You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@datafusion.apache.org by "adriangb (via GitHub)" <gi...@apache.org> on 2026/09/09 03:46:22 UTC
[PR] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
adriangb opened a new pull request, #25093:
URL: https://github.com/apache/datafusion/pull/25093
> [!NOTE]
> **Benchmark-only draft. Do not review or merge.** This exists so the benchmark bot has a buildable `changed` side for https://github.com/apache/datafusion/pull/21621. I will close it once the runs finish.
## Which issue does this PR close?
- Related to https://github.com/apache/datafusion/issues/11900 and https://github.com/apache/datafusion/pull/21621. Closes nothing.
## Rationale for this change
https://github.com/apache/datafusion/pull/21621 is 289 commits behind `main`, conflicts in `benchmarks/src/bin/dfbench.rs`, and its head no longer compiles the `datafusion-benchmarks` crate (`push_down_topk.rs:188` calls `print_memory_stats()` with no argument). The benchmark bot builds `dfbench` on the PR head, so triggering benchmarks on that PR fails on the `changed` side, and custom refs must be reachable from `apache/datafusion`.
This branch is current `main` plus exactly the optimizer change from that PR, so `run benchmarks` here compares `main` against the rule with the `push_down_topk` suite from https://github.com/apache/datafusion/pull/22760 present on both sides.
## What changes are included in this PR?
Merge of https://github.com/apache/datafusion/pull/21621 onto `main`, touching only:
- `datafusion/optimizer/src/push_down_limit.rs`
- `datafusion/optimizer/src/push_down_limit/topk_through_join.rs`
- `datafusion/core/src/optimizer_rule_reference.md`
- `datafusion/sqllogictest/test_files/push_down_topk_through_join.slt`
The PR's duplicate `dfbench push-down-topk` subcommand and `benchmarks/queries/push_down_topk/*.sql` are dropped; the `sql_benchmarks/push_down_topk` suite already on `main` supersedes them.
## What is the testing strategy for this PR?
None beyond the tests carried over from the source PR. Not for merge.
## Are there any user-facing changes?
No.
π€ Generated with [Claude Code](https://claude.com/claude-code)
--
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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595616616
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595469801)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
```
</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 pr21621-bench-on-main
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.22 ms β 1.21 ms β no change β
β QQuery 1 β 12.00 ms β 11.94 ms β no change β
β QQuery 2 β 36.83 ms β 36.98 ms β no change β
β QQuery 3 β 30.88 ms β 30.88 ms β no change β
β QQuery 4 β 222.21 ms β 221.45 ms β no change β
β QQuery 5 β 271.40 ms β 271.37 ms β no change β
β QQuery 6 β 1.29 ms β 1.25 ms β no change β
β QQuery 7 β 13.70 ms β 13.18 ms β no change β
β QQuery 8 β 322.25 ms β 325.02 ms β no change β
β QQuery 9 β 456.74 ms β 442.21 ms β no change β
β QQuery 10 β 69.74 ms β 70.43 ms β no change β
β QQuery 11 β 80.56 ms β 81.35 ms β no change β
β QQuery 12 β 266.29 ms β 269.53 ms β no change β
β QQuery 13 β 364.32 ms β 365.09 ms β no change β
β QQuery 14 β 280.34 ms β 282.72 ms β no change β
β QQuery 15 β 267.71 ms β 272.25 ms β no change β
β QQuery 16 β 618.10 ms β 615.54 ms β no change β
β QQuery 17 β 624.90 ms β 621.64 ms β no change β
β QQuery 18 β 1256.55 ms β 1262.74 ms β no change β
β QQuery 19 β 27.49 ms β 27.55 ms β no change β
β QQuery 20 β 511.75 ms β 516.56 ms β no change β
β QQuery 21 β 514.84 ms β 518.00 ms β no change β
β QQuery 22 β 971.31 ms β 983.86 ms β no change β
β QQuery 23 β 3000.13 ms β 3062.62 ms β no change β
β QQuery 24 β 40.72 ms β 40.75 ms β no change β
β QQuery 25 β 110.62 ms β 110.81 ms β no change β
β QQuery 26 β 41.16 ms β 41.05 ms β no change β
β QQuery 27 β 512.66 ms β 523.46 ms β no change β
β QQuery 28 β 2911.70 ms β 2921.73 ms β no change β
β QQuery 29 β 41.78 ms β 40.53 ms β no change β
β QQuery 30 β 304.32 ms β 306.12 ms β no change β
β QQuery 31 β 280.48 ms β 283.33 ms β no change β
β QQuery 32 β 952.04 ms β 915.58 ms β no change β
β QQuery 33 β 1468.76 ms β 1446.55 ms β no change β
β QQuery 34 β 1450.99 ms β 1466.10 ms β no change β
β QQuery 35 β 280.14 ms β 278.91 ms β no change β
β QQuery 36 β 68.23 ms β 66.93 ms β no change β
β QQuery 37 β 35.06 ms β 35.49 ms β no change β
β QQuery 38 β 44.00 ms β 40.37 ms β +1.09x faster β
β QQuery 39 β 132.12 ms β 143.12 ms β 1.08x slower β
β QQuery 40 β 14.80 ms β 14.32 ms β no change β
β QQuery 41 β 13.88 ms β 13.51 ms β no change β
β QQuery 42 β 13.32 ms β 13.28 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 18939.33ms β
β Total Time (pr21621-bench-on-main) β 19007.34ms β
β Average Time (HEAD) β 440.45ms β
β Average Time (pr21621-bench-on-main) β 442.03ms β
β Queries Faster β 1 β
β Queries Slower β 1 β
β Queries with No Change β 41 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and pr21621-bench-on-main
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.22 / 3.92 Β±5.35 / 14.62 ms β 1.21 / 3.92 Β±5.36 / 14.65 ms β no change β
β QQuery 1 β 12.00 / 12.17 Β±0.16 / 12.48 ms β 11.94 / 12.18 Β±0.19 / 12.40 ms β no change β
β QQuery 2 β 36.83 / 37.15 Β±0.27 / 37.60 ms β 36.98 / 37.41 Β±0.27 / 37.72 ms β no change β
β QQuery 3 β 30.88 / 31.82 Β±1.07 / 33.89 ms β 30.88 / 31.33 Β±0.31 / 31.83 ms β no change β
β QQuery 4 β 222.21 / 226.03 Β±2.54 / 230.06 ms β 221.45 / 226.68 Β±3.65 / 232.68 ms β no change β
β QQuery 5 β 271.40 / 273.55 Β±1.29 / 275.10 ms β 271.37 / 273.95 Β±1.96 / 276.47 ms β no change β
β QQuery 6 β 1.29 / 1.43 Β±0.23 / 1.88 ms β 1.25 / 1.39 Β±0.22 / 1.82 ms β no change β
β QQuery 7 β 13.70 / 14.43 Β±0.83 / 15.71 ms β 13.18 / 13.57 Β±0.34 / 14.19 ms β +1.06x faster β
β QQuery 8 β 322.25 / 329.13 Β±4.78 / 336.90 ms β 325.02 / 329.57 Β±2.90 / 333.37 ms β no change β
β QQuery 9 β 456.74 / 458.24 Β±2.10 / 462.31 ms β 442.21 / 457.33 Β±15.49 / 486.24 ms β no change β
β QQuery 10 β 69.74 / 70.67 Β±1.04 / 72.02 ms β 70.43 / 71.66 Β±0.85 / 73.00 ms β no change β
β QQuery 11 β 80.56 / 81.67 Β±0.89 / 82.81 ms β 81.35 / 82.60 Β±0.98 / 84.31 ms β no change β
β QQuery 12 β 266.29 / 271.33 Β±4.40 / 277.41 ms β 269.53 / 273.44 Β±2.78 / 277.50 ms β no change β
β QQuery 13 β 364.32 / 381.16 Β±12.87 / 399.85 ms β 365.09 / 388.82 Β±23.91 / 432.70 ms β no change β
β QQuery 14 β 280.34 / 286.87 Β±7.66 / 301.48 ms β 282.72 / 287.76 Β±2.97 / 290.31 ms β no change β
β QQuery 15 β 267.71 / 275.34 Β±5.77 / 285.35 ms β 272.25 / 280.61 Β±7.27 / 291.72 ms β no change β
β QQuery 16 β 618.10 / 624.78 Β±7.09 / 634.67 ms β 615.54 / 636.35 Β±24.81 / 683.88 ms β no change β
β QQuery 17 β 624.90 / 630.20 Β±5.43 / 640.39 ms β 621.64 / 640.55 Β±16.04 / 665.29 ms β no change β
β QQuery 18 β 1256.55 / 1286.68 Β±16.54 / 1301.43 ms β 1262.74 / 1285.63 Β±16.71 / 1313.64 ms β no change β
β QQuery 19 β 27.49 / 31.64 Β±7.56 / 46.75 ms β 27.55 / 32.76 Β±9.91 / 52.58 ms β no change β
β QQuery 20 β 511.75 / 527.78 Β±13.61 / 545.68 ms β 516.56 / 521.46 Β±4.02 / 527.91 ms β no change β
β QQuery 21 β 514.84 / 521.98 Β±9.61 / 540.77 ms β 518.00 / 521.33 Β±2.84 / 526.39 ms β no change β
β QQuery 22 β 971.31 / 990.20 Β±10.28 / 1001.40 ms β 983.86 / 994.62 Β±8.36 / 1005.83 ms β no change β
β QQuery 23 β 3000.13 / 3024.42 Β±18.97 / 3055.48 ms β 3062.62 / 3090.14 Β±21.45 / 3118.87 ms β no change β
β QQuery 24 β 40.72 / 40.85 Β±0.11 / 40.96 ms β 40.75 / 50.96 Β±18.14 / 87.22 ms β 1.25x slower β
β QQuery 25 β 110.62 / 112.19 Β±0.93 / 113.34 ms β 110.81 / 113.22 Β±2.06 / 116.71 ms β no change β
β QQuery 26 β 41.16 / 41.69 Β±0.53 / 42.68 ms β 41.05 / 41.31 Β±0.25 / 41.78 ms β no change β
β QQuery 27 β 512.66 / 521.08 Β±5.51 / 529.62 ms β 523.46 / 528.20 Β±6.88 / 541.64 ms β no change β
β QQuery 28 β 2911.70 / 2934.59 Β±15.90 / 2952.86 ms β 2921.73 / 2944.10 Β±16.74 / 2965.30 ms β no change β
β QQuery 29 β 41.78 / 56.46 Β±13.66 / 74.55 ms β 40.53 / 44.45 Β±7.27 / 58.98 ms β +1.27x faster β
β QQuery 30 β 304.32 / 310.84 Β±5.87 / 321.54 ms β 306.12 / 313.46 Β±10.28 / 333.87 ms β no change β
β QQuery 31 β 280.48 / 290.62 Β±9.14 / 305.41 ms β 283.33 / 296.19 Β±16.39 / 328.15 ms β no change β
β QQuery 32 β 952.04 / 981.00 Β±18.03 / 1004.89 ms β 915.58 / 938.53 Β±19.20 / 970.56 ms β no change β
β QQuery 33 β 1468.76 / 1492.00 Β±19.87 / 1515.68 ms β 1446.55 / 1475.29 Β±17.82 / 1498.87 ms β no change β
β QQuery 34 β 1450.99 / 1490.13 Β±21.50 / 1512.78 ms β 1466.10 / 1496.45 Β±19.82 / 1524.81 ms β no change β
β QQuery 35 β 280.14 / 307.85 Β±49.05 / 405.85 ms β 278.91 / 322.68 Β±46.05 / 407.43 ms β no change β
β QQuery 36 β 68.23 / 71.34 Β±2.35 / 75.41 ms β 66.93 / 71.55 Β±4.18 / 77.85 ms β no change β
β QQuery 37 β 35.06 / 38.84 Β±6.94 / 52.71 ms β 35.49 / 40.21 Β±4.88 / 48.56 ms β no change β
β QQuery 38 β 44.00 / 48.92 Β±3.41 / 52.85 ms β 40.37 / 43.54 Β±2.20 / 46.97 ms β +1.12x faster β
β QQuery 39 β 132.12 / 145.30 Β±7.27 / 151.55 ms β 143.12 / 156.90 Β±9.65 / 171.94 ms β 1.08x slower β
β QQuery 40 β 14.80 / 18.27 Β±6.49 / 31.24 ms β 14.32 / 17.49 Β±5.74 / 28.96 ms β no change β
β QQuery 41 β 13.88 / 15.84 Β±1.51 / 17.44 ms β 13.51 / 14.23 Β±0.70 / 15.43 ms β +1.11x faster β
β QQuery 42 β 13.32 / 17.44 Β±7.98 / 33.38 ms β 13.28 / 13.56 Β±0.15 / 13.71 ms β +1.29x faster β
βββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 19327.81ms β
β Total Time (pr21621-bench-on-main) β 19417.38ms β
β Average Time (HEAD) β 449.48ms β
β Average Time (pr21621-bench-on-main) β 451.57ms 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
ββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 100.0s |
| Peak memory | 11.0 GiB |
| Avg memory | 4.3 GiB |
| CPU user | 992.2s |
| CPU sys | 67.2s |
| Peak spill | 0 B |
**clickbench_partitioned β branch**
| Metric | Value |
|--------|-------|
| Wall time | 100.0s |
| Peak memory | 11.5 GiB |
| Avg memory | 4.6 GiB |
| CPU user | 995.1s |
| CPU sys | 68.4s |
| 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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595632189
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595606896)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5595606896-2238-wkq6q 6.12.94+ #1 SMP Fri Jul 17 09:42:57 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 pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpcds
```
</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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "codecov-commenter (via GitHub)" <gi...@apache.org>.
codecov-commenter commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595688799
## [Codecov](https://app.codecov.io/gh/apache/datafusion/pull/25093?dropdown=coverage&src=pr&el=h1&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache) Report
:x: Patch coverage is `77.87611%` with `200 lines` in your changes missing coverage. Please review.
:white_check_mark: Project coverage is 81.79%. Comparing base ([`da89c7c`](https://app.codecov.io/gh/apache/datafusion/commit/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a?dropdown=coverage&el=desc&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache)) to head ([`53d3a41`](https://app.codecov.io/gh/apache/datafusion/commit/53d3a410e1e96b844ff6096e372344a2927c8faa?dropdown=coverage&el=desc&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache)).
