You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@datafusion.apache.org by "Dandandan (via GitHub)" <gi...@apache.org> on 2026/04/13 12:37:49 UTC
[PR] feat: reorder row groups by grouping key statistics [datafusion]
Dandandan opened a new pull request, #21588:
URL: https://github.com/apache/datafusion/pull/21588
## Which issue does this PR close?
Extends #21580 (row group reorder by statistics during sort pushdown) to also reorder by GROUP BY keys, as suggested in the [review](https://github.com/apache/datafusion/pull/21580#discussion_r2191783519).
## Rationale for this change
When an `AggregateExec` sits above a Parquet data source, reading row groups in grouping-key order clusters similar group values together. This:
1. **Reduces active cardinality** of aggregation hash tables — fewer live entries at any time
2. **Improves CPU cache locality** for hash table lookups during aggregation
## What changes are included in this PR?
- Add `try_pushdown_groupby_order()` to `ExecutionPlan`, `DataSource`, and `FileSource` traits (default: no-op)
- Implement on `DataSourceExec` → `FileScanConfig` → `ParquetSource` chain, reusing the existing `sort_order_for_reorder` / `reorder_by_statistics` infrastructure from #21580
- New `ReorderByGroupKeys` physical optimizer rule that detects `AggregateExec` → `DataSourceExec` patterns and pushes grouping key expressions down
- Rule runs before `PushdownSort` so sort pushdown can override the reorder hint when a sort requirement is present
- SLT tests for GROUP BY correctness with the optimization active
## Are these changes tested?
Yes — SLT tests in `sort_pushdown.slt` (Test I) verify correct GROUP BY SUM and COUNT results with the optimization active on multi-row-group Parquet files.
## Are there any user-facing changes?
No user-facing API changes. The optimization is automatic when aggregating over Parquet data.
🤖 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236437752
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236422970)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236422970-1145-qtz2s 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (6c143f75da587a8f623dd7b92ef85d5df975ca4f) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..6c143f75da587a8f623dd7b92ef85d5df975ca4f) using: clickbench_extended
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4237468507
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4237311314)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<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 reorder-by-groupby-keys
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┓
┃ Query ┃ HEAD ┃ reorder-by-groupby-keys ┃ Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━┩
│ QQuery 1 │ 6.77 / 7.21 ±0.71 / 8.63 ms │ 6.54 / 7.01 ±0.78 / 8.57 ms │ no change │
│ QQuery 2 │ 143.59 / 145.14 ±1.09 / 146.83 ms │ 146.87 / 147.65 ±0.74 / 148.63 ms │ no change │
│ QQuery 3 │ 113.86 / 114.28 ±0.22 / 114.48 ms │ 114.89 / 115.46 ±0.65 / 116.67 ms │ no change │
│ QQuery 4 │ 1355.24 / 1389.92 ±22.20 / 1420.51 ms │ 1384.40 / 1413.72 ±16.72 / 1428.72 ms │ no change │
│ QQuery 5 │ 172.47 / 173.93 ±1.06 / 175.41 ms │ 172.91 / 174.90 ±1.08 / 175.78 ms │ no change │
│ QQuery 6 │ 836.94 / 868.18 ±19.71 / 896.73 ms │ 847.62 / 871.57 ±14.94 / 887.60 ms │ no change │
│ QQuery 7 │ 340.83 / 344.07 ±2.20 / 346.45 ms │ 342.89 / 345.57 ±1.83 / 347.87 ms │ no change │
│ QQuery 8 │ 114.69 / 116.63 ±1.31 / 118.34 ms │ 116.33 / 117.16 ±0.62 / 117.92 ms │ no change │
│ QQuery 9 │ 100.78 / 104.57 ±3.07 / 108.12 ms │ 96.68 / 105.20 ±4.27 / 107.79 ms │ no change │
│ QQuery 10 │ 105.24 / 106.57 ±0.73 / 107.27 ms │ 106.10 / 107.69 ±1.04 / 108.87 ms │ no change │
│ QQuery 11 │ 967.13 / 978.23 ±9.04 / 993.54 ms │ 991.77 / 1004.15 ±8.66 / 1017.15 ms │ no change │
│ QQuery 12 │ 45.16 / 47.42 ±1.78 / 50.61 ms │ 45.13 / 46.24 ±1.17 / 48.38 ms │ no change │
│ QQuery 13 │ 400.38 / 402.70 ±2.55 / 407.57 ms │ 403.76 / 405.84 ±1.15 / 406.94 ms │ no change │
│ QQuery 14 │ 1004.86 / 1009.84 ±2.99 / 1013.75 ms │ 993.85 / 1002.29 ±4.76 / 1007.82 ms │ no change │
│ QQuery 15 │ 15.65 / 16.28 ±0.47 / 16.91 ms │ 16.09 / 16.72 ±0.81 / 18.25 ms │ no change │
│ QQuery 16 │ 7.21 / 7.81 ±0.75 / 9.28 ms │ 7.20 / 7.99 ±0.77 / 9.46 ms │ no change │
│ QQuery 17 │ 226.02 / 228.80 ±1.63 / 230.94 ms │ 226.68 / 229.40 ±1.80 / 232.04 ms │ no change │
│ QQuery 18 │ 125.73 / 127.12 ±0.94 / 127.97 ms │ 126.80 / 128.15 ±1.19 / 130.21 ms │ no change │
│ QQuery 19 │ 154.62 / 156.86 ±1.63 / 158.88 ms │ 155.19 / 157.10 ±1.35 / 158.91 ms │ no change │
│ QQuery 20 │ 14.01 / 14.49 ±0.31 / 14.78 ms │ 13.45 / 14.51 ±0.61 / 15.19 ms │ no change │
│ QQuery 21 │ 19.46 / 20.02 ±0.42 / 20.74 ms │ 20.06 / 20.64 ±0.52 / 21.44 ms │ no change │
│ QQuery 22 │ 486.45 / 490.09 ±2.35 / 492.80 ms │ 484.87 / 486.39 ±1.55 / 489.08 ms │ no change │
│ QQuery 23 │ 869.86 / 884.49 ±10.25 / 897.94 ms │ 877.03 / 888.38 ±6.72 / 897.83 ms │ no change │
│ QQuery 24 │ 380.73 / 384.22 ±2.94 / 389.23 ms │ 382.22 / 386.92 ±4.02 / 391.73 ms │ no change │
│ QQuery 25 │ 340.12 / 342.96 ±1.77 / 345.59 ms │ 344.95 / 348.01 ±2.45 / 352.15 ms │ no change │
│ QQuery 26 │ 80.73 / 82.47 ±1.31 / 84.52 ms │ 81.72 / 82.68 ±0.63 / 83.63 ms │ no change │
│ QQuery 27 │ 6.94 / 7.15 ±0.12 / 7.29 ms │ 7.22 / 7.72 ±0.72 / 9.07 ms │ 1.08x slower │
│ QQuery 28 │ 149.09 / 150.21 ±1.43 / 152.99 ms │ 150.43 / 153.56 ±1.92 / 155.87 ms │ no change │
│ QQuery 29 │ 281.19 / 282.76 ±1.11 / 284.31 ms │ 286.42 / 287.58 ±0.88 / 288.70 ms │ no change │
│ QQuery 30 │ 43.86 / 45.44 ±1.15 / 47.27 ms │ 43.98 / 45.44 ±1.49 / 48.10 ms │ no change │
│ QQuery 31 │ 171.02 / 172.83 ±1.86 / 175.25 ms │ 171.38 / 174.33 ±1.71 / 176.53 ms │ no change │
│ QQuery 32 │ 58.63 / 59.78 ±0.90 / 60.67 ms │ 57.80 / 58.83 ±0.89 / 59.99 ms │ no change │
