickbench_pushdown
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org