:warning: Report is 103 commits behind head on main.
| [Files with missing lines](https://app.codecov.io/gh/apache/datafusion/pull/25093?dropdown=coverage&src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache) | Patch % | Lines |
|---|---|---|
| [...optimizer/src/push\_down\_limit/topk\_through\_join.rs](https://app.codecov.io/gh/apache/datafusion/pull/25093?src=pr&el=tree&filepath=datafusion%2Foptimizer%2Fsrc%2Fpush_down_limit%2Ftopk_through_join.rs&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache#diff-ZGF0YWZ1c2lvbi9vcHRpbWl6ZXIvc3JjL3B1c2hfZG93bl9saW1pdC90b3BrX3Rocm91Z2hfam9pbi5ycw==) | 75.95% | [42 Missing and 153 partials :warning: ](https://app.codecov.io/gh/apache/datafusion/pull/25093?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache) |
| [datafusion/optimizer/src/push\_down\_limit.rs](https://app.codecov.io/gh/apache/datafusion/pull/25093?src=pr&el=tree&filepath=datafusion%2Foptimizer%2Fsrc%2Fpush_down_limit.rs&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache#diff-ZGF0YWZ1c2lvbi9vcHRpbWl6ZXIvc3JjL3B1c2hfZG93bl9saW1pdC5ycw==) | 94.62% | [0 Missing and 5 partials :warning: ](https://app.codecov.io/gh/apache/datafusion/pull/25093?src=pr&el=tree&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 #25093 +/- ##
==========================================
+ Coverage 81.60% 81.79% +0.18%
==========================================
Files 1123 1131 +8
Lines 408898 418526 +9628
Branches 408898 418526 +9628
==========================================
+ Hits 333670 342313 +8643
- Misses 55625 55929 +304
- Partials 19603 20284 +681
```
</details>
[:umbrella: View full report in Codecov by Harness](https://app.codecov.io/gh/apache/datafusion/pull/25093?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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595493429
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595469940)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5595469940-2234-mztv6 6.12.94+ #1 SMP Fri Jul 17 09:42:57 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 pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark topk_tpch
```
</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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangb (via GitHub)" <gi...@apache.org>.
adriangb closed pull request #25093: [benchmark only] PR #21621 push-down TopK through join, rebased on main
URL: https://github.com/apache/datafusion/pull/25093
--
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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangb (via GitHub)" <gi...@apache.org>.
adriangb commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595728605
run benchmark tpcds
baseline:
ref: 53d3a410e1e96b844ff6096e372344a2927c8faa
--
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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangb (via GitHub)" <gi...@apache.org>.
adriangb commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595606896
run benchmark tpcds
--
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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595584178
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595469940)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark topk_tpch
```
</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 pr21621-bench-on-main
--------------------
Benchmark run_topk_tpch.json
--------------------
βββββββββ³βββββββββββ³ββββββββββββββββββββββββ³ββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Q1 β 2.18 ms β 2.23 ms β no change β
β Q2 β 2.89 ms β 2.94 ms β no change β
β Q3 β 31.74 ms β 31.86 ms β no change β
β Q4 β 3.33 ms β 3.22 ms β no change β
β Q5 β 10.49 ms β 10.37 ms β no change β
β Q6 β 17.42 ms β 17.47 ms β no change β
β Q7 β 37.76 ms β 38.26 ms β no change β
β Q8 β 6.91 ms β 6.95 ms β no change β
β Q9 β 8.62 ms β 8.71 ms β no change β
β Q10 β 12.77 ms β 12.99 ms β no change β
β Q11 β 3.90 ms β 4.01 ms β no change β
βββββββββ΄βββββββββββ΄ββββββββββββββββββββββββ΄ββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³βββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 138.02ms β
β Total Time (pr21621-bench-on-main) β 139.03ms β
β Average Time (HEAD) β 12.55ms β
β Average Time (pr21621-bench-on-main) β 12.64ms β
β Queries Faster β 0 β
β Queries Slower β 0 β
β Queries with No Change β 11 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄βββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and pr21621-bench-on-main
--------------------
Benchmark run_topk_tpch.json
--------------------
βββββββββ³βββββββββββββββββββββββββββββββββ³βββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Q1 β 2.18 / 2.94 Β±0.91 / 4.46 ms β 2.23 / 2.96 Β±0.91 / 4.64 ms β no change β
β Q2 β 2.89 / 3.38 Β±0.77 / 4.91 ms β 2.94 / 3.14 Β±0.20 / 3.46 ms β +1.08x faster β
β Q3 β 31.74 / 33.63 Β±1.74 / 36.43 ms β 31.86 / 33.03 Β±1.15 / 34.42 ms β no change β
β Q4 β 3.33 / 3.40 Β±0.11 / 3.62 ms β 3.22 / 3.76 Β±0.75 / 5.24 ms β 1.10x slower β
β Q5 β 10.49 / 10.73 Β±0.28 / 11.27 ms β 10.37 / 10.45 Β±0.08 / 10.60 ms β no change β
β Q6 β 17.42 / 17.54 Β±0.19 / 17.92 ms β 17.47 / 17.57 Β±0.07 / 17.66 ms β no change β
β Q7 β 37.76 / 38.10 Β±0.53 / 39.15 ms β 38.26 / 38.55 Β±0.38 / 39.29 ms β no change β
β Q8 β 6.91 / 7.52 Β±1.06 / 9.63 ms β 6.95 / 7.97 Β±1.30 / 10.11 ms β 1.06x slower β
β Q9 β 8.62 / 9.31 Β±1.24 / 11.79 ms β 8.71 / 9.44 Β±1.39 / 12.21 ms β no change β
β Q10 β 12.77 / 12.96 Β±0.19 / 13.28 ms β 12.99 / 13.08 Β±0.07 / 13.17 ms β no change β
β Q11 β 3.90 / 4.44 Β±0.92 / 6.27 ms β 4.01 / 4.24 Β±0.39 / 5.02 ms β no change β
βββββββββ΄βββββββββββββββββββββββββββββββββ΄βββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³βββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 143.94ms β
β Total Time (pr21621-bench-on-main) β 144.19ms β
β Average Time (HEAD) β 13.09ms β
β Average Time (pr21621-bench-on-main) β 13.11ms β
β Queries Faster β 1 β
β Queries Slower β 2 β
β Queries with No Change β 8 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄βββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**topk_tpch β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 5.0s |
| Peak memory | 619.5 MiB |
| Avg memory | 103.3 MiB |
| CPU user | 0.0s |
| CPU sys | 0.0s |
| Peak spill | 0 B |
**topk_tpch β branch**
| Metric | Value |
|--------|-------|
| Wall time | 5.0s |
| Peak memory | 260.0 KiB |
| Avg memory | 43.3 KiB |
| CPU user | 0.0s |
| CPU sys | 0.0s |
| 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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangb (via GitHub)" <gi...@apache.org>.
adriangb commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5596184448
Benchmark runs complete (default suite, push_down_topk, topk_tpch, sql_planner, plus tpcds control and A/A runs). Closing as announced; results stay in the comments above.
--
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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595720206
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595606896)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpcds
```
</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 pr21621-bench-on-main
--------------------
Benchmark tpcds_sf1.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 5.58 ms β 5.54 ms β no change β
β QQuery 2 β 80.72 ms β 79.94 ms β no change β
β QQuery 3 β 28.37 ms β 28.85 ms β no change β
β QQuery 4 β 465.65 ms β 485.26 ms β no change β
β QQuery 5 β 51.78 ms β 51.54 ms β no change β
β QQuery 6 β 36.02 ms β 35.81 ms β no change β
β QQuery 7 β 74.22 ms β 74.28 ms β no change β
β QQuery 8 β 36.14 ms β 36.26 ms β no change β
β QQuery 9 β 52.87 ms β 51.35 ms β no change β
β QQuery 10 β 62.45 ms β 63.20 ms β no change β
β QQuery 11 β 294.08 ms β 315.52 ms β 1.07x slower β
β QQuery 12 β 27.95 ms β 29.54 ms β 1.06x slower β
β QQuery 13 β 117.97 ms β 116.69 ms β no change β
β QQuery 14 β 417.18 ms β 416.65 ms β no change β
β QQuery 15 β 66.09 ms β 56.25 ms β +1.17x faster β
β QQuery 16 β 7.55 ms β 6.69 ms β +1.13x faster β
β QQuery 17 β 83.49 ms β 78.99 ms β +1.06x faster β
β QQuery 18 β 104.66 ms β 103.36 ms β no change β
β QQuery 19 β 41.40 ms β 41.27 ms β no change β
β QQuery 20 β 35.64 ms β 35.10 ms β no change β
β QQuery 21 β 16.89 ms β 16.86 ms β no change β
β QQuery 22 β 63.98 ms β 62.95 ms β no change β
β QQuery 23 β 306.12 ms β 309.51 ms β no change β
β QQuery 24 β 194.92 ms β 193.48 ms β no change β
β QQuery 25 β 109.62 ms β 108.99 ms β no change β
β QQuery 26 β 48.33 ms β 48.36 ms β no change β
β QQuery 27 β 6.06 ms β 6.23 ms β no change β
β QQuery 28 β 56.30 ms β 59.56 ms β 1.06x slower β
β QQuery 29 β 96.42 ms β 95.78 ms β no change β
β QQuery 30 β 31.91 ms β 31.88 ms β no change β
β QQuery 31 β 109.51 ms β 108.96 ms β no change β
β QQuery 32 β 19.77 ms β 19.82 ms β no change β
β QQuery 33 β 37.47 ms β 37.29 ms β no change β
β QQuery 34 β 9.81 ms β 10.02 ms β no change β
β QQuery 35 β 71.80 ms β 72.09 ms β no change β
β QQuery 36 β 5.85 ms β 6.01 ms β no change β
β QQuery 37 β 6.80 ms β 6.80 ms β no change β
β QQuery 38 β 61.24 ms β 61.46 ms β no change β
β QQuery 39 β 88.40 ms β 88.82 ms β no change β
β QQuery 40 β 23.62 ms β 24.26 ms β no change β
β QQuery 41 β 11.05 ms β 12.11 ms β 1.10x slower β
β QQuery 42 β 23.72 ms β 25.09 ms β 1.06x slower β
β QQuery 43 β 5.37 ms β 5.85 ms β 1.09x slower β
β QQuery 44 β 9.24 ms β 10.30 ms β 1.12x slower β
β QQuery 45 β 37.85 ms β 42.49 ms β 1.12x slower β
β QQuery 46 β 11.72 ms β 13.00 ms β 1.11x slower β
β QQuery 47 β 222.34 ms β 222.97 ms β no change β
β QQuery 48 β 94.99 ms β 94.62 ms β no change β
β QQuery 49 β 70.61 ms β 70.79 ms β no change β
β QQuery 50 β 58.36 ms β 58.26 ms β no change β
β QQuery 51 β 91.54 ms β 92.78 ms β no change β
β QQuery 52 β 23.56 ms β 24.08 ms β no change β
β QQuery 53 β 28.53 ms β 28.81 ms β no change β
β QQuery 54 β 53.55 ms β 53.81 ms β no change β
β QQuery 55 β 23.17 ms β 22.94 ms β no change β
β QQuery 56 β 37.63 ms β 38.86 ms β no change β
β QQuery 57 β 173.27 ms β 174.63 ms β no change β
β QQuery 58 β 111.32 ms β 110.83 ms β no change β
β QQuery 59 β 117.69 ms β 117.01 ms β no change β
β QQuery 60 β 39.23 ms β 39.08 ms β no change β
β QQuery 61 β 12.29 ms β 12.41 ms β no change β
β QQuery 62 β 45.64 ms β 45.71 ms β no change β
β QQuery 63 β 29.09 ms β 29.01 ms β no change β
β QQuery 64 β 367.45 ms β 363.64 ms β no change β
β QQuery 65 β 130.39 ms β 122.55 ms β +1.06x faster β
β QQuery 66 β 88.44 ms β 79.64 ms β +1.11x faster β
β QQuery 67 β 282.39 ms β 237.93 ms β +1.19x faster β
β QQuery 68 β 13.18 ms β 12.02 ms β +1.10x faster β
β QQuery 69 β 59.67 ms β 57.41 ms β no change β
β QQuery 70 β 111.85 ms β 105.03 ms β +1.06x faster β
β QQuery 71 β 37.55 ms β 35.13 ms β +1.07x faster β
β QQuery 72 β 1793.30 ms β 1843.85 ms β no change β
β QQuery 73 β 10.63 ms β 10.99 ms β no change β
β QQuery 74 β 195.22 ms β 210.04 ms β 1.08x slower β
β QQuery 75 β 146.47 ms β 153.80 ms β no change β
β QQuery 76 β 36.91 ms β 37.04 ms β no change β
β QQuery 77 β 63.03 ms β 60.63 ms β no change β
β QQuery 78 β 222.28 ms β 220.81 ms β no change β
β QQuery 79 β 65.91 ms β 66.07 ms β no change β
β QQuery 80 β 98.40 ms β 98.66 ms β no change β
β QQuery 81 β 25.30 ms β 25.70 ms β no change β
β QQuery 82 β 16.03 ms β 15.97 ms β no change β
β QQuery 83 β 33.45 ms β 33.45 ms β no change β
β QQuery 84 β 29.17 ms β 29.01 ms β no change β
β QQuery 85 β 102.50 ms β 102.60 ms β no change β
β QQuery 86 β 24.92 ms β 24.67 ms β no change β
β QQuery 87 β 61.14 ms β 60.57 ms β no change β
β QQuery 88 β 64.27 ms β 62.62 ms β no change β
β QQuery 89 β 35.39 ms β 35.00 ms β no change β
β QQuery 90 β 17.35 ms β 16.91 ms β no change β
β QQuery 91 β 45.31 ms β 44.88 ms β no change β
β QQuery 92 β 29.73 ms β 28.38 ms β no change β
β QQuery 93 β 48.67 ms β 49.21 ms β no change β
β QQuery 94 β 37.73 ms β 37.49 ms β no change β
β QQuery 95 β 79.41 ms β 80.97 ms β no change β
β QQuery 96 β 23.75 ms β 23.90 ms β no change β
β QQuery 97 β 53.48 ms β 51.77 ms β no change β
β QQuery 98 β 42.44 ms β 42.89 ms β no change β
β QQuery 99 β 69.53 ms β 69.85 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββ
β Benchmark Summary β β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 9349.06ms β
β Total Time (pr21621-bench-on-main) β 9377.71ms β
β Average Time (HEAD) β 94.43ms β
β Average Time (pr21621-bench-on-main) β 94.72ms β
β Queries Faster β 9 β
β Queries Slower β 10 β
β Queries with No Change β 80 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and pr21621-bench-on-main
--------------------
Benchmark tpcds_sf1.json
--------------------
βββββββββββββ³βββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 5.58 / 6.11 Β±0.95 / 8.02 ms β 5.54 / 6.08 Β±0.85 / 7.77 ms β no change β
β QQuery 2 β 80.72 / 80.95 Β±0.12 / 81.05 ms β 79.94 / 80.49 Β±0.36 / 80.96 ms β no change β
β QQuery 3 β 28.37 / 28.88 Β±0.28 / 29.17 ms β 28.85 / 29.12 Β±0.22 / 29.37 ms β no change β
β QQuery 4 β 465.65 / 477.83 Β±10.47 / 497.18 ms β 485.26 / 504.94 Β±20.23 / 538.22 ms β 1.06x slower β
β QQuery 5 β 51.78 / 52.73 Β±0.72 / 53.77 ms β 51.54 / 52.34 Β±0.83 / 53.58 ms β no change β
β QQuery 6 β 36.02 / 36.30 Β±0.21 / 36.66 ms β 35.81 / 36.07 Β±0.22 / 36.36 ms β no change β
β QQuery 7 β 74.22 / 74.97 Β±0.57 / 76.00 ms β 74.28 / 74.85 Β±0.37 / 75.33 ms β no change β
β QQuery 8 β 36.14 / 36.36 Β±0.22 / 36.71 ms β 36.26 / 37.90 Β±2.10 / 41.92 ms β no change β
β QQuery 9 β 52.87 / 54.88 Β±1.05 / 55.79 ms β 51.35 / 52.93 Β±1.35 / 54.85 ms β no change β
β QQuery 10 β 62.45 / 62.73 Β±0.28 / 63.25 ms β 63.20 / 63.59 Β±0.26 / 63.96 ms β no change β
β QQuery 11 β 294.08 / 297.26 Β±3.32 / 303.70 ms β 315.52 / 343.69 Β±17.08 / 364.03 ms β 1.16x slower β
β QQuery 12 β 27.95 / 28.23 Β±0.30 / 28.72 ms β 29.54 / 30.42 Β±0.52 / 31.06 ms β 1.08x slower β
β QQuery 13 β 117.97 / 118.29 Β±0.26 / 118.60 ms β 116.69 / 119.54 Β±3.38 / 126.17 ms β no change β
β QQuery 14 β 417.18 / 420.37 Β±2.38 / 423.70 ms β 416.65 / 418.86 Β±1.50 / 421.22 ms β no change β
β QQuery 15 β 66.09 / 68.26 Β±1.44 / 70.37 ms β 56.25 / 56.99 Β±0.55 / 57.67 ms β +1.20x faster β
β QQuery 16 β 7.55 / 7.70 Β±0.12 / 7.86 ms β 6.69 / 6.83 Β±0.13 / 7.08 ms β +1.13x faster β
β QQuery 17 β 83.49 / 85.38 Β±2.21 / 89.06 ms β 78.99 / 80.31 Β±1.56 / 83.34 ms β +1.06x faster β
β QQuery 18 β 104.66 / 111.54 Β±4.19 / 116.52 ms β 103.36 / 103.90 Β±0.87 / 105.62 ms β +1.07x faster β