│ QQuery 33 │ 140.68 / 142.44 ±1.45 / 144.66 ms │ 142.98 / 144.31 ±1.02 / 146.05 ms │ no change │
│ QQuery 34 │ 6.99 / 7.26 ±0.20 / 7.60 ms │ 7.16 / 7.99 ±0.67 / 9.00 ms │ 1.10x slower │
│ QQuery 35 │ 108.11 / 109.17 ±1.32 / 111.66 ms │ 107.70 / 109.46 ±1.52 / 112.06 ms │ no change │
│ QQuery 36 │ 6.56 / 6.85 ±0.17 / 7.02 ms │ 6.99 / 7.10 ±0.10 / 7.25 ms │ no change │
│ QQuery 37 │ 8.24 / 8.48 ±0.23 / 8.89 ms │ 8.69 / 9.15 ±0.39 / 9.86 ms │ 1.08x slower │
│ QQuery 38 │ 81.40 / 84.43 ±3.23 / 89.18 ms │ 81.70 / 89.18 ±4.39 / 95.25 ms │ 1.06x slower │
│ QQuery 39 │ 126.13 / 129.01 ±1.81 / 131.78 ms │ 125.62 / 129.19 ±1.97 / 131.41 ms │ no change │
│ QQuery 40 │ 108.13 / 113.73 ±5.91 / 123.54 ms │ 110.45 / 116.01 ±6.97 / 129.67 ms │ no change │
│ QQuery 41 │ 14.56 / 15.72 ±0.99 / 17.26 ms │ 14.81 / 15.97 ±1.06 / 17.31 ms │ no change │
│ QQuery 42 │ 107.79 / 110.52 ±1.66 / 112.98 ms │ 110.88 / 111.62 ±0.92 / 113.40 ms │ no change │
│ QQuery 43 │ 6.31 / 6.42 ±0.12 / 6.65 ms │ 5.98 / 6.12 ±0.19 / 6.50 ms │ no change │
│ QQuery 44 │ 11.89 / 12.39 ±0.41 / 13.12 ms │ 11.87 / 12.66 ±0.81 / 14.22 ms │ no change │
│ QQuery 45 │ 50.74 / 51.49 ±0.49 / 52.14 ms │ 49.94 / 51.50 ±1.16 / 53.55 ms │ no change │
│ QQuery 46 │ 8.68 / 8.91 ±0.34 / 9.58 ms │ 8.43 / 8.81 ±0.49 / 9.71 ms │ no change │
│ QQuery 47 │ 728.23 / 736.07 ±5.54 / 745.32 ms │ 737.37 / 742.02 ±5.59 / 752.03 ms │ no change │
│ QQuery 48 │ 280.38 / 287.89 ±5.83 / 297.14 ms │ 288.18 / 292.72 ±2.87 / 296.73 ms │ no change │
│ QQuery 49 │ 251.16 / 253.98 ±1.59 / 255.81 ms │ 250.11 / 253.36 ±2.71 / 257.47 ms │ no change │
│ QQuery 50 │ 222.48 / 226.00 ±2.59 / 229.74 ms │ 226.33 / 229.59 ±2.04 / 231.78 ms │ no change │
│ QQuery 51 │ 180.10 / 181.79 ±1.43 / 184.30 ms │ 180.35 / 184.39 ±3.18 / 188.51 ms │ no change │
│ QQuery 52 │ 106.90 / 107.74 ±0.52 / 108.29 ms │ 108.20 / 110.74 ±1.98 / 114.19 ms │ no change │
│ QQuery 53 │ 102.75 / 103.37 ±0.46 / 104.17 ms │ 103.05 / 104.06 ±0.65 / 105.11 ms │ no change │
│ QQuery 54 │ 146.84 / 147.44 ±0.47 / 148.14 ms │ 144.80 / 148.27 ±1.83 / 149.84 ms │ no change │
│ QQuery 55 │ 107.34 / 108.48 ±1.16 / 110.51 ms │ 108.43 / 109.37 ±1.13 / 111.21 ms │ no change │
│ QQuery 56 │ 140.39 / 142.44 ±1.99 / 145.04 ms │ 141.81 / 143.16 ±1.27 / 145.23 ms │ no change │
│ QQuery 57 │ 172.38 / 174.91 ±1.92 / 178.00 ms │ 172.18 / 175.85 ±2.12 / 178.05 ms │ no change │
│ QQuery 58 │ 295.14 / 304.17 ±6.51 / 315.23 ms │ 283.23 / 293.11 ±5.08 / 296.81 ms │ no change │
│ QQuery 59 │ 197.42 / 198.81 ±1.47 / 201.44 ms │ 198.79 / 200.69 ±1.60 / 203.12 ms │ no change │
│ QQuery 60 │ 143.53 / 144.69 ±0.97 / 146.32 ms │ 144.44 / 146.04 ±1.16 / 147.72 ms │ no change │
│ QQuery 61 │ 13.02 / 13.23 ±0.16 / 13.47 ms │ 13.22 / 13.53 ±0.39 / 14.29 ms │ no change │
│ QQuery 62 │ 921.58 / 934.39 ±19.28 / 972.69 ms │ 905.68 / 917.02 ±11.06 / 937.51 ms │ no change │
│ QQuery 63 │ 104.61 / 106.86 ±2.23 / 110.83 ms │ 104.21 / 108.25 ±3.79 / 115.41 ms │ no change │
│ QQuery 64 │ 693.35 / 698.98 ±4.76 / 705.40 ms │ 682.99 / 689.33 ±3.91 / 693.71 ms │ no change │
│ QQuery 65 │ 265.45 / 268.23 ±1.51 / 269.90 ms │ 256.59 / 259.96 ±2.32 / 263.01 ms │ no change │
│ QQuery 66 │ 242.31 / 258.96 ±11.19 / 271.86 ms │ 232.93 / 253.65 ±12.86 / 269.56 ms │ no change │
│ QQuery 67 │ 321.43 / 328.56 ±5.18 / 335.92 ms │ 309.42 / 316.05 ±4.86 / 321.70 ms │ no change │
│ QQuery 68 │ 8.88 / 10.18 ±1.61 / 13.31 ms │ 8.28 / 11.02 ±1.96 / 13.27 ms │ 1.08x slower │
│ QQuery 69 │ 101.51 / 104.27 ±3.03 / 109.79 ms │ 102.66 / 104.71 ±1.50 / 106.59 ms │ no change │
│ QQuery 70 │ 335.39 / 351.27 ±11.19 / 369.09 ms │ 347.06 / 355.24 ±9.17 / 371.84 ms │ no change │
│ QQuery 71 │ 134.42 / 136.66 ±2.30 / 139.72 ms │ 136.36 / 138.81 ±2.50 / 143.25 ms │ no change │
│ QQuery 72 │ 618.56 / 629.63 ±9.19 / 640.97 ms │ 617.98 / 624.35 ±4.28 / 628.40 ms │ no change │
│ QQuery 73 │ 6.82 / 7.91 ±1.00 / 9.71 ms │ 7.57 / 8.51 ±0.76 / 9.48 ms │ 1.08x slower │
│ QQuery 74 │ 602.53 / 612.10 ±6.19 / 621.87 ms │ 616.67 / 619.29 ±3.16 / 625.39 ms │ no change │
│ QQuery 75 │ 274.26 / 277.11 ±2.26 / 280.80 ms │ 277.03 / 279.00 ±1.16 / 280.22 ms │ no change │
│ QQuery 76 │ 130.81 / 132.81 ±1.53 / 134.56 ms │ 133.00 / 134.04 ±0.87 / 135.12 ms │ no change │
│ QQuery 77 │ 188.40 / 189.96 ±0.90 / 191.16 ms │ 190.71 / 191.88 ±1.21 / 193.49 ms │ no change │
│ QQuery 78 │ 340.89 / 345.16 ±4.62 / 351.82 ms │ 336.87 / 344.26 ±4.11 / 348.31 ms │ no change │
│ QQuery 79 │ 235.62 / 238.10 ±2.48 / 242.33 ms │ 235.59 / 238.77 ±2.71 / 243.44 ms │ no change │
│ QQuery 80 │ 319.66 / 322.93 ±1.90 / 325.57 ms │ 318.97 / 322.32 ±2.43 / 326.49 ms │ no change │
│ QQuery 81 │ 26.46 / 28.21 ±1.37 / 29.94 ms │ 25.82 / 27.61 ±1.47 / 30.07 ms │ no change │
│ QQuery 82 │ 196.75 / 199.21 ±1.81 / 201.74 ms │ 199.95 / 201.42 ±1.37 / 203.34 ms │ no change │
│ QQuery 83 │ 38.09 / 38.85 ±0.45 / 39.31 ms │ 39.46 / 40.79 ±1.23 / 42.90 ms │ no change │
│ QQuery 84 │ 48.55 / 51.04 ±1.83 / 53.29 ms │ 47.92 / 49.02 ±1.03 / 50.81 ms │ no change │
│ QQuery 85 │ 146.37 / 147.80 ±1.41 / 150.49 ms │ 148.32 / 149.22 ±0.68 / 150.28 ms │ no change │
│ QQuery 86 │ 39.43 / 39.86 ±0.45 / 40.67 ms │ 38.96 / 40.32 ±1.12 / 42.24 ms │ no change │
│ QQuery 87 │ 84.48 / 88.95 ±3.70 / 95.75 ms │ 83.25 / 87.96 ±3.58 / 92.31 ms │ no change │
│ QQuery 88 │ 100.36 / 101.27 ±0.72 / 102.48 ms │ 100.84 / 106.45 ±8.78 / 123.92 ms │ 1.05x slower │
│ QQuery 89 │ 119.34 / 120.17 ±1.43 / 123.02 ms │ 118.49 / 119.14 ±0.34 / 119.51 ms │ no change │
│ QQuery 90 │ 23.96 / 24.41 ±0.45 / 25.13 ms │ 22.79 / 23.90 ±0.60 / 24.58 ms │ no change │
│ QQuery 91 │ 62.57 / 63.55 ±0.78 / 64.74 ms │ 61.46 / 64.09 ±1.54 / 65.95 ms │ no change │
│ QQuery 92 │ 57.71 / 58.73 ±0.94 / 60.31 ms │ 58.18 / 59.59 ±1.19 / 61.55 ms │ no change │
│ QQuery 93 │ 184.11 / 187.48 ±1.90 / 189.49 ms │ 185.41 / 188.26 ±2.33 / 192.35 ms │ no change │
│ QQuery 94 │ 61.74 / 62.69 ±0.86 / 63.99 ms │ 60.95 / 62.34 ±1.06 / 64.06 ms │ no change │
│ QQuery 95 │ 127.55 / 129.40 ±1.76 / 132.42 ms │ 128.05 / 128.64 ±0.56 / 129.54 ms │ no change │
│ QQuery 96 │ 71.55 / 74.05 ±1.37 / 75.54 ms │ 73.60 / 74.97 ±1.11 / 76.65 ms │ no change │
│ QQuery 97 │ 125.57 / 126.54 ±0.80 / 127.66 ms │ 126.95 / 128.43 ±1.61 / 131.45 ms │ no change │
│ QQuery 98 │ 153.86 / 156.33 ±2.26 / 160.46 ms │ 155.59 / 156.16 ±0.53 / 157.12 ms │ no change │
│ QQuery 99 │ 10817.13 / 10874.72 ±32.88 / 10908.64 ms │ 10816.79 / 10859.57 ±34.22 / 10908.12 ms │ no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴──────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary ┃ ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD) │ 31747.62ms │
│ Total Time (reorder-by-groupby-keys) │ 31809.08ms │
│ Average Time (HEAD) │ 320.68ms │
│ Average Time (reorder-by-groupby-keys) │ 321.30ms │
│ Queries Faster │ 0 │
│ Queries Slower │ 7 │
│ Queries with No Change │ 92 │
│ Queries with Failure │ 0 │
└────────────────────────────────────────┴────────────┘
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds — base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 159.1s |
| Peak memory | 5.4 GiB |
| Avg memory | 4.5 GiB |
| CPU user | 262.0s |
| CPU sys | 17.1s |
| Peak spill | 0 B |
**tpcds — branch**
| Metric | Value |
|--------|-------|
| Wall time | 159.3s |
| Peak memory | 5.1 GiB |
| Avg memory | 4.4 GiB |
| CPU user | 262.7s |
| CPU sys | 16.4s |
| 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236564028
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236422970)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<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 reorder-by-groupby-keys
--------------------
Benchmark clickbench_extended.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query ┃ HEAD ┃ reorder-by-groupby-keys ┃ Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0 │ 840.93 / 859.32 ±14.30 / 873.56 ms │ 824.54 / 840.58 ±11.64 / 857.33 ms │ no change │
│ QQuery 1 │ 207.72 / 208.61 ±0.95 / 210.15 ms │ 212.42 / 213.62 ±0.91 / 214.77 ms │ no change │
│ QQuery 2 │ 494.81 / 496.35 ±1.02 / 497.65 ms │ 509.63 / 510.86 ±1.24 / 513.17 ms │ no change │
│ QQuery 3 │ 315.61 / 316.60 ±0.99 / 318.31 ms │ 315.62 / 316.29 ±0.61 / 317.21 ms │ no change │
│ QQuery 4 │ 668.93 / 688.17 ±17.69 / 717.97 ms │ 658.18 / 666.83 ±5.92 / 676.55 ms │ no change │
│ QQuery 5 │ 9530.42 / 9945.56 ±242.61 / 10226.74 ms │ 9651.16 / 9914.79 ±266.78 / 10389.61 ms │ no change │
│ QQuery 6 │ 990.50 / 1006.11 ±17.69 / 1039.95 ms │ 994.36 / 1004.02 ±9.20 / 1019.83 ms │ no change │
│ QQuery 7 │ 773.05 / 786.23 ±13.18 / 802.65 ms │ 768.11 / 785.43 ±11.75 / 803.62 ms │ no change │
│ QQuery 8 │ 406.26 / 410.89 ±2.82 / 414.57 ms │ 401.45 / 408.45 ±4.91 / 414.11 ms │ no change │
│ QQuery 9 │ 2798.60 / 2828.24 ±28.13 / 2871.90 ms │ 2667.55 / 2709.23 ±31.18 / 2753.57 ms │ no change │
│ QQuery 10 │ 632.82 / 641.85 ±8.59 / 653.10 ms │ 630.96 / 635.06 ±3.79 / 639.73 ms │ no change │
│ QQuery 11 │ 2051.95 / 2067.85 ±16.33 / 2094.29 ms │ 1944.46 / 1954.55 ±11.97 / 1973.10 ms │ +1.06x faster │
│ QQuery 12 │ 193.87 / 198.53 ±4.04 / 205.26 ms │ 192.62 / 201.13 ±5.38 / 209.51 ms │ no change │
└───────────┴─────────────────────────────────────────┴─────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary ┃ ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD) │ 20454.29ms │
│ Total Time (reorder-by-groupby-keys) │ 20160.84ms │
│ Average Time (HEAD) │ 1573.41ms │
│ Average Time (reorder-by-groupby-keys) │ 1550.83ms │
│ Queries Faster │ 1 │
│ Queries Slower │ 0 │
│ Queries with No Change │ 12 │
│ Queries with Failure │ 0 │
└────────────────────────────────────────┴────────────┘
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_extended — base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 103.1s |
| Peak memory | 34.0 GiB |
| Avg memory | 29.8 GiB |
| CPU user | 998.1s |
| CPU sys | 44.9s |
| Peak spill | 0 B |
**clickbench_extended — branch**
| Metric | Value |
|--------|-------|
| Wall time | 101.6s |
| Peak memory | 34.2 GiB |
| Avg memory | 29.1 GiB |
| CPU user | 997.6s |
| CPU sys | 45.4s |
| 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236801904
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236786870)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236786870-1149-cbk5t 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (b41f84e8179d9b636bc618e97162c70fc14f3fd7) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..b41f84e8179d9b636bc618e97162c70fc14f3fd7) using: clickbench_partitioned
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236980088
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236439600
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236423391)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236423391-1146-qkpwr 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (6c143f75da587a8f623dd7b92ef85d5df975ca4f) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..6c143f75da587a8f623dd7b92ef85d5df975ca4f) using: tpch
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan closed pull request #21588: feat: reorder row groups by grouping key statistics
URL: https://github.com/apache/datafusion/pull/21588
--
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236539625
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236421765)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<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 reorder-by-groupby-keys
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query ┃ HEAD ┃ reorder-by-groupby-keys ┃ Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1 │ 6.76 / 7.27 ±0.80 / 8.86 ms │ 6.61 / 7.04 ±0.81 / 8.65 ms │ no change │