β QQuery 19 β 41.40 / 41.59 Β±0.19 / 41.91 ms β 41.27 / 41.67 Β±0.36 / 42.30 ms β no change β
β QQuery 20 β 35.64 / 36.33 Β±0.71 / 37.63 ms β 35.10 / 36.26 Β±1.25 / 38.58 ms β no change β
β QQuery 21 β 16.89 / 17.33 Β±0.37 / 17.85 ms β 16.86 / 17.13 Β±0.17 / 17.32 ms β no change β
β QQuery 22 β 63.98 / 66.31 Β±1.99 / 68.94 ms β 62.95 / 63.84 Β±0.63 / 64.65 ms β no change β
β QQuery 23 β 306.12 / 310.36 Β±3.22 / 315.99 ms β 309.51 / 313.56 Β±4.05 / 320.32 ms β no change β
β QQuery 24 β 194.92 / 197.54 Β±2.94 / 202.59 ms β 193.48 / 196.39 Β±3.95 / 204.08 ms β no change β
β QQuery 25 β 109.62 / 112.03 Β±1.87 / 114.98 ms β 108.99 / 110.15 Β±0.80 / 111.25 ms β no change β
β QQuery 26 β 48.33 / 48.63 Β±0.19 / 48.92 ms β 48.36 / 48.84 Β±0.43 / 49.66 ms β no change β
β QQuery 27 β 6.06 / 6.22 Β±0.17 / 6.55 ms β 6.23 / 6.43 Β±0.20 / 6.81 ms β no change β
β QQuery 28 β 56.30 / 59.98 Β±1.87 / 61.59 ms β 59.56 / 62.26 Β±2.98 / 67.97 ms β no change β
β QQuery 29 β 96.42 / 99.09 Β±2.59 / 103.64 ms β 95.78 / 96.72 Β±0.71 / 97.93 ms β no change β
β QQuery 30 β 31.91 / 32.18 Β±0.15 / 32.34 ms β 31.88 / 32.16 Β±0.25 / 32.62 ms β no change β
β QQuery 31 β 109.51 / 110.20 Β±0.35 / 110.49 ms β 108.96 / 110.08 Β±0.71 / 110.79 ms β no change β
β QQuery 32 β 19.77 / 20.05 Β±0.40 / 20.84 ms β 19.82 / 20.15 Β±0.40 / 20.91 ms β no change β
β QQuery 33 β 37.47 / 38.36 Β±1.52 / 41.39 ms β 37.29 / 37.67 Β±0.25 / 37.97 ms β no change β
β QQuery 34 β 9.81 / 10.04 Β±0.23 / 10.42 ms β 10.02 / 10.16 Β±0.15 / 10.41 ms β no change β
β QQuery 35 β 71.80 / 73.51 Β±1.61 / 75.56 ms β 72.09 / 74.17 Β±3.39 / 80.93 ms β no change β
β QQuery 36 β 5.85 / 6.05 Β±0.23 / 6.51 ms β 6.01 / 6.11 Β±0.16 / 6.42 ms β no change β
β QQuery 37 β 6.80 / 6.95 Β±0.09 / 7.06 ms β 6.80 / 6.92 Β±0.10 / 7.04 ms β no change β
β QQuery 38 β 61.24 / 61.98 Β±0.49 / 62.64 ms β 61.46 / 62.28 Β±0.80 / 63.81 ms β no change β
β QQuery 39 β 88.40 / 89.63 Β±1.24 / 92.03 ms β 88.82 / 89.54 Β±0.48 / 90.23 ms β no change β
β QQuery 40 β 23.62 / 25.12 Β±2.56 / 30.23 ms β 24.26 / 27.77 Β±4.08 / 35.58 ms β 1.11x slower β
β QQuery 41 β 11.05 / 11.15 Β±0.11 / 11.36 ms β 12.11 / 12.28 Β±0.12 / 12.41 ms β 1.10x slower β
β QQuery 42 β 23.72 / 24.35 Β±0.45 / 24.98 ms β 25.09 / 25.60 Β±0.65 / 26.80 ms β 1.05x slower β
β QQuery 43 β 5.37 / 5.44 Β±0.11 / 5.65 ms β 5.85 / 5.93 Β±0.06 / 6.03 ms β 1.09x slower β
β QQuery 44 β 9.24 / 9.41 Β±0.16 / 9.70 ms β 10.30 / 10.70 Β±0.55 / 11.78 ms β 1.14x slower β
β QQuery 45 β 37.85 / 38.12 Β±0.27 / 38.52 ms β 42.49 / 46.36 Β±1.99 / 48.12 ms β 1.22x slower β
β QQuery 46 β 11.72 / 12.62 Β±0.88 / 14.31 ms β 13.00 / 13.34 Β±0.28 / 13.74 ms β 1.06x slower β
β QQuery 47 β 222.34 / 226.20 Β±3.78 / 232.13 ms β 222.97 / 251.91 Β±21.60 / 276.93 ms β 1.11x slower β
β QQuery 48 β 94.99 / 96.21 Β±1.75 / 99.68 ms β 94.62 / 97.19 Β±2.75 / 102.11 ms β no change β
β QQuery 49 β 70.61 / 71.74 Β±1.11 / 73.77 ms β 70.79 / 71.33 Β±0.55 / 72.29 ms β no change β
β QQuery 50 β 58.36 / 59.25 Β±0.60 / 60.11 ms β 58.26 / 58.75 Β±0.28 / 59.06 ms β no change β
β QQuery 51 β 91.54 / 93.46 Β±1.64 / 95.81 ms β 92.78 / 96.81 Β±4.05 / 104.23 ms β no change β
β QQuery 52 β 23.56 / 23.79 Β±0.24 / 24.20 ms β 24.08 / 24.21 Β±0.13 / 24.40 ms β no change β
β QQuery 53 β 28.53 / 29.46 Β±1.00 / 31.37 ms β 28.81 / 29.25 Β±0.46 / 30.11 ms β no change β
β QQuery 54 β 53.55 / 53.82 Β±0.24 / 54.23 ms β 53.81 / 54.06 Β±0.26 / 54.48 ms β no change β
β QQuery 55 β 23.17 / 23.38 Β±0.17 / 23.60 ms β 22.94 / 25.45 Β±3.47 / 32.31 ms β 1.09x slower β
β QQuery 56 β 37.63 / 38.61 Β±0.80 / 39.96 ms β 38.86 / 39.25 Β±0.39 / 39.93 ms β no change β
β QQuery 57 β 173.27 / 175.30 Β±1.64 / 178.03 ms β 174.63 / 175.67 Β±0.80 / 176.74 ms β no change β
β QQuery 58 β 111.32 / 113.16 Β±2.30 / 117.44 ms β 110.83 / 112.22 Β±1.12 / 113.84 ms β no change β
β QQuery 59 β 117.69 / 118.85 Β±1.63 / 122.06 ms β 117.01 / 118.94 Β±1.88 / 122.38 ms β no change β
β QQuery 60 β 39.23 / 40.00 Β±0.61 / 40.98 ms β 39.08 / 39.72 Β±0.54 / 40.65 ms β no change β
β QQuery 61 β 12.29 / 12.48 Β±0.18 / 12.80 ms β 12.41 / 12.65 Β±0.19 / 12.89 ms β no change β
β QQuery 62 β 45.64 / 46.50 Β±1.09 / 48.62 ms β 45.71 / 45.92 Β±0.18 / 46.24 ms β no change β
β QQuery 63 β 29.09 / 29.52 Β±0.50 / 30.40 ms β 29.01 / 29.15 Β±0.17 / 29.49 ms β no change β
β QQuery 64 β 367.45 / 389.79 Β±13.73 / 407.40 ms β 363.64 / 371.07 Β±5.66 / 380.78 ms β no change β
β QQuery 65 β 130.39 / 133.44 Β±3.08 / 137.82 ms β 122.55 / 126.98 Β±2.91 / 130.11 ms β no change β
β QQuery 66 β 88.44 / 89.64 Β±1.45 / 92.50 ms β 79.64 / 80.66 Β±0.79 / 82.03 ms β +1.11x faster β
β QQuery 67 β 282.39 / 284.89 Β±1.57 / 286.48 ms β 237.93 / 243.23 Β±3.82 / 248.69 ms β +1.17x faster β
β QQuery 68 β 13.18 / 13.36 Β±0.13 / 13.54 ms β 12.02 / 12.53 Β±0.36 / 13.08 ms β +1.07x faster β
β QQuery 69 β 59.67 / 61.16 Β±1.42 / 63.80 ms β 57.41 / 57.80 Β±0.57 / 58.92 ms β +1.06x faster β
β QQuery 70 β 111.85 / 116.89 Β±5.10 / 123.59 ms β 105.03 / 106.88 Β±1.36 / 108.47 ms β +1.09x faster β
β QQuery 71 β 37.55 / 37.74 Β±0.17 / 38.03 ms β 35.13 / 38.37 Β±5.25 / 48.85 ms β no change β
β QQuery 72 β 1793.30 / 1876.64 Β±100.40 / 2069.56 ms β 1843.85 / 1972.08 Β±97.96 / 2105.65 ms β 1.05x slower β
β QQuery 73 β 10.63 / 10.90 Β±0.22 / 11.13 ms β 10.99 / 13.53 Β±4.03 / 21.55 ms β 1.24x slower β
β QQuery 74 β 195.22 / 203.33 Β±4.56 / 207.72 ms β 210.04 / 211.83 Β±1.09 / 213.35 ms β no change β
β QQuery 75 β 146.47 / 152.85 Β±4.74 / 159.06 ms β 153.80 / 157.29 Β±2.35 / 160.58 ms β no change β
β QQuery 76 β 36.91 / 37.10 Β±0.13 / 37.30 ms β 37.04 / 38.24 Β±0.91 / 39.53 ms β no change β
β QQuery 77 β 63.03 / 65.07 Β±2.80 / 70.50 ms β 60.63 / 62.24 Β±1.48 / 64.21 ms β no change β
β QQuery 78 β 222.28 / 225.80 Β±2.87 / 230.38 ms β 220.81 / 227.81 Β±8.20 / 243.75 ms β no change β
β QQuery 79 β 65.91 / 68.54 Β±3.48 / 75.30 ms β 66.07 / 67.12 Β±1.30 / 69.56 ms β no change β
β QQuery 80 β 98.40 / 99.61 Β±1.11 / 101.22 ms β 98.66 / 102.18 Β±2.76 / 106.63 ms β no change β
β QQuery 81 β 25.30 / 25.72 Β±0.21 / 25.87 ms β 25.70 / 26.02 Β±0.26 / 26.48 ms β no change β
β QQuery 82 β 16.03 / 16.28 Β±0.24 / 16.73 ms β 15.97 / 16.36 Β±0.21 / 16.56 ms β no change β
β QQuery 83 β 33.45 / 34.85 Β±2.04 / 38.91 ms β 33.45 / 33.74 Β±0.25 / 34.17 ms β no change β
β QQuery 84 β 29.17 / 30.58 Β±1.78 / 34.02 ms β 29.01 / 29.36 Β±0.22 / 29.68 ms β no change β
β QQuery 85 β 102.50 / 104.36 Β±2.64 / 109.51 ms β 102.60 / 106.76 Β±4.49 / 114.89 ms β no change β
β QQuery 86 β 24.92 / 25.18 Β±0.27 / 25.70 ms β 24.67 / 25.54 Β±0.61 / 26.55 ms β no change β
β QQuery 87 β 61.14 / 62.82 Β±1.44 / 64.91 ms β 60.57 / 61.52 Β±0.84 / 62.95 ms β no change β
β QQuery 88 β 64.27 / 64.98 Β±1.14 / 67.25 ms β 62.62 / 65.11 Β±3.81 / 72.71 ms β no change β
β QQuery 89 β 35.39 / 36.67 Β±1.64 / 39.85 ms β 35.00 / 36.00 Β±0.78 / 37.40 ms β no change β
β QQuery 90 β 17.35 / 17.49 Β±0.10 / 17.61 ms β 16.91 / 17.22 Β±0.19 / 17.45 ms β no change β
β QQuery 91 β 45.31 / 45.54 Β±0.29 / 46.09 ms β 44.88 / 44.97 Β±0.05 / 45.03 ms β no change β
β QQuery 92 β 29.73 / 30.25 Β±0.44 / 31.05 ms β 28.38 / 29.11 Β±0.52 / 30.00 ms β no change β
β QQuery 93 β 48.67 / 50.50 Β±1.56 / 52.94 ms β 49.21 / 50.05 Β±0.75 / 51.09 ms β no change β
β QQuery 94 β 37.73 / 38.01 Β±0.25 / 38.45 ms β 37.49 / 38.66 Β±0.80 / 39.78 ms β no change β
β QQuery 95 β 79.41 / 80.75 Β±1.45 / 82.59 ms β 80.97 / 81.62 Β±0.49 / 82.44 ms β no change β
β QQuery 96 β 23.75 / 23.98 Β±0.17 / 24.26 ms β 23.90 / 24.25 Β±0.20 / 24.50 ms β no change β
β QQuery 97 β 53.48 / 55.26 Β±2.23 / 59.53 ms β 51.77 / 52.82 Β±1.28 / 55.30 ms β no change β
β QQuery 98 β 42.44 / 42.70 Β±0.18 / 42.99 ms β 42.89 / 45.33 Β±2.76 / 50.70 ms β 1.06x slower β
β QQuery 99 β 69.53 / 70.64 Β±1.77 / 74.15 ms β 69.85 / 70.04 Β±0.15 / 70.25 ms β no change β
βββββββββββββ΄βββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββ
β Benchmark Summary β β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 9594.38ms β
β Total Time (pr21621-bench-on-main) β 9710.13ms β
β Average Time (HEAD) β 96.91ms β
β Average Time (pr21621-bench-on-main) β 98.08ms β
β Queries Faster β 9 β
β Queries Slower β 15 β
β Queries with No Change β 75 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 50.0s |
| Peak memory | 2.3 GiB |
| Avg memory | 1.6 GiB |
| CPU user | 206.6s |
| CPU sys | 5.7s |
| Peak spill | 0 B |
**tpcds β branch**
| Metric | Value |
|--------|-------|
| Wall time | 50.0s |
| Peak memory | 2.2 GiB |
| Avg memory | 1.4 GiB |
| CPU user | 210.7s |
| CPU sys | 5.8s |
| 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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595493408
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595469940)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5595469940-2233-9cj5f 6.12.94+ #1 SMP Fri Jul 17 09:42:57 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 pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark push_down_topk
```
</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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangb (via GitHub)" <gi...@apache.org>.
adriangb commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595469801
run benchmarks
--
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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595677625
Benchmark for [this request](https://github.com/apache/datafusion/pull/25093#issuecomment-5595469940) failed before finishing (Kubernetes reason: `BackoffLimitExceeded`).
Benchmarks requested: `push_down_topk`
<details><summary>Runner log (last 40 lines)</summary>
```
2026-09-09T03:48:23.051991Z INFO runner starting benchmark runner bench_type=Datafusion, pr_url=https://github.com/apache/datafusion/pull/25093, benchmarks=push_down_topk
2026-09-09T03:48:23.110658Z INFO benchmark_controller::runner::bench_datafusion === Cloning PR branch ===
2026-09-09T03:48:23.110766Z INFO benchmark_controller::runner::shell running command cmd=git, args=["clone", "--depth=200", "https://github.com/apache/datafusion.git", "/workspace/datafusion-branch"], cwd="/"
2026-09-09T03:48:28.121028Z INFO benchmark_controller::runner::shell running command cmd=git, args=["fetch", "origin", "refs/pull/25093/head:pr21621-bench-on-main", "main"], cwd="/workspace/datafusion-branch"
2026-09-09T03:48:48.125905Z INFO benchmark_controller::runner::shell running command cmd=git, args=["checkout", "pr21621-bench-on-main"], cwd="/workspace/datafusion-branch"
2026-09-09T03:48:53.128362Z INFO benchmark_controller::runner::shell running command cmd=git, args=["merge-base", "HEAD", "origin/main"], cwd="/workspace/datafusion-branch"
2026-09-09T03:48:58.130683Z INFO benchmark_controller::runner::bench_datafusion === Cloning merge-base ===
2026-09-09T03:48:58.130698Z INFO benchmark_controller::runner::shell running command cmd=git, args=["clone", "--depth=200", "https://github.com/apache/datafusion.git", "/workspace/datafusion-base"], cwd="/"
2026-09-09T03:49:03.132922Z INFO benchmark_controller::runner::shell running command cmd=git, args=["-c", "advice.detachedHead=false", "checkout", "da89c7c85b1c888ba4038317eacbdcfe03ab9b1a"], cwd="/workspace/datafusion-base"
2026-09-09T03:49:08.134467Z INFO benchmark_controller::runner::shell running command cmd=rustc, args=["--version"], cwd="/"
2026-09-09T03:49:13.136872Z INFO benchmark_controller::runner::shell running command cmd=cargo, args=["metadata", "--no-deps", "--format-version", "1"], cwd="/workspace/datafusion-branch/benchmarks"
2026-09-09T03:49:18.141051Z INFO benchmark_controller::runner::bench_datafusion === Compiling dfbench for PR branch and merge-base in parallel ===
2026-09-09T03:49:18.170540Z INFO benchmark_controller::runner::shell running command cmd=git, args=["rev-parse", "HEAD"], cwd="/workspace/datafusion-branch"
2026-09-09T03:49:23.180677Z INFO benchmark_controller::runner::shell running command cmd=git, args=["rev-parse", "HEAD"], cwd="/workspace/datafusion-base"
2026-09-09T03:49:28.188510Z INFO benchmark_controller::runner::controller_client post_comment _repo=apache/datafusion, _pr_number=25093, job_id=2233
2026-09-09T03:49:28.916404Z INFO benchmark_controller::runner::bench_datafusion === Waiting for builds ===
2026-09-09T04:00:32.710084Z INFO benchmark_controller::runner::bench_datafusion === Builds complete ===
2026-09-09T04:00:32.710097Z INFO benchmark_controller::runner::bench_datafusion === Setting up bench runner ===
2026-09-09T04:00:32.710106Z INFO benchmark_controller::runner::shell running command cmd=git, args=["clone", "--depth=200", "https://github.com/apache/datafusion.git", "/workspace/datafusion-bench"], cwd="/"
2026-09-09T04:00:37.711697Z INFO benchmark_controller::runner::shell running command cmd=git, args=["-c", "advice.detachedHead=false", "checkout", "origin/main"], cwd="/workspace/datafusion-bench"
2026-09-09T04:00:42.714049Z INFO benchmark_controller::runner::shell running command cmd=cp, args=["-r", "/data/tpch-answers/.", "/workspace/datafusion-bench/benchmarks/data/tpch_sf1/answers"], cwd="/"
2026-09-09T04:00:47.716848Z INFO benchmark_controller::runner::shell running command cmd=cp, args=["-r", "/data/tpch-answers/.", "/workspace/datafusion-bench/benchmarks/data/tpch_sf10/answers"], cwd="/"
2026-09-09T04:00:52.718316Z INFO benchmark_controller::runner::bench_datafusion ** Creating data if needed for push_down_topk **
2026-09-09T04:00:52.718930Z INFO benchmark_controller::runner::shell running command cmd=/scripts/cache_data.sh, args=["push_down_topk", "/workspace/datafusion-bench/benchmarks"], cwd="/workspace/datafusion-bench/benchmarks"
2026-09-09T04:01:02.722034Z INFO benchmark_controller::runner::bench_datafusion ** Running push_down_topk baseline **
2026-09-09T04:01:02.722132Z INFO benchmark_controller::runner::shell running command (monitored) cmd=env, args=["DATAFUSION_DIR=/workspace/datafusion-base", "RESULTS_NAME=HEAD", "DATAFUSION_RUNTIME_TEMP_DIRECTORY=/workspace/spill-base-push_down_topk", "SQL_CARGO_COMMAND=cargo bench --bench sql -- --save-baseline HEAD", "./bench.sh", "run", "push_down_topk"], cwd="/workspace/datafusion-bench/benchmarks"
```
</details>
<details><summary>Kubernetes message</summary>
```
Job has reached the specified backoff limit
```
</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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangb (via GitHub)" <gi...@apache.org>.