│ QQuery 2 │ 143.51 / 145.21 ±1.15 / 146.54 ms │ 144.06 / 145.63 ±1.06 / 146.79 ms │ no change │
│ QQuery 3 │ 113.53 / 114.11 ±0.46 / 114.70 ms │ 113.52 / 114.40 ±0.62 / 115.23 ms │ no change │
│ QQuery 4 │ 1315.80 / 1339.74 ±14.80 / 1360.20 ms │ 1310.68 / 1358.90 ±28.02 / 1386.76 ms │ no change │
│ QQuery 5 │ 172.12 / 174.71 ±2.01 / 177.55 ms │ 173.97 / 175.05 ±0.88 / 176.12 ms │ no change │
│ QQuery 6 │ 843.41 / 868.03 ±25.17 / 914.86 ms │ 842.79 / 859.72 ±22.85 / 904.40 ms │ no change │
│ QQuery 7 │ 338.26 / 341.61 ±1.96 / 343.59 ms │ 346.06 / 348.28 ±1.60 / 350.08 ms │ no change │
│ QQuery 8 │ 116.66 / 117.30 ±0.57 / 118.30 ms │ 116.76 / 117.74 ±1.01 / 119.38 ms │ no change │
│ QQuery 9 │ 100.87 / 103.65 ±2.45 / 107.19 ms │ 101.45 / 104.01 ±2.58 / 108.15 ms │ no change │
│ QQuery 10 │ 105.59 / 107.34 ±1.54 / 110.08 ms │ 106.32 / 109.00 ±1.58 / 110.58 ms │ no change │
│ QQuery 11 │ 913.33 / 937.81 ±18.08 / 962.85 ms │ 916.57 / 942.73 ±18.85 / 975.02 ms │ no change │
│ QQuery 12 │ 45.62 / 47.19 ±0.84 / 48.09 ms │ 46.23 / 47.53 ±0.94 / 49.05 ms │ no change │
│ QQuery 13 │ 398.09 / 403.58 ±2.93 / 406.83 ms │ 402.30 / 405.28 ±2.22 / 407.72 ms │ no change │
│ QQuery 14 │ 992.41 / 1006.38 ±8.70 / 1016.01 ms │ 984.34 / 1001.54 ±9.13 / 1011.53 ms │ no change │
│ QQuery 15 │ 15.55 / 16.94 ±1.74 / 19.89 ms │ 15.78 / 16.91 ±0.72 / 17.85 ms │ no change │
│ QQuery 16 │ 7.16 / 7.75 ±0.84 / 9.38 ms │ 7.32 / 7.89 ±0.72 / 9.28 ms │ no change │
│ QQuery 17 │ 227.08 / 229.04 ±1.37 / 231.17 ms │ 230.67 / 231.30 ±0.37 / 231.83 ms │ no change │
│ QQuery 18 │ 126.68 / 127.90 ±1.48 / 130.75 ms │ 129.10 / 130.15 ±0.99 / 131.88 ms │ no change │
│ QQuery 19 │ 154.79 / 156.05 ±1.19 / 157.88 ms │ 155.08 / 156.54 ±1.59 / 159.01 ms │ no change │
│ QQuery 20 │ 13.43 / 14.60 ±0.93 / 16.19 ms │ 13.71 / 14.18 ±0.61 / 15.35 ms │ no change │
│ QQuery 21 │ 20.00 / 20.34 ±0.40 / 21.11 ms │ 19.37 / 19.65 ±0.25 / 20.08 ms │ no change │
│ QQuery 22 │ 484.89 / 486.57 ±1.30 / 488.84 ms │ 482.89 / 484.17 ±1.22 / 485.78 ms │ no change │
│ QQuery 23 │ 867.86 / 877.42 ±13.59 / 904.25 ms │ 866.18 / 870.56 ±3.98 / 875.50 ms │ no change │
│ QQuery 24 │ 384.12 / 386.20 ±2.55 / 390.86 ms │ 379.39 / 384.51 ±3.61 / 389.86 ms │ no change │
│ QQuery 25 │ 342.80 / 345.49 ±2.00 / 347.33 ms │ 340.11 / 342.60 ±2.83 / 348.11 ms │ no change │
│ QQuery 26 │ 79.94 / 82.97 ±2.18 / 85.73 ms │ 83.40 / 84.09 ±0.44 / 84.58 ms │ no change │
│ QQuery 27 │ 7.05 / 7.74 ±0.55 / 8.68 ms │ 6.93 / 7.94 ±0.94 / 9.47 ms │ no change │
│ QQuery 28 │ 149.86 / 151.73 ±1.82 / 154.96 ms │ 148.92 / 150.70 ±1.27 / 152.75 ms │ no change │
│ QQuery 29 │ 282.82 / 284.52 ±1.34 / 286.93 ms │ 280.86 / 282.14 ±1.26 / 284.52 ms │ no change │
│ QQuery 30 │ 43.13 / 45.68 ±1.83 / 48.86 ms │ 43.99 / 45.29 ±1.26 / 47.37 ms │ no change │
│ QQuery 31 │ 170.48 / 172.66 ±1.78 / 175.68 ms │ 172.16 / 173.18 ±0.53 / 173.66 ms │ no change │
│ QQuery 32 │ 57.99 / 59.29 ±0.89 / 60.38 ms │ 57.90 / 58.89 ±1.17 / 61.16 ms │ no change │
│ QQuery 33 │ 143.41 / 144.81 ±1.19 / 146.47 ms │ 141.67 / 143.52 ±1.57 / 145.55 ms │ no change │
│ QQuery 34 │ 7.12 / 7.30 ±0.24 / 7.78 ms │ 6.91 / 7.25 ±0.24 / 7.59 ms │ no change │
│ QQuery 35 │ 106.95 / 109.87 ±1.62 / 111.79 ms │ 105.30 / 106.94 ±1.07 / 108.25 ms │ no change │
│ QQuery 36 │ 6.52 / 6.93 ±0.31 / 7.36 ms │ 6.47 / 6.85 ±0.31 / 7.21 ms │ no change │
│ QQuery 37 │ 8.52 / 9.21 ±0.58 / 10.19 ms │ 8.15 / 8.51 ±0.27 / 8.89 ms │ +1.08x faster │
│ QQuery 38 │ 84.69 / 88.59 ±3.82 / 94.99 ms │ 82.16 / 86.08 ±4.53 / 94.86 ms │ no change │
│ QQuery 39 │ 126.10 / 129.36 ±2.57 / 132.39 ms │ 122.52 / 126.15 ±3.17 / 131.23 ms │ no change │
│ QQuery 40 │ 108.67 / 113.27 ±5.15 / 123.28 ms │ 108.14 / 115.19 ±7.00 / 128.36 ms │ no change │
│ QQuery 41 │ 15.20 / 15.91 ±0.92 / 17.66 ms │ 14.36 / 14.79 ±0.31 / 15.12 ms │ +1.08x faster │
│ QQuery 42 │ 106.05 / 107.97 ±1.58 / 110.04 ms │ 107.94 / 110.05 ±1.46 / 112.37 ms │ no change │
│ QQuery 43 │ 5.99 / 6.21 ±0.20 / 6.60 ms │ 5.93 / 6.34 ±0.45 / 7.07 ms │ no change │
│ QQuery 44 │ 11.83 / 12.24 ±0.40 / 12.99 ms │ 11.66 / 12.40 ±0.65 / 13.50 ms │ no change │
│ QQuery 45 │ 50.56 / 51.10 ±0.59 / 52.18 ms │ 50.81 / 51.23 ±0.43 / 52.01 ms │ no change │
│ QQuery 46 │ 8.62 / 8.95 ±0.27 / 9.42 ms │ 8.72 / 9.14 ±0.27 / 9.43 ms │ no change │
│ QQuery 47 │ 692.38 / 706.49 ±8.82 / 719.23 ms │ 711.24 / 717.32 ±6.24 / 729.18 ms │ no change │
│ QQuery 48 │ 289.04 / 292.52 ±2.20 / 294.82 ms │ 288.24 / 292.97 ±2.97 / 296.87 ms │ no change │
│ QQuery 49 │ 248.54 / 251.34 ±2.08 / 254.62 ms │ 249.72 / 253.20 ±3.18 / 258.31 ms │ no change │
│ QQuery 50 │ 217.22 / 219.97 ±1.70 / 221.96 ms │ 225.47 / 228.55 ±2.64 / 233.30 ms │ no change │
│ QQuery 51 │ 184.48 / 186.19 ±1.50 / 187.99 ms │ 180.04 / 182.13 ±1.58 / 184.54 ms │ no change │
│ QQuery 52 │ 107.87 / 108.74 ±0.82 / 109.86 ms │ 107.48 / 110.10 ±2.46 / 113.42 ms │ no change │
│ QQuery 53 │ 102.63 / 104.07 ±1.48 / 106.62 ms │ 103.66 / 104.85 ±1.17 / 106.68 ms │ no change │
│ QQuery 54 │ 146.66 / 147.26 ±0.63 / 148.40 ms │ 145.25 / 148.43 ±2.55 / 152.65 ms │ no change │
│ QQuery 55 │ 106.87 / 108.42 ±1.47 / 110.94 ms │ 106.88 / 108.32 ±0.74 / 108.93 ms │ no change │
│ QQuery 56 │ 140.85 / 141.37 ±0.40 / 141.92 ms │ 141.06 / 142.23 ±1.19 / 143.99 ms │ no change │
│ QQuery 57 │ 172.81 / 174.79 ±2.21 / 179.00 ms │ 170.35 / 173.43 ±2.30 / 177.09 ms │ no change │
│ QQuery 58 │ 294.40 / 301.17 ±3.77 / 305.88 ms │ 286.60 / 300.83 ±10.98 / 320.42 ms │ no change │
│ QQuery 59 │ 197.88 / 199.60 ±1.17 / 200.96 ms │ 197.65 / 199.82 ±1.99 / 203.24 ms │ no change │
│ QQuery 60 │ 143.53 / 144.92 ±1.09 / 146.12 ms │ 143.61 / 145.18 ±1.72 / 147.57 ms │ no change │
│ QQuery 61 │ 13.27 / 13.52 ±0.26 / 13.93 ms │ 13.00 / 13.34 ±0.32 / 13.85 ms │ no change │
│ QQuery 62 │ 872.44 / 931.79 ±38.40 / 969.47 ms │ 881.73 / 910.15 ±20.94 / 944.24 ms │ no change │
│ QQuery 63 │ 103.18 / 105.92 ±1.84 / 108.06 ms │ 103.47 / 105.20 ±1.26 / 106.55 ms │ no change │
│ QQuery 64 │ 681.52 / 687.01 ±3.16 / 690.36 ms │ 683.10 / 688.30 ±6.10 / 699.77 ms │ no change │
│ QQuery 65 │ 248.39 / 253.43 ±3.06 / 257.55 ms │ 251.38 / 255.19 ±2.40 / 257.80 ms │ no change │
│ QQuery 66 │ 249.51 / 269.74 ±14.69 / 287.84 ms │ 251.70 / 254.19 ±3.79 / 261.66 ms │ +1.06x faster │
│ QQuery 67 │ 312.64 / 323.30 ±10.79 / 344.10 ms │ 313.57 / 318.22 ±2.68 / 321.45 ms │ no change │
│ QQuery 68 │ 8.83 / 10.52 ±1.36 / 12.97 ms │ 9.30 / 10.38 ±0.75 / 11.37 ms │ no change │
│ QQuery 69 │ 101.75 / 104.53 ±2.51 / 109.12 ms │ 103.38 / 104.09 ±0.39 / 104.55 ms │ no change │
│ QQuery 70 │ 333.30 / 346.58 ±10.96 / 363.53 ms │ 342.76 / 348.40 ±5.84 / 358.77 ms │ no change │
│ QQuery 71 │ 134.94 / 141.54 ±7.37 / 155.53 ms │ 133.38 / 136.32 ±1.53 / 137.81 ms │ no change │
│ QQuery 72 │ 615.65 / 619.77 ±4.19 / 626.81 ms │ 620.19 / 624.00 ±3.12 / 628.05 ms │ no change │
│ QQuery 73 │ 7.33 / 8.48 ±1.31 / 10.61 ms │ 7.02 / 7.87 ±0.50 / 8.54 ms │ +1.08x faster │
│ QQuery 74 │ 591.74 / 597.14 ±3.59 / 602.96 ms │ 569.24 / 581.82 ±10.36 / 596.27 ms │ no change │
│ QQuery 75 │ 275.37 / 279.99 ±2.92 / 283.36 ms │ 277.39 / 279.13 ±1.41 / 281.61 ms │ no change │
│ QQuery 76 │ 130.90 / 132.53 ±1.60 / 135.03 ms │ 132.31 / 133.71 ±1.51 / 136.41 ms │ no change │
│ QQuery 77 │ 186.50 / 189.64 ±2.36 / 193.27 ms │ 188.80 / 190.86 ±1.32 / 192.66 ms │ no change │
│ QQuery 78 │ 335.59 / 340.81 ±4.97 / 348.85 ms │ 338.57 / 343.60 ±3.55 / 347.96 ms │ no change │
│ QQuery 79 │ 233.05 / 235.58 ±1.59 / 237.88 ms │ 230.35 / 234.47 ±2.66 / 238.03 ms │ no change │
│ QQuery 80 │ 319.21 / 323.19 ±3.97 / 330.51 ms │ 322.01 / 324.70 ±1.96 / 327.87 ms │ no change │
│ QQuery 81 │ 26.35 / 27.61 ±1.18 / 29.48 ms │ 26.52 / 27.88 ±1.08 / 29.36 ms │ no change │
│ QQuery 82 │ 197.92 / 199.43 ±0.88 / 200.32 ms │ 197.15 / 199.54 ±2.22 / 203.72 ms │ no change │
│ QQuery 83 │ 39.09 / 39.96 ±1.46 / 42.87 ms │ 38.48 / 39.69 ±0.92 / 40.98 ms │ no change │
│ QQuery 84 │ 48.05 / 49.14 ±0.99 / 50.97 ms │ 49.19 / 50.28 ±1.13 / 52.35 ms │ no change │
│ QQuery 85 │ 146.78 / 149.44 ±1.65 / 151.81 ms │ 149.06 / 152.17 ±3.05 / 157.31 ms │ no change │
│ QQuery 86 │ 38.62 / 39.63 ±0.81 / 41.03 ms │ 38.68 / 39.83 ±0.89 / 41.39 ms │ no change │
│ QQuery 87 │ 83.20 / 87.43 ±3.47 / 93.68 ms │ 86.25 / 88.97 ±3.13 / 94.80 ms │ no change │
│ QQuery 88 │ 100.66 / 101.74 ±0.89 / 103.33 ms │ 99.64 / 104.23 ±7.53 / 119.23 ms │ no change │
│ QQuery 89 │ 119.89 / 120.86 ±0.80 / 122.00 ms │ 118.55 / 120.98 ±1.79 / 124.13 ms │ no change │
│ QQuery 90 │ 23.77 / 24.66 ±0.90 / 26.38 ms │ 23.07 / 23.76 ±0.38 / 24.12 ms │ no change │
│ QQuery 91 │ 61.38 / 63.79 ±1.87 / 66.98 ms │ 65.04 / 65.37 ±0.33 / 65.90 ms │ no change │
│ QQuery 92 │ 57.68 / 58.88 ±1.03 / 60.68 ms │ 57.12 / 58.17 ±0.89 / 59.40 ms │ no change │
│ QQuery 93 │ 186.16 / 187.72 ±1.48 / 190.18 ms │ 185.14 / 187.68 ±1.76 / 189.88 ms │ no change │
│ QQuery 94 │ 61.89 / 62.74 ±1.17 / 64.95 ms │ 61.02 / 62.25 ±0.69 / 63.07 ms │ no change │
│ QQuery 95 │ 127.13 / 127.63 ±0.55 / 128.62 ms │ 127.18 / 128.47 ±1.34 / 130.93 ms │ no change │
│ QQuery 96 │ 73.01 / 74.91 ±1.41 / 76.90 ms │ 73.10 / 74.55 ±1.42 / 76.32 ms │ no change │
│ QQuery 97 │ 123.37 / 127.70 ±3.05 / 130.41 ms │ 122.86 / 125.05 ±1.79 / 127.25 ms │ no change │
│ QQuery 98 │ 152.36 / 154.79 ±1.31 / 156.14 ms │ 153.79 / 155.81 ±1.94 / 159.34 ms │ no change │
│ QQuery 99 │ 10791.68 / 10831.60 ±30.61 / 10872.87 ms │ 10774.39 / 10779.46 ±4.53 / 10787.06 ms │ no change │
└───────────┴──────────────────────────────────────────┴─────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary ┃ ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD) │ 31530.04ms │
│ Total Time (reorder-by-groupby-keys) │ 31457.40ms │
│ Average Time (HEAD) │ 318.49ms │
│ Average Time (reorder-by-groupby-keys) │ 317.75ms │
│ Queries Faster │ 4 │
│ Queries Slower │ 0 │
│ Queries with No Change │ 95 │
│ Queries with Failure │ 0 │
└────────────────────────────────────────┴────────────┘
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds — base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 158.0s |
| Peak memory | 5.1 GiB |
| Avg memory | 4.4 GiB |
| CPU user | 260.1s |
| CPU sys | 17.6s |
| Peak spill | 0 B |
**tpcds — branch**
| Metric | Value |
|--------|-------|
| Wall time | 157.6s |
| Peak memory | 5.8 GiB |
| Avg memory | 4.8 GiB |
| CPU user | 259.9s |
| CPU sys | 17.0s |
| 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4237468705
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4237311314)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<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 reorder-by-groupby-keys
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query ┃ HEAD ┃ reorder-by-groupby-keys ┃ Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0 │ 1.21 / 4.45 ±6.35 / 17.15 ms │ 1.19 / 4.43 ±6.33 / 17.08 ms │ no change │
│ QQuery 1 │ 14.27 / 14.46 ±0.13 / 14.63 ms │ 14.25 / 14.86 ±0.31 / 15.07 ms │ no change │
│ QQuery 2 │ 43.79 / 44.18 ±0.29 / 44.60 ms │ 44.20 / 44.69 ±0.27 / 45.03 ms │ no change │
│ QQuery 3 │ 42.49 / 44.53 ±2.16 / 47.29 ms │ 43.29 / 44.70 ±1.05 / 45.90 ms │ no change │
│ QQuery 4 │ 287.85 / 294.77 ±6.36 / 305.17 ms │ 265.29 / 280.21 ±11.01 / 298.79 ms │ no change │
│ QQuery 5 │ 345.72 / 348.86 ±3.19 / 354.48 ms │ 347.02 / 362.49 ±14.71 / 389.88 ms │ no change │
│ QQuery 6 │ 5.40 / 6.49 ±1.16 / 8.54 ms │ 4.96 / 6.31 ±0.92 / 7.61 ms │ no change │
│ QQuery 7 │ 16.54 / 17.11 ±0.59 / 17.92 ms │ 16.82 / 17.02 ±0.17 / 17.32 ms │ no change │
│ QQuery 8 │ 431.06 / 437.58 ±6.68 / 449.26 ms │ 389.17 / 401.49 ±17.15 / 434.61 ms │ +1.09x faster │
│ QQuery 9 │ 648.22 / 655.29 ±7.68 / 669.10 ms │ 641.77 / 648.70 ±3.95 / 653.27 ms │ no change │