adriangb commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595682675
run benchmark push_down_topk
env:
CARGO_PROFILE_BENCH_LTO: "thin"
--
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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595493830
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595469801)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5595469801-2236-xsbz8 6.12.94+ #1 SMP Fri Jul 17 09:42:57 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 pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpcds
```
</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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595850397
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595728605)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to 53d3a410e1e96b844ff6096e372344a2927c8faa [diff](https://github.com/apache/datafusion/compare/53d3a410e1e96b844ff6096e372344a2927c8faa..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpcds
baseline:
ref: "53d3a410e1e96b844ff6096e372344a2927c8faa"
```
</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 pr21621-bench-on-main
--------------------
Benchmark tpcds_sf1.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 5.56 ms β 5.58 ms β no change β
β QQuery 2 β 80.54 ms β 79.51 ms β no change β
β QQuery 3 β 28.71 ms β 28.48 ms β no change β
β QQuery 4 β 484.31 ms β 472.35 ms β no change β
β QQuery 5 β 51.51 ms β 51.79 ms β no change β
β QQuery 6 β 35.70 ms β 35.68 ms β no change β
β QQuery 7 β 74.68 ms β 74.41 ms β no change β
β QQuery 8 β 36.70 ms β 36.52 ms β no change β
β QQuery 9 β 52.09 ms β 51.76 ms β no change β
β QQuery 10 β 63.06 ms β 62.89 ms β no change β
β QQuery 11 β 301.66 ms β 294.73 ms β no change β
β QQuery 12 β 28.56 ms β 28.09 ms β no change β
β QQuery 13 β 117.62 ms β 117.20 ms β no change β
β QQuery 14 β 415.54 ms β 413.84 ms β no change β
β QQuery 15 β 57.96 ms β 57.19 ms β no change β
β QQuery 16 β 6.76 ms β 6.60 ms β no change β
β QQuery 17 β 79.76 ms β 79.63 ms β no change β
β QQuery 18 β 104.34 ms β 105.42 ms β no change β
β QQuery 19 β 40.94 ms β 40.82 ms β no change β
β QQuery 20 β 35.84 ms β 34.95 ms β no change β
β QQuery 21 β 17.28 ms β 16.98 ms β no change β
β QQuery 22 β 63.58 ms β 63.50 ms β no change β
β QQuery 23 β 311.18 ms β 307.77 ms β no change β
β QQuery 24 β 199.33 ms β 193.29 ms β no change β
β QQuery 25 β 116.58 ms β 107.96 ms β +1.08x faster β
β QQuery 26 β 51.41 ms β 48.24 ms β +1.07x faster β
β QQuery 27 β 6.75 ms β 5.91 ms β +1.14x faster β
β QQuery 28 β 59.39 ms β 55.46 ms β +1.07x faster β
β QQuery 29 β 101.48 ms β 96.60 ms β no change β
β QQuery 30 β 35.40 ms β 31.50 ms β +1.12x faster β
β QQuery 31 β 118.25 ms β 109.54 ms β +1.08x faster β
β QQuery 32 β 22.17 ms β 19.75 ms β +1.12x faster β
β QQuery 33 β 40.33 ms β 37.56 ms β +1.07x faster β
β QQuery 34 β 11.07 ms β 9.93 ms β +1.11x faster β
β QQuery 35 β 81.24 ms β 72.30 ms β +1.12x faster β
β QQuery 36 β 6.65 ms β 5.77 ms β +1.15x faster β
β QQuery 37 β 7.74 ms β 6.71 ms β +1.15x faster β
β QQuery 38 β 68.64 ms β 61.87 ms β +1.11x faster β
β QQuery 39 β 102.35 ms β 87.61 ms β +1.17x faster β
β QQuery 40 β 27.05 ms β 23.30 ms β +1.16x faster β
β QQuery 41 β 12.40 ms β 10.97 ms β +1.13x faster β
β QQuery 42 β 25.93 ms β 23.19 ms β +1.12x faster β
β QQuery 43 β 5.99 ms β 5.16 ms β +1.16x faster β
β QQuery 44 β 10.58 ms β 9.12 ms β +1.16x faster β
β QQuery 45 β 48.63 ms β 38.34 ms β +1.27x faster β
β QQuery 46 β 13.10 ms β 11.83 ms β +1.11x faster β
β QQuery 47 β 277.15 ms β 226.02 ms β +1.23x faster β
β QQuery 48 β 100.77 ms β 94.97 ms β +1.06x faster β
β QQuery 49 β 71.71 ms β 70.66 ms β no change β
β QQuery 50 β 60.32 ms β 58.81 ms β no change β
β QQuery 51 β 93.91 ms β 92.65 ms β no change β
β QQuery 52 β 24.30 ms β 23.71 ms β no change β
β QQuery 53 β 29.05 ms β 28.55 ms β no change β
β QQuery 54 β 54.31 ms β 53.57 ms β no change β
β QQuery 55 β 23.13 ms β 22.71 ms β no change β
β QQuery 56 β 38.50 ms β 37.93 ms β no change β
β QQuery 57 β 174.04 ms β 174.11 ms β no change β
β QQuery 58 β 111.12 ms β 118.31 ms β 1.06x slower β
β QQuery 59 β 117.62 ms β 122.45 ms β no change β
β QQuery 60 β 38.71 ms β 41.45 ms β 1.07x slower β
β QQuery 61 β 12.04 ms β 13.34 ms β 1.11x slower β
β QQuery 62 β 45.64 ms β 48.22 ms β 1.06x slower β
β QQuery 63 β 29.23 ms β 30.52 ms β no change β
β QQuery 64 β 369.02 ms β 401.43 ms β 1.09x slower β
β QQuery 65 β 124.88 ms β 132.95 ms β 1.06x slower β
β QQuery 66 β 85.79 ms β 86.31 ms β no change β
β QQuery 67 β 242.33 ms β 285.33 ms β 1.18x slower β
β QQuery 68 β 11.94 ms β 12.89 ms β 1.08x slower β
β QQuery 69 β 57.92 ms β 60.24 ms β no change β
β QQuery 70 β 108.33 ms β 111.67 ms β no change β
β QQuery 71 β 35.97 ms β 34.88 ms β no change β
β QQuery 72 β 1830.39 ms β 1934.21 ms β 1.06x slower β
β QQuery 73 β 10.34 ms β 10.80 ms β no change β
β QQuery 74 β 169.43 ms β 171.23 ms β no change β
β QQuery 75 β 145.27 ms β 145.39 ms β no change β
β QQuery 76 β 35.82 ms β 35.26 ms β no change β
β QQuery 77 β 60.62 ms β 60.50 ms β no change β
β QQuery 78 β 221.38 ms β 219.21 ms β no change β
β QQuery 79 β 66.67 ms β 66.67 ms β no change β
β QQuery 80 β 103.91 ms β 99.60 ms β no change β
β QQuery 81 β 27.60 ms β 26.47 ms β no change β
β QQuery 82 β 17.62 ms β 16.74 ms β no change β
β QQuery 83 β 35.59 ms β 34.13 ms β no change β
β QQuery 84 β 30.32 ms β 29.29 ms β no change β
β QQuery 85 β 106.01 ms β 103.34 ms β no change β
β QQuery 86 β 26.39 ms β 25.49 ms β no change β
β QQuery 87 β 61.20 ms β 61.97 ms β no change β
β QQuery 88 β 63.16 ms β 63.43 ms β no change β
β QQuery 89 β 35.33 ms β 36.12 ms β no change β
β QQuery 90 β 16.80 ms β 16.83 ms β no change β
β QQuery 91 β 44.86 ms β 44.53 ms β no change β
β QQuery 92 β 29.04 ms β 29.16 ms β no change β
β QQuery 93 β 49.82 ms β 49.64 ms β no change β
β QQuery 94 β 37.89 ms β 38.00 ms β no change β
β QQuery 95 β 79.86 ms β 80.71 ms β no change β
β QQuery 96 β 24.06 ms β 23.99 ms β no change β
β QQuery 97 β 52.93 ms β 54.10 ms β no change β
β QQuery 98 β 42.31 ms β 42.38 ms β no change β
β QQuery 99 β 69.60 ms β 69.79 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββ
β Benchmark Summary β β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 9498.30ms β
β Total Time (pr21621-bench-on-main) β 9510.27ms β
β Average Time (HEAD) β 95.94ms β
β Average Time (pr21621-bench-on-main) β 96.06ms β
β Queries Faster β 23 β
β Queries Slower β 9 β
β Queries with No Change β 67 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and pr21621-bench-on-main
--------------------
Benchmark tpcds_sf1.json
--------------------
βββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 5.56 / 6.13 Β±0.99 / 8.10 ms β 5.58 / 6.09 Β±0.94 / 7.98 ms β no change β
β QQuery 2 β 80.54 / 80.93 Β±0.26 / 81.26 ms β 79.51 / 80.06 Β±0.38 / 80.70 ms β no change β
β QQuery 3 β 28.71 / 29.03 Β±0.31 / 29.59 ms β 28.48 / 28.80 Β±0.34 / 29.33 ms β no change β
β QQuery 4 β 484.31 / 492.37 Β±4.68 / 497.58 ms β 472.35 / 480.53 Β±5.67 / 487.19 ms β no change β
β QQuery 5 β 51.51 / 52.55 Β±0.79 / 53.89 ms β 51.79 / 52.47 Β±0.47 / 52.98 ms β no change β
β QQuery 6 β 35.70 / 36.58 Β±0.44 / 36.83 ms β 35.68 / 36.56 Β±0.46 / 36.89 ms β no change β
β QQuery 7 β 74.68 / 75.43 Β±0.59 / 76.27 ms β 74.41 / 75.29 Β±0.56 / 75.84 ms β no change β
β QQuery 8 β 36.70 / 38.07 Β±2.18 / 42.38 ms β 36.52 / 36.67 Β±0.17 / 36.96 ms β no change β
β QQuery 9 β 52.09 / 54.81 Β±2.23 / 58.89 ms β 51.76 / 53.44 Β±1.33 / 55.43 ms β no change β
β QQuery 10 β 63.06 / 63.42 Β±0.19 / 63.58 ms β 62.89 / 63.32 Β±0.26 / 63.61 ms β no change β
β QQuery 11 β 301.66 / 306.74 Β±3.63 / 311.53 ms β 294.73 / 299.99 Β±4.55 / 305.92 ms β no change β
β QQuery 12 β 28.56 / 28.99 Β±0.27 / 29.33 ms β 28.09 / 28.83 Β±0.50 / 29.65 ms β no change β
β QQuery 13 β 117.62 / 118.54 Β±0.71 / 119.77 ms β 117.20 / 118.95 Β±1.87 / 122.40 ms β no change β
β QQuery 14 β 415.54 / 418.47 Β±1.99 / 421.45 ms β 413.84 / 429.48 Β±11.78 / 447.95 ms β no change β
β QQuery 15 β 57.96 / 60.14 Β±1.68 / 63.10 ms β 57.19 / 58.12 Β±0.71 / 59.00 ms β no change β
β QQuery 16 β 6.76 / 6.93 Β±0.22 / 7.37 ms β 6.60 / 6.74 Β±0.22 / 7.18 ms β no change β
β QQuery 17 β 79.76 / 81.22 Β±1.60 / 84.11 ms β 79.63 / 80.44 Β±1.21 / 82.84 ms β no change β
β QQuery 18 β 104.34 / 105.29 Β±0.72 / 106.47 ms β 105.42 / 106.27 Β±0.44 / 106.62 ms β no change β
β QQuery 19 β 40.94 / 41.67 Β±0.56 / 42.55 ms β 40.82 / 41.47 Β±0.39 / 41.95 ms β no change β
β QQuery 20 β 35.84 / 36.54 Β±0.86 / 38.24 ms β 34.95 / 35.31 Β±0.22 / 35.55 ms β no change β
β QQuery 21 β 17.28 / 17.55 Β±0.23 / 17.90 ms β 16.98 / 17.18 Β±0.14 / 17.35 ms β no change β
β QQuery 22 β 63.58 / 65.28 Β±1.39 / 67.78 ms β 63.50 / 63.88 Β±0.42 / 64.69 ms β no change β
β QQuery 23 β 311.18 / 313.13 Β±1.10 / 314.23 ms β 307.77 / 310.48 Β±3.63 / 317.46 ms β no change β
β QQuery 24 β 199.33 / 209.71 Β±11.75 / 230.53 ms β 193.29 / 197.78 Β±4.29 / 205.84 ms β +1.06x faster β
β QQuery 25 β 116.58 / 118.84 Β±1.99 / 122.25 ms β 107.96 / 111.88 Β±3.33 / 117.88 ms β +1.06x faster β
β QQuery 26 β 51.41 / 52.03 Β±0.46 / 52.60 ms β 48.24 / 48.50 Β±0.20 / 48.75 ms β +1.07x faster β
β QQuery 27 β 6.75 / 6.97 Β±0.21 / 7.33 ms β 5.91 / 6.05 Β±0.21 / 6.47 ms β +1.15x faster β
β QQuery 28 β 59.39 / 62.35 Β±2.63 / 66.27 ms β 55.46 / 59.54 Β±2.04 / 60.80 ms β no change β
β QQuery 29 β 101.48 / 103.17 Β±1.04 / 104.58 ms β 96.60 / 100.14 Β±4.12 / 108.05 ms β no change β
β QQuery 30 β 35.40 / 35.65 Β±0.19 / 35.97 ms β 31.50 / 32.06 Β±0.38 / 32.70 ms β +1.11x faster β
β QQuery 31 β 118.25 / 119.58 Β±1.04 / 121.46 ms β 109.54 / 111.03 Β±1.65 / 113.99 ms β +1.08x faster β
β QQuery 32 β 22.17 / 22.36 Β±0.16 / 22.63 ms β 19.75 / 20.14 Β±0.37 / 20.75 ms β +1.11x faster β
β QQuery 33 β 40.33 / 40.84 Β±0.46 / 41.65 ms β 37.56 / 37.92 Β±0.31 / 38.35 ms β +1.08x faster β
β QQuery 34 β 11.07 / 11.30 Β±0.16 / 11.49 ms β 9.93 / 10.20 Β±0.23 / 10.49 ms β +1.11x faster β
β QQuery 35 β 81.24 / 83.33 Β±1.53 / 85.92 ms β 72.30 / 73.19 Β±0.56 / 73.93 ms β +1.14x faster β
β QQuery 36 β 6.65 / 6.79 Β±0.15 / 7.05 ms β 5.77 / 5.91 Β±0.19 / 6.27 ms β +1.15x faster β
β QQuery 37 β 7.74 / 8.49 Β±1.30 / 11.08 ms β 6.71 / 6.88 Β±0.12 / 7.02 ms β +1.23x faster β
β QQuery 38 β 68.64 / 69.77 Β±1.26 / 72.08 ms β 61.87 / 63.06 Β±1.29 / 65.28 ms β +1.11x faster β
β QQuery 39 β 102.35 / 105.82 Β±4.27 / 114.03 ms β 87.61 / 89.20 Β±0.96 / 90.39 ms β +1.19x faster β
β QQuery 40 β 27.05 / 28.07 Β±0.90 / 29.47 ms β 23.30 / 23.79 Β±0.35 / 24.36 ms β +1.18x faster β
β QQuery 41 β 12.40 / 12.54 Β±0.12 / 12.73 ms β 10.97 / 11.13 Β±0.19 / 11.48 ms β +1.13x faster β
β QQuery 42 β 25.93 / 26.42 Β±0.46 / 27.11 ms β 23.19 / 23.51 Β±0.32 / 24.13 ms β +1.12x faster β
β QQuery 43 β 5.99 / 6.08 Β±0.06 / 6.15 ms β 5.16 / 6.35 Β±2.21 / 10.77 ms β no change β
β QQuery 44 β 10.58 / 10.63 Β±0.04 / 10.66 ms β 9.12 / 9.27 Β±0.15 / 9.53 ms β +1.15x faster β