│ QQuery 10 │ 92.35 / 93.81 ±1.44 / 96.26 ms │ 95.13 / 100.03 ±5.51 / 107.76 ms │ 1.07x slower │
│ QQuery 11 │ 104.39 / 105.17 ±0.49 / 105.87 ms │ 107.67 / 110.38 ±3.06 / 116.28 ms │ no change │
│ QQuery 12 │ 342.44 / 348.06 ±4.54 / 354.46 ms │ 340.85 / 348.97 ±8.98 / 365.47 ms │ no change │
│ QQuery 13 │ 461.88 / 477.59 ±12.42 / 498.99 ms │ 444.86 / 454.71 ±5.73 / 461.27 ms │ no change │
│ QQuery 14 │ 347.05 / 349.43 ±2.36 / 352.30 ms │ 347.10 / 368.29 ±11.28 / 378.49 ms │ 1.05x slower │
│ QQuery 15 │ 350.92 / 368.43 ±14.35 / 389.04 ms │ 331.79 / 340.53 ±6.44 / 348.90 ms │ +1.08x faster │
│ QQuery 16 │ 714.43 / 723.10 ±7.69 / 735.08 ms │ 684.27 / 703.92 ±12.27 / 720.25 ms │ no change │
│ QQuery 17 │ 709.73 / 716.40 ±4.16 / 721.82 ms │ 689.44 / 710.33 ±15.59 / 729.11 ms │ no change │
│ QQuery 18 │ 1441.53 / 1482.07 ±30.21 / 1516.83 ms │ 1285.38 / 1322.15 ±21.42 / 1351.09 ms │ +1.12x faster │
│ QQuery 19 │ 36.55 / 47.53 ±14.61 / 76.39 ms │ 36.59 / 37.68 ±1.17 / 39.85 ms │ +1.26x faster │
│ QQuery 20 │ 716.83 / 736.38 ±20.84 / 763.26 ms │ 714.35 / 724.73 ±8.67 / 740.57 ms │ no change │
│ QQuery 21 │ 765.33 / 774.84 ±7.99 / 789.27 ms │ 767.17 / 781.29 ±15.26 / 811.02 ms │ no change │
│ QQuery 22 │ 1138.58 / 1143.75 ±4.96 / 1152.15 ms │ 1135.21 / 1140.92 ±4.15 / 1146.52 ms │ no change │
│ QQuery 23 │ 3090.29 / 3099.37 ±9.79 / 3116.83 ms │ 3071.34 / 3083.32 ±9.93 / 3098.79 ms │ no change │
│ QQuery 24 │ 99.42 / 102.59 ±1.83 / 104.59 ms │ 101.23 / 103.48 ±2.13 / 106.28 ms │ no change │
│ QQuery 25 │ 138.58 / 140.61 ±1.54 / 142.77 ms │ 139.02 / 141.76 ±2.95 / 147.48 ms │ no change │
│ QQuery 26 │ 100.42 / 103.44 ±1.97 / 106.26 ms │ 99.81 / 103.00 ±2.19 / 106.18 ms │ no change │
│ QQuery 27 │ 853.29 / 857.69 ±2.27 / 859.60 ms │ 854.67 / 860.01 ±5.54 / 867.48 ms │ no change │
│ QQuery 28 │ 3289.47 / 3308.66 ±12.69 / 3319.25 ms │ 3274.88 / 3317.14 ±23.04 / 3342.36 ms │ no change │
│ QQuery 29 │ 49.33 / 56.39 ±9.41 / 75.05 ms │ 50.42 / 55.31 ±4.73 / 64.36 ms │ no change │
│ QQuery 30 │ 359.82 / 366.15 ±4.83 / 374.26 ms │ 353.73 / 374.17 ±20.53 / 411.22 ms │ no change │
│ QQuery 31 │ 356.65 / 375.55 ±13.11 / 393.02 ms │ 366.68 / 371.71 ±3.17 / 376.29 ms │ no change │
│ QQuery 32 │ 1248.61 / 1285.33 ±35.02 / 1348.47 ms │ 1029.32 / 1046.95 ±22.26 / 1085.76 ms │ +1.23x faster │
│ QQuery 33 │ 1534.24 / 1590.81 ±44.59 / 1669.28 ms │ 1459.62 / 1545.08 ±52.35 / 1617.27 ms │ no change │
│ QQuery 34 │ 1507.48 / 1522.68 ±9.10 / 1534.32 ms │ 1471.95 / 1502.54 ±27.82 / 1543.62 ms │ no change │
│ QQuery 35 │ 388.00 / 390.35 ±1.66 / 392.59 ms │ 308.19 / 315.64 ±5.40 / 322.37 ms │ +1.24x faster │
│ QQuery 36 │ 121.22 / 122.72 ±0.91 / 123.72 ms │ 113.59 / 115.98 ±2.02 / 119.67 ms │ +1.06x faster │
│ QQuery 37 │ 49.78 / 51.45 ±1.56 / 54.05 ms │ 49.02 / 50.89 ±1.07 / 52.23 ms │ no change │
│ QQuery 38 │ 78.83 / 80.39 ±1.53 / 82.94 ms │ 76.49 / 78.13 ±1.30 / 79.90 ms │ no change │
│ QQuery 39 │ 206.94 / 218.69 ±10.62 / 233.61 ms │ 200.65 / 209.20 ±9.82 / 227.61 ms │ no change │
│ QQuery 40 │ 26.05 / 26.96 ±0.66 / 28.08 ms │ 22.84 / 25.41 ±1.99 / 28.42 ms │ +1.06x faster │
│ QQuery 41 │ 20.62 / 21.35 ±0.68 / 22.43 ms │ 19.89 / 21.15 ±0.81 / 22.38 ms │ no change │
│ QQuery 42 │ 18.84 / 20.48 ±1.23 / 22.61 ms │ 19.26 / 20.08 ±0.52 / 20.65 ms │ no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary ┃ ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD) │ 22975.95ms │
│ Total Time (reorder-by-groupby-keys) │ 22309.79ms │
│ Average Time (HEAD) │ 534.32ms │
│ Average Time (reorder-by-groupby-keys) │ 518.83ms │
│ Queries Faster │ 8 │
│ Queries Slower │ 2 │
│ Queries with No Change │ 33 │
│ Queries with Failure │ 0 │
└────────────────────────────────────────┴────────────┘
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned — base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 115.9s |
| Peak memory | 36.4 GiB |
| Avg memory | 26.3 GiB |
| CPU user | 1076.1s |
| CPU sys | 99.6s |
| Peak spill | 0 B |
**clickbench_partitioned — branch**
| Metric | Value |
|--------|-------|
| Wall time | 112.7s |
| Peak memory | 40.2 GiB |
| Avg memory | 30.0 GiB |
| CPU user | 1025.8s |
| CPU sys | 77.6s |
| Peak spill | 0 B |
</details>
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4237333493
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4237311314)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4237311314-1155-lvklk 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (beb832507495d703815cfef410bcb5256bb4c15e) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..beb832507495d703815cfef410bcb5256bb4c15e) using: clickbench_partitioned
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236436796
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236421765)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236421765-1142-wqcp8 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (6c143f75da587a8f623dd7b92ef85d5df975ca4f) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..6c143f75da587a8f623dd7b92ef85d5df975ca4f) using: clickbench_partitioned
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236443827
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236421765)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236421765-1143-85lls 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (6c143f75da587a8f623dd7b92ef85d5df975ca4f) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..6c143f75da587a8f623dd7b92ef85d5df975ca4f) using: tpcds
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236474741
run benchmark clickbench_1
--
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236422970
run benchmark clickbench_extended
--
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236451497
Benchmark for [this request](https://github.com/apache/datafusion/pull/21588#issuecomment-4236437119) failed.