β QQuery 45 β 48.63 / 49.58 Β±0.50 / 49.96 ms β 38.34 / 38.96 Β±0.42 / 39.57 ms β +1.27x faster β
β QQuery 46 β 13.10 / 13.61 Β±0.59 / 14.50 ms β 11.83 / 12.17 Β±0.23 / 12.51 ms β +1.12x faster β
β QQuery 47 β 277.15 / 283.37 Β±5.14 / 292.62 ms β 226.02 / 231.92 Β±4.82 / 240.41 ms β +1.22x faster β
β QQuery 48 β 100.77 / 101.17 Β±0.44 / 101.74 ms β 94.97 / 95.55 Β±0.50 / 96.42 ms β +1.06x faster β
β QQuery 49 β 71.71 / 74.43 Β±2.32 / 78.30 ms β 70.66 / 72.70 Β±3.32 / 79.32 ms β no change β
β QQuery 50 β 60.32 / 61.46 Β±1.66 / 64.74 ms β 58.81 / 60.12 Β±1.69 / 63.36 ms β no change β
β QQuery 51 β 93.91 / 96.33 Β±1.86 / 99.03 ms β 92.65 / 93.62 Β±0.50 / 93.98 ms β no change β
β QQuery 52 β 24.30 / 24.67 Β±0.35 / 25.35 ms β 23.71 / 24.39 Β±0.62 / 25.57 ms β no change β
β QQuery 53 β 29.05 / 29.35 Β±0.27 / 29.80 ms β 28.55 / 29.80 Β±1.37 / 32.38 ms β no change β
β QQuery 54 β 54.31 / 55.53 Β±1.23 / 57.84 ms β 53.57 / 54.00 Β±0.36 / 54.58 ms β no change β
β QQuery 55 β 23.13 / 23.38 Β±0.19 / 23.63 ms β 22.71 / 23.16 Β±0.46 / 24.03 ms β no change β
β QQuery 56 β 38.50 / 38.94 Β±0.23 / 39.16 ms β 37.93 / 38.90 Β±0.63 / 39.84 ms β no change β
β QQuery 57 β 174.04 / 176.30 Β±2.31 / 180.34 ms β 174.11 / 176.78 Β±2.00 / 180.15 ms β no change β
β QQuery 58 β 111.12 / 111.69 Β±0.66 / 112.89 ms β 118.31 / 120.53 Β±1.28 / 122.02 ms β 1.08x slower β
β QQuery 59 β 117.62 / 118.18 Β±0.47 / 118.99 ms β 122.45 / 123.25 Β±0.53 / 123.86 ms β no change β
β QQuery 60 β 38.71 / 39.59 Β±0.69 / 40.79 ms β 41.45 / 42.56 Β±0.59 / 43.16 ms β 1.07x slower β
β QQuery 61 β 12.04 / 12.24 Β±0.15 / 12.45 ms β 13.34 / 13.53 Β±0.14 / 13.73 ms β 1.11x slower β
β QQuery 62 β 45.64 / 45.82 Β±0.15 / 46.09 ms β 48.22 / 49.32 Β±1.18 / 51.58 ms β 1.08x slower β
β QQuery 63 β 29.23 / 29.52 Β±0.24 / 29.85 ms β 30.52 / 31.01 Β±0.28 / 31.32 ms β 1.05x slower β
β QQuery 64 β 369.02 / 372.12 Β±2.89 / 377.22 ms β 401.43 / 404.78 Β±2.33 / 407.34 ms β 1.09x slower β
β QQuery 65 β 124.88 / 126.98 Β±2.17 / 130.24 ms β 132.95 / 135.60 Β±2.16 / 139.20 ms β 1.07x slower β
β QQuery 66 β 85.79 / 87.49 Β±2.27 / 91.82 ms β 86.31 / 87.24 Β±0.87 / 88.84 ms β no change β
β QQuery 67 β 242.33 / 263.16 Β±15.90 / 282.68 ms β 285.33 / 291.64 Β±5.37 / 300.17 ms β 1.11x slower β
β QQuery 68 β 11.94 / 12.23 Β±0.26 / 12.65 ms β 12.89 / 13.29 Β±0.28 / 13.69 ms β 1.09x slower β
β QQuery 69 β 57.92 / 59.44 Β±2.29 / 63.97 ms β 60.24 / 61.87 Β±2.26 / 66.29 ms β no change β
β QQuery 70 β 108.33 / 110.56 Β±1.54 / 113.04 ms β 111.67 / 116.29 Β±6.22 / 128.47 ms β 1.05x slower β
β QQuery 71 β 35.97 / 36.41 Β±0.31 / 36.81 ms β 34.88 / 36.31 Β±1.26 / 38.53 ms β no change β
β QQuery 72 β 1830.39 / 1893.64 Β±42.80 / 1951.24 ms β 1934.21 / 1989.19 Β±31.40 / 2028.01 ms β 1.05x slower β
β QQuery 73 β 10.34 / 10.72 Β±0.36 / 11.35 ms β 10.80 / 12.24 Β±1.68 / 15.42 ms β 1.14x slower β
β QQuery 74 β 169.43 / 184.24 Β±13.66 / 201.27 ms β 171.23 / 176.27 Β±6.33 / 187.52 ms β no change β
β QQuery 75 β 145.27 / 148.21 Β±2.22 / 152.04 ms β 145.39 / 146.28 Β±1.22 / 148.68 ms β no change β
β QQuery 76 β 35.82 / 36.50 Β±0.55 / 37.05 ms β 35.26 / 37.05 Β±2.78 / 42.59 ms β no change β
β QQuery 77 β 60.62 / 61.29 Β±0.49 / 61.91 ms β 60.50 / 62.86 Β±1.64 / 65.37 ms β no change β
β QQuery 78 β 221.38 / 223.98 Β±2.72 / 227.92 ms β 219.21 / 226.19 Β±4.52 / 232.85 ms β no change β
β QQuery 79 β 66.67 / 67.27 Β±0.86 / 68.92 ms β 66.67 / 67.67 Β±1.07 / 69.74 ms β no change β
β QQuery 80 β 103.91 / 106.70 Β±3.70 / 113.98 ms β 99.60 / 102.10 Β±1.87 / 104.46 ms β no change β
β QQuery 81 β 27.60 / 27.77 Β±0.14 / 27.97 ms β 26.47 / 26.60 Β±0.14 / 26.84 ms β no change β
β QQuery 82 β 17.62 / 17.77 Β±0.16 / 18.05 ms β 16.74 / 16.86 Β±0.15 / 17.14 ms β +1.05x faster β
β QQuery 83 β 35.59 / 35.86 Β±0.25 / 36.21 ms β 34.13 / 34.51 Β±0.22 / 34.79 ms β no change β
β QQuery 84 β 30.32 / 32.64 Β±3.34 / 39.27 ms β 29.29 / 29.46 Β±0.15 / 29.65 ms β +1.11x faster β
β QQuery 85 β 106.01 / 107.19 Β±1.08 / 108.44 ms β 103.34 / 107.24 Β±4.08 / 114.91 ms β no change β
β QQuery 86 β 26.39 / 26.73 Β±0.39 / 27.47 ms β 25.49 / 25.78 Β±0.30 / 26.35 ms β no change β
β QQuery 87 β 61.20 / 62.88 Β±2.31 / 67.32 ms β 61.97 / 62.62 Β±0.57 / 63.28 ms β no change β
β QQuery 88 β 63.16 / 63.73 Β±0.35 / 64.11 ms β 63.43 / 65.97 Β±4.65 / 75.26 ms β no change β
β QQuery 89 β 35.33 / 36.18 Β±0.80 / 37.60 ms β 36.12 / 36.96 Β±0.81 / 38.29 ms β no change β
β QQuery 90 β 16.80 / 17.18 Β±0.26 / 17.53 ms β 16.83 / 17.08 Β±0.27 / 17.58 ms β no change β
β QQuery 91 β 44.86 / 45.08 Β±0.17 / 45.30 ms β 44.53 / 45.15 Β±0.42 / 45.66 ms β no change β
β QQuery 92 β 29.04 / 30.78 Β±2.77 / 36.29 ms β 29.16 / 29.71 Β±0.40 / 30.40 ms β no change β
β QQuery 93 β 49.82 / 51.06 Β±0.99 / 52.73 ms β 49.64 / 51.36 Β±1.49 / 54.06 ms β no change β
β QQuery 94 β 37.89 / 38.67 Β±0.93 / 40.17 ms β 38.00 / 39.21 Β±0.97 / 40.88 ms β no change β
β QQuery 95 β 79.86 / 81.09 Β±0.83 / 82.12 ms β 80.71 / 81.90 Β±1.28 / 83.70 ms β no change β
β QQuery 96 β 24.06 / 24.25 Β±0.21 / 24.64 ms β 23.99 / 24.42 Β±0.25 / 24.72 ms β no change β
β QQuery 97 β 52.93 / 54.91 Β±2.94 / 60.75 ms β 54.10 / 55.31 Β±1.88 / 59.04 ms β no change β
β QQuery 98 β 42.31 / 43.22 Β±0.62 / 44.20 ms β 42.38 / 43.18 Β±0.63 / 43.99 ms β no change β
β QQuery 99 β 69.60 / 70.08 Β±0.33 / 70.51 ms β 69.79 / 70.05 Β±0.32 / 70.67 ms β no change β
βββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββ
β Benchmark Summary β β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 9723.73ms β
β Total Time (pr21621-bench-on-main) β 9720.44ms β
β Average Time (HEAD) β 98.22ms β
β Average Time (pr21621-bench-on-main) β 98.19ms β
β Queries Faster β 24 β
β Queries Slower β 12 β
β Queries with No Change β 63 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 50.0s |
| Peak memory | 2.1 GiB |
| Avg memory | 1.5 GiB |
| CPU user | 205.9s |
| CPU sys | 5.8s |
| Peak spill | 0 B |
**tpcds β branch**
| Metric | Value |
|--------|-------|
| Wall time | 50.0s |
| Peak memory | 2.0 GiB |
| Avg memory | 1.3 GiB |
| CPU user | 208.5s |
| CPU sys | 5.9s |
| 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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595494020
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595470052)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5595470052-2232-8zplq 6.12.94+ #1 SMP Fri Jul 17 09:42:57 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 pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark sql_planner
```
</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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangb (via GitHub)" <gi...@apache.org>.
adriangb commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595469940
run benchmark push_down_topk topk_tpch
--
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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595693265
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595682675)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5595682675-2239-xlw2r 6.12.94+ #1 SMP Fri Jul 17 09:42:57 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 pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark push_down_topk
env:
CARGO_PROFILE_BENCH_LTO: "thin"
```
</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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5596177907
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595470052)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark sql_planner
```
</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>
```
group HEAD pr21621-bench-on-main
----- ---- ---------------------
logical_aggregate_with_join 1.00 396.1Β±2.18Β΅s ? ?/sec 1.00 397.4Β±1.07Β΅s ? ?/sec
logical_correlated_subquery_exists 1.00 238.1Β±0.83Β΅s ? ?/sec 1.01 239.4Β±1.04Β΅s ? ?/sec
logical_correlated_subquery_in 1.00 240.6Β±1.14Β΅s ? ?/sec 1.00 241.2Β±1.01Β΅s ? ?/sec
logical_distinct_many_columns 1.00 528.0Β±1.04Β΅s ? ?/sec 1.00 528.8Β±2.53Β΅s ? ?/sec
logical_join_4_with_agg_and_filter 1.00 200.1Β±0.54Β΅s ? ?/sec 1.00 200.8Β±0.90Β΅s ? ?/sec
logical_join_8_with_agg_sort_limit 1.01 371.9Β±1.56Β΅s ? ?/sec 1.00 369.9Β±1.54Β΅s ? ?/sec
logical_join_chain_16 1.00 620.4Β±3.96Β΅s ? ?/sec 1.00 622.1Β±3.86Β΅s ? ?/sec
logical_join_chain_4 1.00 79.6Β±0.45Β΅s ? ?/sec 1.01 80.3Β±0.33Β΅s ? ?/sec
logical_join_chain_8 1.02 208.4Β±1.04Β΅s ? ?/sec 1.00 204.1Β±0.98Β΅s ? ?/sec
logical_multiple_subqueries 1.00 464.3Β±1.36Β΅s ? ?/sec 1.00 462.3Β±3.53Β΅s ? ?/sec
logical_nested_cte_4_levels 1.00 210.2Β±1.19Β΅s ? ?/sec 1.01 211.6Β±1.24Β΅s ? ?/sec
logical_plan_struct_join_agg_sort 1.03 130.0Β±0.52Β΅s ? ?/sec 1.00 126.7Β±0.45Β΅s ? ?/sec
logical_plan_tpcds_all 1.00 88.2Β±0.17ms ? ?/sec 1.00 88.4Β±0.13ms ? ?/sec
logical_plan_tpch_all 1.01 5.5Β±0.01ms ? ?/sec 1.00 5.4Β±0.02ms ? ?/sec
logical_scalar_subquery 1.00 260.2Β±0.76Β΅s ? ?/sec 1.01 263.3Β±1.81Β΅s ? ?/sec
logical_select_all_from_1000 1.00 104.6Β±0.18ms ? ?/sec 1.01 105.4Β±0.39ms ? ?/sec
logical_select_one_from_700 1.02 284.7Β±1.41Β΅s ? ?/sec 1.00 280.2Β±1.69Β΅s ? ?/sec
logical_trivial_join_high_numbered_columns 1.00 238.6Β±0.53Β΅s ? ?/sec 1.00 238.2Β±0.61Β΅s ? ?/sec
logical_trivial_join_low_numbered_columns 1.00 225.6Β±0.55Β΅s ? ?/sec 1.00 225.9Β±0.72Β΅s ? ?/sec
logical_union_4_branches 1.00 371.9Β±1.62Β΅s ? ?/sec 1.00 372.8Β±2.43Β΅s ? ?/sec
logical_union_8_branches 1.00 750.4Β±1.43Β΅s ? ?/sec 1.00 746.7Β±4.41Β΅s ? ?/sec
logical_wide_aggregate_100_exprs 1.00 4.5Β±0.01ms ? ?/sec 1.00 4.5Β±0.01ms ? ?/sec
logical_wide_case_50_exprs 1.00 2.3Β±0.00ms ? ?/sec 1.00 2.3Β±0.00ms ? ?/sec
logical_wide_filter_200_predicates 1.00 1267.9Β±7.11Β΅s ? ?/sec 1.00 1270.7Β±9.45Β΅s ? ?/sec
logical_wide_filter_50_predicates 1.00 343.0Β±2.02Β΅s ? ?/sec 1.00 342.5Β±2.18Β΅s ? ?/sec
optimizer_correlated_exists 1.01 219.7Β±0.58Β΅s ? ?/sec 1.00 217.7Β±0.49Β΅s ? ?/sec
optimizer_join_4_with_agg_filter 1.04 437.8Β±1.89Β΅s ? ?/sec 1.00 420.6Β±1.07Β΅s ? ?/sec
optimizer_join_chain_4 1.02 159.3Β±0.30Β΅s ? ?/sec 1.00 156.1Β±0.20Β΅s ? ?/sec
optimizer_join_chain_8 1.01 542.2Β±0.68Β΅s ? ?/sec 1.00 535.4Β±0.81Β΅s ? ?/sec
optimizer_select_all_from_1000 1.00 6.7Β±0.01ms ? ?/sec 1.00 6.7Β±0.01ms ? ?/sec
optimizer_select_one_from_700 1.00 228.6Β±0.91Β΅s ? ?/sec 1.00 228.3Β±0.46Β΅s ? ?/sec
optimizer_tpcds_all 1.01 303.9Β±0.37ms ? ?/sec 1.00 301.6Β±0.50ms ? ?/sec
optimizer_tpch_all 1.02 16.6Β±0.03ms ? ?/sec 1.00 16.3Β±0.03ms ? ?/sec
optimizer_wide_aggregate_100 1.00 2.2Β±0.01ms ? ?/sec 1.00 2.2Β±0.00ms ? ?/sec
optimizer_wide_filter_200 1.00 3.6Β±0.01ms ? ?/sec 1.00 3.6Β±0.01ms ? ?/sec
physical_intersection 1.01 540.4Β±3.18Β΅s ? ?/sec 1.00 537.5Β±1.87Β΅s ? ?/sec
physical_join_consider_sort 1.00 981.5Β±1.52Β΅s ? ?/sec 1.00 981.5Β±2.16Β΅s ? ?/sec
physical_join_distinct 1.00 218.9Β±0.58Β΅s ? ?/sec 1.00 218.5Β±0.78Β΅s ? ?/sec
physical_many_self_joins 1.00 7.5Β±0.02ms ? ?/sec 1.01 7.5Β±0.02ms ? ?/sec
physical_plan_clickbench_all 1.00 125.4Β±0.28ms ? ?/sec 1.01 127.0Β±0.37ms ? ?/sec
physical_plan_clickbench_q1 1.00 1340.1Β±5.28Β΅s ? ?/sec 1.02 1360.8Β±6.09Β΅s ? ?/sec
physical_plan_clickbench_q10 1.00 2.1Β±0.01ms ? ?/sec 1.00 2.1Β±0.01ms ? ?/sec
physical_plan_clickbench_q11 1.01 2.2Β±0.01ms ? ?/sec 1.00 2.2Β±0.01ms ? ?/sec
physical_plan_clickbench_q12 1.01 2.3Β±0.01ms ? ?/sec 1.00 2.3Β±0.01ms ? ?/sec
physical_plan_clickbench_q13 1.00 2.1Β±0.01ms ? ?/sec 1.00 2.0Β±0.01ms ? ?/sec
physical_plan_clickbench_q14 1.00 2.2Β±0.00ms ? ?/sec 1.00 2.2Β±0.01ms ? ?/sec
physical_plan_clickbench_q15 1.00 2.1Β±0.01ms ? ?/sec 1.00 2.1Β±0.01ms ? ?/sec
physical_plan_clickbench_q16 1.00 1778.1Β±13.71Β΅s ? ?/sec 1.01 1803.3Β±6.44Β΅s ? ?/sec
physical_plan_clickbench_q17 1.00 1818.5Β±5.17Β΅s ? ?/sec 1.02 1853.4Β±6.35Β΅s ? ?/sec
physical_plan_clickbench_q18 1.00 1626.7Β±5.27Β΅s ? ?/sec 1.02 1658.2Β±5.07Β΅s ? ?/sec
physical_plan_clickbench_q19 1.00 2.0Β±0.00ms ? ?/sec 1.02 2.1Β±0.01ms ? ?/sec
physical_plan_clickbench_q2 1.00 1745.2Β±5.57Β΅s ? ?/sec 1.00 1751.5Β±6.43Β΅s ? ?/sec
physical_plan_clickbench_q20 1.00 1483.4Β±5.01Β΅s ? ?/sec 1.03 1526.6Β±8.87Β΅s ? ?/sec
physical_plan_clickbench_q21 1.00 1725.5Β±5.33Β΅s ? ?/sec 1.02 1751.5Β±5.90Β΅s ? ?/sec
physical_plan_clickbench_q22 1.00 2.1Β±0.01ms ? ?/sec 1.02 2.2Β±0.02ms ? ?/sec
physical_plan_clickbench_q23 1.00 2.3Β±0.01ms ? ?/sec 1.01 2.3Β±0.01ms ? ?/sec
physical_plan_clickbench_q24 1.00 6.5Β±0.01ms ? ?/sec 1.00 6.5Β±0.01ms ? ?/sec
physical_plan_clickbench_q25 1.00 1859.8Β±5.51Β΅s ? ?/sec 1.01 1884.3Β±8.03Β΅s ? ?/sec
physical_plan_clickbench_q26 1.01 1728.7Β±9.13Β΅s ? ?/sec 1.00 1716.1Β±6.97Β΅s ? ?/sec
physical_plan_clickbench_q27 1.01 1927.6Β±14.33Β΅s ? ?/sec 1.00 1904.4Β±6.33Β΅s ? ?/sec
physical_plan_clickbench_q28 1.02 2.3Β±0.02ms ? ?/sec 1.00 2.3Β±0.01ms ? ?/sec
physical_plan_clickbench_q29 1.00 2.5Β±0.02ms ? ?/sec 1.00 2.5Β±0.01ms ? ?/sec
physical_plan_clickbench_q3 1.00 1636.7Β±6.75Β΅s ? ?/sec 1.01 1660.9Β±10.21Β΅s ? ?/sec
physical_plan_clickbench_q30 1.00 15.0Β±0.03ms ? ?/sec 1.01 15.2Β±0.05ms ? ?/sec
physical_plan_clickbench_q31 1.01 2.5Β±0.01ms ? ?/sec 1.00 2.4Β±0.01ms ? ?/sec
physical_plan_clickbench_q32 1.01 2.5Β±0.01ms ? ?/sec 1.00 2.4Β±0.01ms ? ?/sec
physical_plan_clickbench_q33 1.01 2.0Β±0.01ms ? ?/sec 1.00 2.0Β±0.01ms ? ?/sec
physical_plan_clickbench_q34 1.01 1802.4Β±5.19Β΅s ? ?/sec 1.00 1784.5Β±5.62Β΅s ? ?/sec
physical_plan_clickbench_q35 1.01 1850.1Β±6.49Β΅s ? ?/sec 1.00 1830.8Β±4.99Β΅s ? ?/sec
physical_plan_clickbench_q36 1.01 2.1Β±0.01ms ? ?/sec 1.00 2.1Β±0.01ms ? ?/sec
physical_plan_clickbench_q37 1.01 2.5Β±0.01ms ? ?/sec 1.00 2.5Β±0.01ms ? ?/sec
physical_plan_clickbench_q38 1.01 2.5Β±0.01ms ? ?/sec 1.00 2.5Β±0.01ms ? ?/sec
physical_plan_clickbench_q39 1.01 2.6Β±0.00ms ? ?/sec 1.00 2.6Β±0.01ms ? ?/sec
physical_plan_clickbench_q4 1.00 1455.7Β±5.65Β΅s ? ?/sec 1.00 1450.4Β±5.74Β΅s ? ?/sec
physical_plan_clickbench_q40 1.00 3.3Β±0.01ms ? ?/sec 1.01 3.3Β±0.01ms ? ?/sec
physical_plan_clickbench_q41 1.00 2.8Β±0.01ms ? ?/sec 1.00 2.8Β±0.01ms ? ?/sec
physical_plan_clickbench_q42 1.00 3.0Β±0.01ms ? ?/sec 1.00 3.0Β±0.01ms ? ?/sec
physical_plan_clickbench_q43 1.00 3.1Β±0.01ms ? ?/sec 1.00 3.2Β±0.01ms ? ?/sec
physical_plan_clickbench_q44 1.00 1527.8Β±5.30Β΅s ? ?/sec 1.02 1560.7Β±6.02Β΅s ? ?/sec
physical_plan_clickbench_q45 1.00 1552.0Β±4.66Β΅s ? ?/sec 1.01 1564.6Β±5.09Β΅s ? ?/sec
physical_plan_clickbench_q46 1.00 1868.8Β±6.29Β΅s ? ?/sec 1.01 1885.0Β±5.25Β΅s ? ?/sec
physical_plan_clickbench_q47 1.01 2.5Β±0.01ms ? ?/sec 1.00 2.5Β±0.01ms ? ?/sec
physical_plan_clickbench_q48 1.01 2.7Β±0.01ms ? ?/sec 1.00 2.7Β±0.01ms ? ?/sec
physical_plan_clickbench_q49 1.00 2.8Β±0.01ms ? ?/sec 1.00 2.8Β±0.01ms ? ?/sec
physical_plan_clickbench_q5 1.00 1594.6Β±5.36Β΅s ? ?/sec 1.00 1591.9Β±8.70Β΅s ? ?/sec
physical_plan_clickbench_q50 1.01 2.6Β±0.01ms ? ?/sec 1.00 2.6Β±0.01ms ? ?/sec
physical_plan_clickbench_q51 1.00 1974.5Β±5.92Β΅s ? ?/sec 1.00 1971.3Β±6.16Β΅s ? ?/sec
physical_plan_clickbench_q6 1.00 1600.7Β±5.64Β΅s ? ?/sec 1.00 1595.5Β±6.68Β΅s ? ?/sec
physical_plan_clickbench_q7 1.01 1412.1Β±4.47Β΅s ? ?/sec 1.00 1402.5Β±4.96Β΅s ? ?/sec
physical_plan_clickbench_q8 1.00 1922.5Β±6.05Β΅s ? ?/sec 1.00 1929.5Β±10.12Β΅s ? ?/sec
physical_plan_clickbench_q9 1.00 1941.6Β±6.27Β΅s ? ?/sec 1.00 1943.3Β±7.84Β΅s ? ?/sec
physical_plan_struct_join_agg_sort 1.01 1276.1Β±5.64Β΅s ? ?/sec 1.00 1261.0Β±3.38Β΅s ? ?/sec
physical_plan_tpcds_all 1.00 710.7Β±2.85ms ? ?/sec 1.00 709.6Β±5.86ms ? ?/sec
physical_plan_tpch_all 1.00 43.9Β±0.08ms ? ?/sec 1.00 43.9Β±0.15ms ? ?/sec
physical_plan_tpch_q1 1.01 1531.0Β±5.21Β΅s ? ?/sec 1.00 1517.2Β±6.55Β΅s ? ?/sec
physical_plan_tpch_q10 1.01 2.8Β±0.00ms ? ?/sec 1.00 2.7Β±0.01ms ? ?/sec
physical_plan_tpch_q11 1.01 2.2Β±0.00ms ? ?/sec 1.00 2.2Β±0.00ms ? ?/sec
physical_plan_tpch_q12 1.02 1215.6Β±2.63Β΅s ? ?/sec 1.00 1194.3Β±2.73Β΅s ? ?/sec
physical_plan_tpch_q13 1.01 1009.8Β±3.36Β΅s ? ?/sec 1.00 1001.3Β±2.78Β΅s ? ?/sec
physical_plan_tpch_q14 1.01 1364.4Β±4.34Β΅s ? ?/sec 1.00 1348.8Β±4.11Β΅s ? ?/sec
physical_plan_tpch_q16 1.00 1588.9Β±2.90Β΅s ? ?/sec 1.00 1581.2Β±2.55Β΅s ? ?/sec
physical_plan_tpch_q17 1.01 1598.7Β±2.42Β΅s ? ?/sec 1.00 1583.7Β±3.37Β΅s ? ?/sec
physical_plan_tpch_q18 1.01 1940.4Β±2.62Β΅s ? ?/sec 1.00 1929.6Β±3.00Β΅s ? ?/sec
physical_plan_tpch_q19 1.01 1841.7Β±6.58Β΅s ? ?/sec 1.00 1822.9Β±3.79Β΅s ? ?/sec
physical_plan_tpch_q2 1.00 3.6Β±0.00ms ? ?/sec 1.00 3.6Β±0.01ms ? ?/sec
physical_plan_tpch_q20 1.01 2.1Β±0.00ms ? ?/sec 1.00 2.1Β±0.00ms ? ?/sec
physical_plan_tpch_q21 1.01 2.8Β±0.00ms ? ?/sec 1.00 2.8Β±0.00ms ? ?/sec
physical_plan_tpch_q22 1.02 1487.8Β±3.31Β΅s ? ?/sec 1.00 1459.7Β±3.90Β΅s ? ?/sec
physical_plan_tpch_q3 1.01 1850.8Β±2.52Β΅s ? ?/sec 1.00 1828.9Β±3.58Β΅s ? ?/sec
physical_plan_tpch_q4 1.02 1175.8Β±4.06Β΅s ? ?/sec 1.00 1154.3Β±3.41Β΅s ? ?/sec
physical_plan_tpch_q5 1.00 2.7Β±0.00ms ? ?/sec 1.00 2.7Β±0.02ms ? ?/sec
physical_plan_tpch_q6 1.03 578.0Β±2.71Β΅s ? ?/sec 1.00 562.6Β±1.80Β΅s ? ?/sec
physical_plan_tpch_q7 1.00 2.8Β±0.00ms ? ?/sec 1.00 2.8Β±0.01ms ? ?/sec
physical_plan_tpch_q8 1.00 3.8Β±0.00ms ? ?/sec 1.00 3.8Β±0.02ms ? ?/sec
physical_plan_tpch_q9 1.01 2.6Β±0.00ms ? ?/sec 1.00 2.6Β±0.01ms ? ?/sec
physical_select_aggregates_from_200 1.00 15.4Β±0.04ms ? ?/sec 1.01 15.5Β±0.02ms ? ?/sec
physical_select_all_from_1000 1.00 116.1Β±0.42ms ? ?/sec 1.01 116.9Β±0.40ms ? ?/sec
physical_select_one_from_700 1.01 714.3Β±3.92Β΅s ? ?/sec 1.00 706.5Β±3.21Β΅s ? ?/sec
physical_sorted_union_order_by_10_int64 1.00 4.1Β±0.00ms ? ?/sec 1.00 4.1Β±0.01ms ? ?/sec
physical_sorted_union_order_by_10_uint64 1.00 8.6Β±0.01ms ? ?/sec 1.01 8.7Β±0.04ms ? ?/sec
physical_sorted_union_order_by_50_int64 1.00 93.5Β±0.31ms ? ?/sec 1.02 95.5Β±0.30ms ? ?/sec
physical_sorted_union_order_by_50_uint64 1.00 356.8Β±0.70ms ? ?/sec 1.04 371.4Β±1.17ms ? ?/sec
physical_theta_join_consider_sort 1.00 1008.8Β±3.94Β΅s ? ?/sec 1.00 1006.4Β±3.74Β΅s ? ?/sec
physical_unnest_to_join 1.01 579.2Β±2.06Β΅s ? ?/sec 1.00 574.1Β±1.91Β΅s ? ?/sec
physical_window_function_partition_by_12_on_values 1.01 663.8Β±1.12Β΅s ? ?/sec 1.00 657.5Β±1.66Β΅s ? ?/sec
physical_window_function_partition_by_30_on_values 1.00 1344.7Β±1.93Β΅s ? ?/sec 1.01 1351.5Β±2.56Β΅s ? ?/sec
physical_window_function_partition_by_4_on_values 1.03 386.7Β±1.40Β΅s ? ?/sec 1.00 376.1Β±1.55Β΅s ? ?/sec
physical_window_function_partition_by_7_on_values 1.02 487.1Β±1.74Β΅s ? ?/sec 1.00 479.4Β±1.38Β΅s ? ?/sec
physical_window_function_partition_by_8_on_values 1.01 526.2Β±1.65Β΅s ? ?/sec 1.00 519.8Β±1.53Β΅s ? ?/sec
with_param_values_many_columns 1.00 432.4Β±1.74Β΅s ? ?/sec 1.00 433.3Β±2.09Β΅s ? ?/sec
```
</p>
</details>
<details><summary>Resource Usage</summary>
**sql_planner β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 2440.5s |
| Peak memory | 130.2 MiB |
| Avg memory | 64.0 MiB |
| CPU user | 1867.8s |
| CPU sys | 1.5s |
| Peak spill | 0 B |
**sql_planner β branch**
| Metric | Value |
|--------|-------|
| Wall time | 2445.5s |
| Peak memory | 131.1 MiB |
| Avg memory | 65.6 MiB |
| CPU user | 1874.2s |
| CPU sys | 1.4s |
| 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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595596385
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595469801)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpcds
```
</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 pr21621-bench-on-main
--------------------
Benchmark tpcds_sf1.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 5.93 ms β 5.72 ms β no change β
β QQuery 2 β 81.28 ms β 80.72 ms β no change β
β QQuery 3 β 28.46 ms β 29.21 ms β no change β
β QQuery 4 β 481.28 ms β 507.25 ms β 1.05x slower β
β QQuery 5 β 53.17 ms β 51.74 ms β no change β
β QQuery 6 β 38.15 ms β 37.54 ms β no change β
β QQuery 7 β 76.92 ms β 75.89 ms β no change β
β QQuery 8 β 38.03 ms β 37.79 ms β no change β
β QQuery 9 β 54.58 ms β 53.43 ms β no change β
β QQuery 10 β 63.01 ms β 65.44 ms β no change β
β QQuery 11 β 309.30 ms β 329.97 ms β 1.07x slower β
β QQuery 12 β 29.41 ms β 29.35 ms β no change β
β QQuery 13 β 119.01 ms β 119.43 ms β no change β
β QQuery 14 β 423.00 ms β 417.21 ms β no change β
β QQuery 15 β 60.17 ms β 58.69 ms β no change β
β QQuery 16 β 7.09 ms β 7.14 ms β no change β
β QQuery 17 β 80.06 ms β 81.35 ms β no change β
β QQuery 18 β 105.87 ms β 105.18 ms β no change β
β QQuery 19 β 41.91 ms β 41.75 ms β no change β
β QQuery 20 β 36.13 ms β 36.12 ms β no change β
β QQuery 21 β 17.84 ms β 17.86 ms β no change β
β QQuery 22 β 66.44 ms β 64.35 ms β no change β
β QQuery 23 β 319.07 ms β 313.65 ms β no change β
β QQuery 24 β 207.41 ms β 214.19 ms β no change β
β QQuery 25 β 110.73 ms β 110.93 ms β no change β
β QQuery 26 β 49.31 ms β 49.46 ms β no change β
β QQuery 27 β 6.18 ms β 6.44 ms β no change β
β QQuery 28 β 56.88 ms β 56.28 ms β no change β
β QQuery 29 β 98.96 ms β 97.56 ms β no change β
β QQuery 30 β 32.62 ms β 32.96 ms β no change β
β QQuery 31 β 111.22 ms β 112.47 ms β no change β
β QQuery 32 β 21.01 ms β 21.10 ms β no change β
β QQuery 33 β 39.00 ms β 38.17 ms β no change β
β QQuery 34 β 10.24 ms β 10.41 ms β no change β
β QQuery 35 β 72.80 ms β 76.50 ms β 1.05x slower β
β QQuery 36 β 5.98 ms β 6.19 ms β no change β
β QQuery 37 β 6.81 ms β 7.18 ms β 1.05x slower β
β QQuery 38 β 62.85 ms β 64.70 ms β no change β
β QQuery 39 β 90.57 ms β 91.91 ms β no change β
β QQuery 40 β 24.47 ms β 25.10 ms β no change β
β QQuery 41 β 11.46 ms β 11.71 ms β no change β
β QQuery 42 β 24.05 ms β 24.31 ms β no change β
β QQuery 43 β 5.52 ms β 5.48 ms β no change β
β QQuery 44 β 9.81 ms β 9.75 ms β no change β
β QQuery 45 β 39.35 ms β 41.02 ms β no change β
β QQuery 46 β 12.98 ms β 12.76 ms β no change β
β QQuery 47 β 234.49 ms β 245.02 ms β no change β
β QQuery 48 β 95.71 ms β 96.34 ms β no change β
β QQuery 49 β 72.02 ms β 71.81 ms β no change β
β QQuery 50 β 59.07 ms β 60.28 ms β no change β
β QQuery 51 β 93.23 ms β 93.01 ms β no change β