Last 20 lines of output:
<details><summary>Click to expand</summary>
```
struct_query_sql
substr
substr_index
substring
sum
to_char
to_hex
to_local_time
to_time
to_timestamp
topk_aggregate
topk_repartition
translate
trim
trunc
unhex
upper
uuid
window_query_sql
with_hashes
```
</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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236801804
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236786870)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236786870-1151-2hpsp 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (b41f84e8179d9b636bc618e97162c70fc14f3fd7) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..b41f84e8179d9b636bc618e97162c70fc14f3fd7) using: tpch
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236996016
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236980088)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236980088-1153-ztgtw 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (1c67f2912996a4f19a36cb31b43c32455b76873b) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..1c67f2912996a4f19a36cb31b43c32455b76873b) using: tpcds
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236542707
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236421765)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<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 reorder-by-groupby-keys
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query ┃ HEAD ┃ reorder-by-groupby-keys ┃ Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0 │ 1.19 / 4.49 ±6.44 / 17.36 ms │ 1.18 / 4.41 ±6.32 / 17.05 ms │ no change │
│ QQuery 1 │ 14.41 / 15.08 ±0.34 / 15.34 ms │ 14.06 / 14.43 ±0.19 / 14.56 ms │ no change │
│ QQuery 2 │ 44.16 / 44.69 ±0.44 / 45.37 ms │ 43.60 / 43.91 ±0.26 / 44.22 ms │ no change │
│ QQuery 3 │ 43.13 / 45.33 ±1.31 / 46.70 ms │ 44.77 / 46.02 ±0.96 / 47.33 ms │ no change │
│ QQuery 4 │ 284.89 / 290.43 ±4.13 / 295.51 ms │ 284.58 / 296.20 ±8.99 / 309.78 ms │ no change │
│ QQuery 5 │ 339.86 / 343.66 ±4.63 / 352.59 ms │ 336.57 / 341.50 ±4.62 / 349.57 ms │ no change │
│ QQuery 6 │ 5.34 / 6.76 ±1.32 / 8.66 ms │ 5.28 / 6.17 ±0.96 / 7.92 ms │ +1.10x faster │
│ QQuery 7 │ 17.11 / 17.31 ±0.27 / 17.82 ms │ 16.65 / 17.45 ±0.93 / 18.61 ms │ no change │
│ QQuery 8 │ 402.75 / 414.89 ±10.47 / 428.41 ms │ 418.02 / 425.85 ±4.47 / 430.78 ms │ no change │
│ QQuery 9 │ 645.11 / 648.99 ±3.72 / 656.02 ms │ 646.01 / 661.02 ±8.19 / 670.03 ms │ no change │
│ QQuery 10 │ 92.46 / 94.42 ±1.69 / 96.75 ms │ 93.90 / 95.81 ±2.13 / 99.89 ms │ no change │
│ QQuery 11 │ 105.05 / 106.10 ±0.62 / 106.72 ms │ 103.95 / 106.51 ±1.57 / 108.62 ms │ no change │
│ QQuery 12 │ 333.28 / 341.25 ±6.86 / 351.45 ms │ 344.75 / 353.40 ±6.89 / 362.19 ms │ no change │
│ QQuery 13 │ 445.00 / 463.62 ±14.40 / 487.67 ms │ 464.49 / 475.69 ±10.04 / 490.03 ms │ no change │
│ QQuery 14 │ 340.50 / 345.25 ±6.02 / 356.96 ms │ 347.54 / 352.34 ±4.59 / 358.15 ms │ no change │
│ QQuery 15 │ 343.28 / 368.70 ±18.08 / 391.32 ms │ 352.78 / 366.84 ±15.80 / 393.84 ms │ no change │
│ QQuery 16 │ 701.83 / 723.25 ±19.77 / 749.00 ms │ 704.52 / 721.91 ±11.67 / 733.93 ms │ no change │
│ QQuery 17 │ 701.19 / 728.93 ±25.73 / 772.58 ms │ 714.65 / 720.17 ±4.06 / 726.75 ms │ no change │
│ QQuery 18 │ 1423.80 / 1492.86 ±40.13 / 1546.51 ms │ 1419.65 / 1474.59 ±31.28 / 1516.01 ms │ no change │
│ QQuery 19 │ 35.33 / 44.04 ±16.24 / 76.51 ms │ 34.82 / 45.50 ±18.60 / 82.63 ms │ no change │
│ QQuery 20 │ 720.62 / 740.04 ±21.55 / 780.16 ms │ 719.72 / 734.16 ±22.10 / 777.58 ms │ no change │
│ QQuery 21 │ 764.37 / 769.73 ±5.42 / 780.03 ms │ 759.10 / 763.27 ±3.51 / 768.06 ms │ no change │
│ QQuery 22 │ 1133.23 / 1147.13 ±7.16 / 1153.41 ms │ 1129.86 / 1133.10 ±2.92 / 1137.79 ms │ no change │
│ QQuery 23 │ 3076.24 / 3092.04 ±19.72 / 3129.90 ms │ 3086.66 / 3107.97 ±16.11 / 3132.90 ms │ no change │
│ QQuery 24 │ 101.58 / 103.93 ±1.65 / 106.75 ms │ 100.76 / 106.00 ±3.41 / 109.87 ms │ no change │
│ QQuery 25 │ 139.03 / 140.59 ±1.76 / 143.91 ms │ 139.26 / 141.88 ±2.75 / 147.12 ms │ no change │
│ QQuery 26 │ 100.06 / 103.20 ±2.12 / 105.57 ms │ 97.15 / 102.82 ±3.17 / 106.39 ms │ no change │
│ QQuery 27 │ 846.42 / 849.70 ±2.08 / 852.69 ms │ 846.65 / 852.68 ±7.30 / 866.97 ms │ no change │
│ QQuery 28 │ 3258.79 / 3291.72 ±17.87 / 3309.37 ms │ 3280.12 / 3302.12 ±11.21 / 3310.63 ms │ no change │
│ QQuery 29 │ 50.23 / 54.86 ±5.81 / 66.27 ms │ 50.23 / 53.93 ±4.41 / 62.06 ms │ no change │
│ QQuery 30 │ 356.20 / 363.34 ±6.20 / 374.44 ms │ 364.04 / 366.90 ±2.17 / 369.72 ms │ no change │
│ QQuery 31 │ 374.19 / 379.70 ±5.76 / 387.84 ms │ 375.52 / 380.87 ±5.59 / 391.25 ms │ no change │
│ QQuery 32 │ 1128.56 / 1319.92 ±120.42 / 1502.98 ms │ 1205.03 / 1253.80 ±49.93 / 1345.77 ms │ +1.05x faster │
│ QQuery 33 │ 1448.36 / 1490.55 ±43.35 / 1572.84 ms │ 1454.59 / 1474.50 ±23.87 / 1518.70 ms │ no change │
│ QQuery 34 │ 1452.16 / 1460.83 ±7.56 / 1474.12 ms │ 1458.03 / 1468.82 ±8.48 / 1483.98 ms │ no change │
│ QQuery 35 │ 379.83 / 387.51 ±5.26 / 396.14 ms │ 382.41 / 390.99 ±5.42 / 396.95 ms │ no change │