β QQuery 52 β 24.70 ms β 24.74 ms β no change β
β QQuery 53 β 29.31 ms β 29.59 ms β no change β
β QQuery 54 β 55.17 ms β 56.03 ms β no change β
β QQuery 55 β 23.39 ms β 23.40 ms β no change β
β QQuery 56 β 39.70 ms β 39.50 ms β no change β
β QQuery 57 β 176.40 ms β 179.19 ms β no change β
β QQuery 58 β 111.98 ms β 112.72 ms β no change β
β QQuery 59 β 120.65 ms β 118.54 ms β no change β
β QQuery 60 β 39.81 ms β 39.65 ms β no change β
β QQuery 61 β 12.71 ms β 12.41 ms β no change β
β QQuery 62 β 46.43 ms β 47.40 ms β no change β
β QQuery 63 β 29.87 ms β 29.73 ms β no change β
β QQuery 64 β 374.67 ms β 376.57 ms β no change β
β QQuery 65 β 127.19 ms β 124.92 ms β no change β
β QQuery 66 β 80.75 ms β 82.23 ms β no change β
β QQuery 67 β 254.68 ms β 251.86 ms β no change β
β QQuery 68 β 13.29 ms β 12.01 ms β +1.11x faster β
β QQuery 69 β 57.98 ms β 57.95 ms β no change β
β QQuery 70 β 109.89 ms β 106.56 ms β no change β
β QQuery 71 β 37.59 ms β 35.84 ms β no change β
β QQuery 72 β 1984.56 ms β 1939.05 ms β no change β
β QQuery 73 β 10.56 ms β 10.34 ms β no change β
β QQuery 74 β 184.24 ms β 182.03 ms β no change β
β QQuery 75 β 148.61 ms β 148.17 ms β no change β
β QQuery 76 β 35.95 ms β 35.58 ms β no change β
β QQuery 77 β 62.09 ms β 66.61 ms β 1.07x slower β
β QQuery 78 β 226.59 ms β 240.91 ms β 1.06x slower β
β QQuery 79 β 67.60 ms β 68.21 ms β no change β
β QQuery 80 β 100.03 ms β 109.23 ms β 1.09x slower β
β QQuery 81 β 26.11 ms β 28.70 ms β 1.10x slower β
β QQuery 82 β 16.86 ms β 17.84 ms β 1.06x slower β
β QQuery 83 β 34.57 ms β 34.64 ms β no change β
β QQuery 84 β 30.05 ms β 30.67 ms β no change β
β QQuery 85 β 104.95 ms β 104.88 ms β no change β
β QQuery 86 β 25.86 ms β 25.70 ms β no change β
β QQuery 87 β 63.22 ms β 63.64 ms β no change β
β QQuery 88 β 64.30 ms β 65.05 ms β no change β
β QQuery 89 β 36.70 ms β 35.72 ms β no change β
β QQuery 90 β 17.54 ms β 17.38 ms β no change β
β QQuery 91 β 45.60 ms β 45.25 ms β no change β
β QQuery 92 β 29.69 ms β 31.46 ms β 1.06x slower β
β QQuery 93 β 50.37 ms β 53.48 ms β 1.06x slower β
β QQuery 94 β 38.55 ms β 39.03 ms β no change β
β QQuery 95 β 81.51 ms β 82.37 ms β no change β
β QQuery 96 β 24.51 ms β 25.12 ms β no change β
β QQuery 97 β 52.77 ms β 53.29 ms β no change β
β QQuery 98 β 42.67 ms β 42.53 ms β no change β
β QQuery 99 β 70.76 ms β 69.83 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββ
β Benchmark Summary β β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 9643.82ms β
β Total Time (pr21621-bench-on-main) β 9687.39ms β
β Average Time (HEAD) β 97.41ms β
β Average Time (pr21621-bench-on-main) β 97.85ms β
β Queries Faster β 1 β
β Queries Slower β 11 β
β Queries with No Change β 87 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and pr21621-bench-on-main
--------------------
Benchmark tpcds_sf1.json
--------------------
βββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 5.93 / 6.47 Β±0.98 / 8.43 ms β 5.72 / 6.44 Β±0.93 / 8.22 ms β no change β
β QQuery 2 β 81.28 / 82.23 Β±0.70 / 83.03 ms β 80.72 / 81.14 Β±0.37 / 81.69 ms β no change β
β QQuery 3 β 28.46 / 29.04 Β±0.31 / 29.32 ms β 29.21 / 29.70 Β±0.48 / 30.34 ms β no change β
β QQuery 4 β 481.28 / 495.28 Β±8.96 / 504.91 ms β 507.25 / 515.62 Β±4.98 / 521.89 ms β no change β
β QQuery 5 β 53.17 / 53.56 Β±0.41 / 54.32 ms β 51.74 / 52.54 Β±0.59 / 53.28 ms β no change β
β QQuery 6 β 38.15 / 39.23 Β±0.82 / 40.42 ms β 37.54 / 38.62 Β±0.80 / 39.80 ms β no change β
β QQuery 7 β 76.92 / 77.13 Β±0.14 / 77.30 ms β 75.89 / 76.49 Β±0.47 / 77.19 ms β no change β
β QQuery 8 β 38.03 / 38.24 Β±0.23 / 38.66 ms β 37.79 / 39.94 Β±2.75 / 45.37 ms β no change β
β QQuery 9 β 54.58 / 57.03 Β±1.46 / 58.92 ms β 53.43 / 58.17 Β±3.76 / 64.69 ms β no change β
β QQuery 10 β 63.01 / 63.92 Β±0.65 / 64.61 ms β 65.44 / 65.55 Β±0.11 / 65.76 ms β no change β
β QQuery 11 β 309.30 / 325.34 Β±12.47 / 342.69 ms β 329.97 / 334.81 Β±4.04 / 340.30 ms β no change β
β QQuery 12 β 29.41 / 29.72 Β±0.23 / 30.06 ms β 29.35 / 29.85 Β±0.34 / 30.37 ms β no change β
β QQuery 13 β 119.01 / 120.11 Β±0.91 / 121.77 ms β 119.43 / 121.25 Β±1.68 / 123.46 ms β no change β
β QQuery 14 β 423.00 / 428.76 Β±6.79 / 441.91 ms β 417.21 / 424.37 Β±4.41 / 430.43 ms β no change β
β QQuery 15 β 60.17 / 60.88 Β±0.59 / 61.89 ms β 58.69 / 60.47 Β±1.13 / 62.21 ms β no change β
β QQuery 16 β 7.09 / 7.19 Β±0.11 / 7.37 ms β 7.14 / 7.25 Β±0.12 / 7.48 ms β no change β
β QQuery 17 β 80.06 / 81.07 Β±0.83 / 82.40 ms β 81.35 / 84.08 Β±1.74 / 86.61 ms β no change β
β QQuery 18 β 105.87 / 106.68 Β±0.68 / 107.82 ms β 105.18 / 106.43 Β±0.81 / 107.56 ms β no change β
β QQuery 19 β 41.91 / 43.71 Β±2.31 / 48.17 ms β 41.75 / 43.04 Β±1.21 / 45.28 ms β no change β
β QQuery 20 β 36.13 / 36.92 Β±0.60 / 37.84 ms β 36.12 / 37.44 Β±1.60 / 40.54 ms β no change β
β QQuery 21 β 17.84 / 17.94 Β±0.07 / 18.02 ms β 17.86 / 18.38 Β±0.58 / 19.50 ms β no change β
β QQuery 22 β 66.44 / 66.89 Β±0.46 / 67.64 ms β 64.35 / 66.19 Β±1.16 / 67.89 ms β no change β
β QQuery 23 β 319.07 / 326.36 Β±5.77 / 334.27 ms β 313.65 / 324.32 Β±14.95 / 353.86 ms β no change β
β QQuery 24 β 207.41 / 211.54 Β±4.24 / 219.02 ms β 214.19 / 222.83 Β±9.13 / 240.54 ms β 1.05x slower β
β QQuery 25 β 110.73 / 113.03 Β±2.55 / 117.84 ms β 110.93 / 111.98 Β±1.38 / 114.42 ms β no change β
β QQuery 26 β 49.31 / 50.86 Β±0.87 / 51.79 ms β 49.46 / 52.55 Β±2.69 / 57.56 ms β no change β
β QQuery 27 β 6.18 / 6.39 Β±0.19 / 6.73 ms β 6.44 / 6.57 Β±0.10 / 6.74 ms β no change β
β QQuery 28 β 56.88 / 60.83 Β±2.42 / 64.35 ms β 56.28 / 61.34 Β±3.21 / 66.43 ms β no change β
β QQuery 29 β 98.96 / 101.23 Β±3.19 / 107.57 ms β 97.56 / 100.97 Β±4.49 / 109.67 ms β no change β
β QQuery 30 β 32.62 / 33.21 Β±0.43 / 33.82 ms β 32.96 / 34.93 Β±1.65 / 37.46 ms β 1.05x slower β
β QQuery 31 β 111.22 / 112.55 Β±0.94 / 114.16 ms β 112.47 / 113.29 Β±0.83 / 114.45 ms β no change β
β QQuery 32 β 21.01 / 21.30 Β±0.18 / 21.54 ms β 21.10 / 21.27 Β±0.20 / 21.62 ms β no change β
β QQuery 33 β 39.00 / 39.96 Β±0.59 / 40.59 ms β 38.17 / 40.97 Β±3.79 / 48.44 ms β no change β
β QQuery 34 β 10.24 / 10.43 Β±0.18 / 10.70 ms β 10.41 / 10.82 Β±0.46 / 11.69 ms β no change β
β QQuery 35 β 72.80 / 73.61 Β±0.61 / 74.65 ms β 76.50 / 77.41 Β±0.92 / 79.07 ms β 1.05x slower β
β QQuery 36 β 5.98 / 6.15 Β±0.21 / 6.56 ms β 6.19 / 6.32 Β±0.14 / 6.57 ms β no change β
β QQuery 37 β 6.81 / 6.99 Β±0.09 / 7.07 ms β 7.18 / 7.28 Β±0.09 / 7.42 ms β no change β
β QQuery 38 β 62.85 / 63.35 Β±0.40 / 63.86 ms β 64.70 / 64.93 Β±0.30 / 65.51 ms β no change β
β QQuery 39 β 90.57 / 93.07 Β±2.70 / 96.81 ms β 91.91 / 97.28 Β±5.49 / 107.11 ms β no change β
β QQuery 40 β 24.47 / 24.98 Β±0.35 / 25.51 ms β 25.10 / 25.91 Β±0.58 / 26.66 ms β no change β
β QQuery 41 β 11.46 / 11.57 Β±0.09 / 11.71 ms β 11.71 / 11.87 Β±0.09 / 11.95 ms β no change β
β QQuery 42 β 24.05 / 24.32 Β±0.27 / 24.83 ms β 24.31 / 24.55 Β±0.14 / 24.74 ms β no change β
β QQuery 43 β 5.52 / 5.63 Β±0.15 / 5.93 ms β 5.48 / 5.65 Β±0.17 / 5.96 ms β no change β
β QQuery 44 β 9.81 / 9.93 Β±0.10 / 10.09 ms β 9.75 / 9.88 Β±0.12 / 10.04 ms β no change β
β QQuery 45 β 39.35 / 40.68 Β±0.98 / 42.16 ms β 41.02 / 44.94 Β±4.14 / 52.69 ms β 1.10x slower β
β QQuery 46 β 12.98 / 13.17 Β±0.26 / 13.68 ms β 12.76 / 13.23 Β±0.44 / 13.85 ms β no change β
β QQuery 47 β 234.49 / 244.48 Β±9.32 / 261.36 ms β 245.02 / 253.71 Β±9.18 / 270.28 ms β no change β
β QQuery 48 β 95.71 / 96.58 Β±1.03 / 98.51 ms β 96.34 / 97.31 Β±0.64 / 98.15 ms β no change β
β QQuery 49 β 72.02 / 75.52 Β±3.53 / 81.17 ms β 71.81 / 74.69 Β±4.67 / 83.96 ms β no change β
β QQuery 50 β 59.07 / 59.46 Β±0.59 / 60.63 ms β 60.28 / 61.95 Β±1.17 / 63.22 ms β no change β
β QQuery 51 β 93.23 / 95.71 Β±1.61 / 97.78 ms β 93.01 / 95.63 Β±2.81 / 101.08 ms β no change β
β QQuery 52 β 24.70 / 25.08 Β±0.27 / 25.47 ms β 24.74 / 25.19 Β±0.34 / 25.72 ms β no change β
β QQuery 53 β 29.31 / 30.99 Β±1.32 / 33.00 ms β 29.59 / 30.37 Β±1.10 / 32.55 ms β no change β
β QQuery 54 β 55.17 / 55.76 Β±0.68 / 57.01 ms β 56.03 / 56.65 Β±0.53 / 57.53 ms β no change β
β QQuery 55 β 23.39 / 23.67 Β±0.21 / 24.03 ms β 23.40 / 24.00 Β±0.36 / 24.50 ms β no change β
β QQuery 56 β 39.70 / 39.93 Β±0.21 / 40.29 ms β 39.50 / 39.76 Β±0.24 / 40.12 ms β no change β
β QQuery 57 β 176.40 / 179.99 Β±3.89 / 187.55 ms β 179.19 / 182.44 Β±3.32 / 188.60 ms β no change β
β QQuery 58 β 111.98 / 114.87 Β±3.55 / 121.84 ms β 112.72 / 115.34 Β±1.48 / 117.18 ms β no change β
β QQuery 59 β 120.65 / 121.59 Β±1.13 / 123.77 ms β 118.54 / 119.48 Β±1.04 / 121.44 ms β no change β
β QQuery 60 β 39.81 / 40.18 Β±0.32 / 40.60 ms β 39.65 / 41.17 Β±2.27 / 45.64 ms β no change β
β QQuery 61 β 12.71 / 14.04 Β±1.38 / 16.52 ms β 12.41 / 12.62 Β±0.14 / 12.82 ms β +1.11x faster β
β QQuery 62 β 46.43 / 46.83 Β±0.48 / 47.76 ms β 47.40 / 48.72 Β±1.06 / 50.03 ms β no change β
β QQuery 63 β 29.87 / 31.17 Β±1.28 / 33.06 ms β 29.73 / 30.17 Β±0.39 / 30.78 ms β no change β
β QQuery 64 β 374.67 / 394.98 Β±11.75 / 408.16 ms β 376.57 / 380.65 Β±3.76 / 387.47 ms β no change β
β QQuery 65 β 127.19 / 130.13 Β±2.64 / 133.41 ms β 124.92 / 128.60 Β±2.97 / 133.93 ms β no change β
β QQuery 66 β 80.75 / 83.68 Β±3.28 / 89.76 ms β 82.23 / 85.11 Β±4.14 / 92.99 ms β no change β
β QQuery 67 β 254.68 / 274.23 Β±12.67 / 290.45 ms β 251.86 / 258.47 Β±3.39 / 260.92 ms β +1.06x faster β
β QQuery 68 β 13.29 / 13.55 Β±0.25 / 13.88 ms β 12.01 / 12.13 Β±0.17 / 12.46 ms β +1.12x faster β
β QQuery 69 β 57.98 / 61.95 Β±3.89 / 69.40 ms β 57.95 / 59.72 Β±3.00 / 65.70 ms β no change β
β QQuery 70 β 109.89 / 112.71 Β±2.51 / 117.25 ms β 106.56 / 112.19 Β±4.46 / 117.07 ms β no change β
β QQuery 71 β 37.59 / 38.05 Β±0.51 / 39.03 ms β 35.84 / 36.07 Β±0.29 / 36.61 ms β +1.05x faster β
β QQuery 72 β 1984.56 / 2046.57 Β±62.68 / 2159.95 ms β 1939.05 / 1977.85 Β±42.21 / 2055.81 ms β no change β
β QQuery 73 β 10.56 / 11.73 Β±1.78 / 15.24 ms β 10.34 / 10.71 Β±0.37 / 11.25 ms β +1.10x faster β
β QQuery 74 β 184.24 / 189.00 Β±3.63 / 194.40 ms β 182.03 / 200.08 Β±15.50 / 220.52 ms β 1.06x slower β
β QQuery 75 β 148.61 / 150.52 Β±1.74 / 153.67 ms β 148.17 / 159.03 Β±8.77 / 169.76 ms β 1.06x slower β
β QQuery 76 β 35.95 / 36.55 Β±0.47 / 37.13 ms β 35.58 / 36.32 Β±0.41 / 36.71 ms β no change β
β QQuery 77 β 62.09 / 63.54 Β±1.19 / 65.68 ms β 66.61 / 68.72 Β±2.75 / 73.99 ms β 1.08x slower β
β QQuery 78 β 226.59 / 234.72 Β±6.34 / 241.58 ms β 240.91 / 259.98 Β±15.04 / 279.69 ms β 1.11x slower β
β QQuery 79 β 67.60 / 69.66 Β±2.35 / 73.72 ms β 68.21 / 71.13 Β±1.99 / 73.50 ms β no change β
β QQuery 80 β 100.03 / 102.21 Β±1.94 / 105.80 ms β 109.23 / 112.54 Β±4.99 / 122.47 ms β 1.10x slower β
β QQuery 81 β 26.11 / 26.46 Β±0.22 / 26.73 ms β 28.70 / 29.19 Β±0.33 / 29.61 ms β 1.10x slower β
β QQuery 82 β 16.86 / 17.06 Β±0.14 / 17.30 ms β 17.84 / 18.47 Β±0.75 / 19.93 ms β 1.08x slower β
β QQuery 83 β 34.57 / 35.66 Β±0.77 / 36.70 ms β 34.64 / 35.26 Β±0.37 / 35.75 ms β no change β
β QQuery 84 β 30.05 / 30.51 Β±0.46 / 31.16 ms β 30.67 / 31.44 Β±0.45 / 31.99 ms β no change β
β QQuery 85 β 104.95 / 107.47 Β±3.72 / 114.73 ms β 104.88 / 113.05 Β±6.30 / 121.04 ms β 1.05x slower β
β QQuery 86 β 25.86 / 26.02 Β±0.21 / 26.43 ms β 25.70 / 26.39 Β±0.65 / 27.54 ms β no change β
β QQuery 87 β 63.22 / 65.21 Β±1.84 / 67.96 ms β 63.64 / 67.44 Β±2.76 / 70.73 ms β no change β
β QQuery 88 β 64.30 / 64.97 Β±0.51 / 65.84 ms β 65.05 / 66.55 Β±1.33 / 68.26 ms β no change β
β QQuery 89 β 36.70 / 37.23 Β±0.38 / 37.76 ms β 35.72 / 36.74 Β±0.68 / 37.60 ms β no change β
β QQuery 90 β 17.54 / 17.71 Β±0.16 / 17.99 ms β 17.38 / 17.59 Β±0.20 / 17.93 ms β no change β
β QQuery 91 β 45.60 / 46.03 Β±0.42 / 46.84 ms β 45.25 / 45.61 Β±0.19 / 45.78 ms β no change β
β QQuery 92 β 29.69 / 30.54 Β±1.03 / 32.57 ms β 31.46 / 32.64 Β±0.71 / 33.69 ms β 1.07x slower β
β QQuery 93 β 50.37 / 51.32 Β±0.52 / 51.82 ms β 53.48 / 56.06 Β±2.20 / 58.68 ms β 1.09x slower β
β QQuery 94 β 38.55 / 39.44 Β±0.91 / 41.19 ms β 39.03 / 39.47 Β±0.35 / 39.98 ms β no change β
β QQuery 95 β 81.51 / 83.24 Β±1.20 / 85.07 ms β 82.37 / 83.98 Β±1.35 / 86.08 ms β no change β
β QQuery 96 β 24.51 / 24.83 Β±0.31 / 25.39 ms β 25.12 / 25.74 Β±0.46 / 26.53 ms β no change β
β QQuery 97 β 52.77 / 55.45 Β±1.94 / 58.54 ms β 53.29 / 55.66 Β±1.71 / 57.94 ms β no change β
β QQuery 98 β 42.67 / 43.37 Β±0.75 / 44.77 ms β 42.53 / 43.26 Β±0.51 / 43.91 ms β no change β
β QQuery 99 β 70.76 / 71.29 Β±0.49 / 72.21 ms β 69.83 / 70.10 Β±0.24 / 70.50 ms β no change β
βββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββ
β Benchmark Summary β β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 9911.91ms β
β Total Time (pr21621-bench-on-main) β 9961.96ms β
β Average Time (HEAD) β 100.12ms β
β Average Time (pr21621-bench-on-main) β 100.63ms β
β Queries Faster β 5 β
β Queries Slower β 14 β
β Queries with No Change β 80 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 50.0s |
| Peak memory | 2.4 GiB |
| Avg memory | 1.7 GiB |
| CPU user | 215.8s |
| CPU sys | 5.7s |
| Peak spill | 0 B |
**tpcds β branch**
| Metric | Value |
|--------|-------|
| Wall time | 55.0s |
| Peak memory | 2.2 GiB |
| Avg memory | 1.5 GiB |
| CPU user | 213.9s |
| CPU sys | 5.9s |
| 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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangb (via GitHub)" <gi...@apache.org>.
adriangb commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595470052
run benchmark sql_planner
--
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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595673106
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595469801)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpch
```
</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 pr21621-bench-on-main
--------------------
Benchmark tpch_sf1.json
--------------------
βββββββββββββ³βββββββββββ³ββββββββββββββββββββββββ³ββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 39.04 ms β 39.03 ms β no change β
β QQuery 2 β 18.76 ms β 18.95 ms β no change β
β QQuery 3 β 29.02 ms β 29.14 ms β no change β
β QQuery 4 β 17.81 ms β 18.01 ms β no change β
β QQuery 5 β 36.50 ms β 36.89 ms β no change β
β QQuery 6 β 16.16 ms β 16.36 ms β no change β
β QQuery 7 β 41.23 ms β 41.33 ms β no change β
β QQuery 8 β 40.94 ms β 41.90 ms β no change β
β QQuery 9 β 50.40 ms β 50.79 ms β no change β
β QQuery 10 β 43.31 ms β 43.34 ms β no change β
β QQuery 11 β 13.76 ms β 13.96 ms β no change β
β QQuery 12 β 24.80 ms β 24.26 ms β no change β
β QQuery 13 β 40.46 ms β 41.72 ms β no change β
β QQuery 14 β 24.68 ms β 25.31 ms β no change β
β QQuery 15 β 31.43 ms β 31.89 ms β no change β
β QQuery 16 β 14.02 ms β 14.47 ms β no change β
β QQuery 17 β 73.19 ms β 75.08 ms β no change β
β QQuery 18 β 61.38 ms β 62.16 ms β no change β
β QQuery 19 β 33.24 ms β 33.40 ms β no change β
β QQuery 20 β 31.84 ms β 32.37 ms β no change β
β QQuery 21 β 58.29 ms β 57.91 ms β no change β
β QQuery 22 β 14.29 ms β 14.82 ms β no change β
βββββββββββββ΄βββββββββββ΄ββββββββββββββββββββββββ΄ββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³βββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 754.56ms β
β Total Time (pr21621-bench-on-main) β 763.07ms β
β Average Time (HEAD) β 34.30ms β
β Average Time (pr21621-bench-on-main) β 34.68ms β
β Queries Faster β 0 β
β Queries Slower β 0 β
β Queries with No Change β 22 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄βββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and pr21621-bench-on-main
--------------------
Benchmark tpch_sf1.json
--------------------
βββββββββββββ³βββββββββββββββββββββββββββββββββ³βββββββββββββββββββββββββββββββββ³ββββββββββββ
β Query β HEAD β pr21621-bench-on-main β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 39.04 / 41.39 Β±2.10 / 44.83 ms β 39.03 / 40.23 Β±1.14 / 41.86 ms β no change β
β QQuery 2 β 18.76 / 19.09 Β±0.21 / 19.41 ms β 18.95 / 19.28 Β±0.41 / 20.07 ms β no change β
β QQuery 3 β 29.02 / 29.84 Β±1.05 / 31.87 ms β 29.14 / 29.45 Β±0.28 / 29.90 ms β no change β
β QQuery 4 β 17.81 / 17.99 Β±0.14 / 18.17 ms β 18.01 / 18.09 Β±0.08 / 18.23 ms β no change β
β QQuery 5 β 36.50 / 36.92 Β±0.39 / 37.40 ms β 36.89 / 37.90 Β±1.08 / 39.98 ms β no change β
β QQuery 6 β 16.16 / 16.34 Β±0.10 / 16.43 ms β 16.36 / 16.57 Β±0.16 / 16.78 ms β no change β
β QQuery 7 β 41.23 / 43.91 Β±2.62 / 47.44 ms β 41.33 / 43.03 Β±1.11 / 44.36 ms β no change β
β QQuery 8 β 40.94 / 41.67 Β±0.40 / 41.98 ms β 41.90 / 42.81 Β±0.87 / 44.01 ms β no change β
β QQuery 9 β 50.40 / 51.42 Β±0.56 / 52.04 ms β 50.79 / 52.32 Β±1.21 / 54.47 ms β no change β
β QQuery 10 β 43.31 / 44.17 Β±0.97 / 45.43 ms β 43.34 / 43.97 Β±0.63 / 45.11 ms β no change β
β QQuery 11 β 13.76 / 13.93 Β±0.16 / 14.19 ms β 13.96 / 14.11 Β±0.10 / 14.25 ms β no change β
β QQuery 12 β 24.80 / 24.95 Β±0.11 / 25.09 ms β 24.26 / 24.76 Β±0.49 / 25.68 ms β no change β
β QQuery 13 β 40.46 / 41.26 Β±0.50 / 42.01 ms β 41.72 / 43.30 Β±1.64 / 46.28 ms β no change β
β QQuery 14 β 24.68 / 24.86 Β±0.15 / 25.07 ms β 25.31 / 25.65 Β±0.19 / 25.83 ms β no change β
β QQuery 15 β 31.43 / 31.94 Β±0.64 / 33.17 ms β 31.89 / 32.35 Β±0.67 / 33.67 ms β no change β
β QQuery 16 β 14.02 / 14.50 Β±0.47 / 15.39 ms β 14.47 / 14.60 Β±0.11 / 14.78 ms β no change β
β QQuery 17 β 73.19 / 73.89 Β±0.42 / 74.45 ms β 75.08 / 77.33 Β±1.47 / 78.99 ms β no change β
β QQuery 18 β 61.38 / 62.82 Β±1.11 / 64.47 ms β 62.16 / 64.55 Β±1.50 / 66.21 ms β no change β
β QQuery 19 β 33.24 / 33.86 Β±0.79 / 35.37 ms β 33.40 / 33.70 Β±0.37 / 34.43 ms β no change β
β QQuery 20 β 31.84 / 32.27 Β±0.27 / 32.55 ms β 32.37 / 33.16 Β±0.99 / 35.08 ms β no change β
β QQuery 21 β 58.29 / 60.07 Β±1.78 / 62.61 ms β 57.91 / 58.81 Β±0.75 / 59.97 ms β no change β
β QQuery 22 β 14.29 / 14.58 Β±0.18 / 14.80 ms β 14.82 / 14.96 Β±0.12 / 15.10 ms β no change β
βββββββββββββ΄βββββββββββββββββββββββββββββββββ΄βββββββββββββββββββββββββββββββββ΄ββββββββββββ
ββββββββββββββββββββββββββββββββββββββββ³βββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 771.65ms β
β Total Time (pr21621-bench-on-main) β 780.91ms β
β Average Time (HEAD) β 35.08ms β
β Average Time (pr21621-bench-on-main) β 35.50ms β
β Queries Faster β 0 β
β Queries Slower β 0 β
β Queries with No Change β 22 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββ΄βββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpch β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 5.0s |
| Peak memory | 1.2 GiB |
| Avg memory | 493.7 MiB |
| CPU user | 21.7s |
| CPU sys | 1.8s |
| Peak spill | 0 B |
**tpch β branch**
| Metric | Value |
|--------|-------|
| Wall time | 5.0s |
| Peak memory | 1.2 GiB |
| Avg memory | 491.1 MiB |
| CPU user | 22.1s |
| CPU sys | 1.8s |
| 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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595742773
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595728605)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5595728605-2240-kqb4n 6.12.94+ #1 SMP Fri Jul 17 09:42:57 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 pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to 53d3a410e1e96b844ff6096e372344a2927c8faa [diff](https://github.com/apache/datafusion/compare/53d3a410e1e96b844ff6096e372344a2927c8faa..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpcds
baseline:
ref: "53d3a410e1e96b844ff6096e372344a2927c8faa"
```
</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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5596065345
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595682675)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark push_down_topk
env:
CARGO_PROFILE_BENCH_LTO: "thin"
```
</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>
```
group HEAD pr21621-bench-on-main
----- ---- ---------------------
push_down_topk/Q01 1.61 9.8Β±0.06ms ? ?/sec 1.00 6.1Β±0.05ms ? ?/sec
push_down_topk/Q02 1.13 13.0Β±0.07ms ? ?/sec 1.00 11.5Β±0.11ms ? ?/sec
push_down_topk/Q03 1.22 9.4Β±0.06ms ? ?/sec 1.00 7.7Β±0.06ms ? ?/sec
push_down_topk/Q04 1.21 2.9Β±0.02ms ? ?/sec 1.00 2.4Β±0.01ms ? ?/sec
push_down_topk/Q05 1.01 11.6Β±0.09ms ? ?/sec 1.00 11.4Β±0.11ms ? ?/sec
```
</p>
</details>
<details><summary>Resource Usage</summary>
**push_down_topk β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 790.3s |
| Peak memory | 604.4 MiB |
| Avg memory | 32.9 MiB |
| CPU user | 274.9s |
| CPU sys | 27.2s |
| Peak spill | 0 B |
**push_down_topk β branch**
| Metric | Value |
|--------|-------|
| Wall time | 775.3s |
| Peak memory | 486.7 MiB |
| Avg memory | 29.0 MiB |
| CPU user | 248.6s |
| CPU sys | 24.4s |
| 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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595492914
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595469801)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5595469801-2235-r9xfr 6.12.94+ #1 SMP Fri Jul 17 09:42:57 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 pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
```
</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] [benchmark only] PR #21621 push-down TopK through join, rebased on main [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #25093:
URL: https://github.com/apache/datafusion/pull/25093#issuecomment-5595596187
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/25093#issuecomment-5595469801)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5595469801-2237-n7t29 6.12.94+ #1 SMP Fri Jul 17 09:42:57 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 pr21621-bench-on-main (53d3a410e1e96b844ff6096e372344a2927c8faa) to da89c7c (merge-base) [diff](https://github.com/apache/datafusion/compare/da89c7c85b1c888ba4038317eacbdcfe03ab9b1a..53d3a410e1e96b844ff6096e372344a2927c8faa)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpch
```
</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