│ QQuery 36 │ 112.19 / 117.77 ±3.66 / 120.88 ms │ 114.51 / 120.09 ±4.11 / 127.21 ms │ no change │
│ QQuery 37 │ 48.13 / 49.49 ±1.07 / 50.96 ms │ 48.86 / 49.86 ±1.41 / 52.62 ms │ no change │
│ QQuery 38 │ 76.15 / 77.66 ±1.34 / 79.96 ms │ 77.49 / 79.59 ±1.37 / 81.32 ms │ no change │
│ QQuery 39 │ 211.15 / 218.42 ±5.57 / 227.58 ms │ 200.40 / 215.62 ±10.27 / 230.49 ms │ no change │
│ QQuery 40 │ 24.72 / 26.82 ±1.17 / 28.17 ms │ 23.93 / 26.10 ±1.99 / 29.50 ms │ no change │
│ QQuery 41 │ 20.88 / 21.66 ±0.73 / 22.92 ms │ 19.46 / 21.24 ±1.53 / 24.04 ms │ no change │
│ QQuery 42 │ 20.19 / 21.06 ±1.03 / 23.02 ms │ 19.47 / 19.79 ±0.21 / 20.05 ms │ +1.06x faster │
└───────────┴────────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary ┃ ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD) │ 22767.73ms │
│ Total Time (reorder-by-groupby-keys) │ 22735.80ms │
│ Average Time (HEAD) │ 529.48ms │
│ Average Time (reorder-by-groupby-keys) │ 528.74ms │
│ Queries Faster │ 3 │
│ Queries Slower │ 0 │
│ Queries with No Change │ 40 │
│ Queries with Failure │ 0 │
└────────────────────────────────────────┴────────────┘
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned — base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 115.0s |
| Peak memory | 41.0 GiB |
| Avg memory | 29.7 GiB |
| CPU user | 1071.2s |
| CPU sys | 95.0s |
| Peak spill | 0 B |
**clickbench_partitioned — branch**
| Metric | Value |
|--------|-------|
| Wall time | 114.8s |
| Peak memory | 40.8 GiB |
| Avg memory | 29.3 GiB |
| CPU user | 1074.2s |
| CPU sys | 92.9s |
| 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4237311314
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236423391
run benchmark 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236794791
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236786870)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236786870-1150-fv6dz 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (b41f84e8179d9b636bc618e97162c70fc14f3fd7) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..b41f84e8179d9b636bc618e97162c70fc14f3fd7) using: tpcds
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236786870
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236910445
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236786870)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<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 reorder-by-groupby-keys
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query ┃ HEAD ┃ reorder-by-groupby-keys ┃ Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0 │ 1.19 / 4.44 ±6.37 / 17.17 ms │ 1.19 / 4.40 ±6.31 / 17.01 ms │ no change │
│ QQuery 1 │ 14.06 / 14.41 ±0.19 / 14.58 ms │ 14.05 / 14.40 ±0.22 / 14.71 ms │ no change │
│ QQuery 2 │ 44.39 / 44.61 ±0.19 / 44.93 ms │ 43.52 / 43.86 ±0.27 / 44.27 ms │ no change │
│ QQuery 3 │ 43.48 / 45.65 ±1.32 / 47.57 ms │ 43.49 / 45.53 ±1.88 / 48.12 ms │ no change │
│ QQuery 4 │ 276.00 / 286.60 ±7.46 / 298.43 ms │ 288.29 / 296.12 ±5.36 / 303.23 ms │ no change │
│ QQuery 5 │ 339.18 / 345.21 ±4.07 / 351.67 ms │ 342.16 / 348.43 ±4.30 / 354.73 ms │ no change │
│ QQuery 6 │ 5.12 / 6.59 ±1.23 / 8.85 ms │ 5.62 / 6.95 ±1.39 / 9.34 ms │ 1.05x slower │
│ QQuery 7 │ 16.57 / 17.49 ±0.58 / 18.07 ms │ 16.44 / 17.46 ±1.01 / 18.92 ms │ no change │
│ QQuery 8 │ 419.85 / 421.80 ±1.59 / 423.27 ms │ 408.55 / 418.09 ±5.89 / 423.46 ms │ no change │
│ QQuery 9 │ 633.72 / 640.33 ±3.79 / 644.79 ms │ 635.01 / 645.03 ±7.09 / 655.34 ms │ no change │
│ QQuery 10 │ 91.38 / 93.09 ±1.59 / 95.74 ms │ 89.52 / 95.65 ±4.57 / 101.53 ms │ no change │
│ QQuery 11 │ 102.50 / 103.45 ±0.53 / 104.11 ms │ 102.81 / 103.75 ±1.48 / 106.68 ms │ no change │
│ QQuery 12 │ 337.26 / 341.99 ±6.32 / 354.36 ms │ 338.76 / 346.15 ±7.00 / 359.45 ms │ no change │
│ QQuery 13 │ 447.49 / 471.34 ±20.68 / 509.59 ms │ 469.82 / 480.53 ±12.43 / 503.29 ms │ no change │
│ QQuery 14 │ 336.76 / 344.60 ±4.29 / 349.37 ms │ 338.94 / 346.27 ±4.81 / 353.60 ms │ no change │
│ QQuery 15 │ 344.89 / 371.33 ±26.20 / 414.01 ms │ 348.86 / 366.35 ±15.07 / 390.58 ms │ no change │
│ QQuery 16 │ 708.74 / 728.65 ±21.98 / 767.21 ms │ 716.52 / 724.53 ±7.39 / 736.02 ms │ no change │
│ QQuery 17 │ 708.79 / 710.26 ±1.26 / 712.59 ms │ 708.58 / 716.06 ±6.16 / 725.34 ms │ no change │
│ QQuery 18 │ 1398.09 / 1445.93 ±40.92 / 1498.49 ms │ 1393.32 / 1444.49 ±34.23 / 1498.48 ms │ no change │
│ QQuery 19 │ 35.30 / 36.60 ±0.98 / 38.04 ms │ 35.95 / 63.91 ±39.91 / 139.23 ms │ 1.75x slower │
│ QQuery 20 │ 710.33 / 723.54 ±12.74 / 740.40 ms │ 726.71 / 739.68 ±17.17 / 773.31 ms │ no change │
│ QQuery 21 │ 755.96 / 760.34 ±3.10 / 765.33 ms │ 756.83 / 765.04 ±4.40 / 768.97 ms │ no change │
│ QQuery 22 │ 1125.67 / 1133.47 ±5.42 / 1142.69 ms │ 1128.34 / 1134.33 ±4.03 / 1138.57 ms │ no change │
│ QQuery 23 │ 3087.76 / 3113.62 ±19.09 / 3143.66 ms │ 3039.55 / 3070.85 ±20.96 / 3094.13 ms │ no change │
│ QQuery 24 │ 101.28 / 103.79 ±2.70 / 109.02 ms │ 104.91 / 106.86 ±1.80 / 109.61 ms │ no change │
│ QQuery 25 │ 137.77 / 141.33 ±1.93 / 142.94 ms │ 138.52 / 140.28 ±1.41 / 142.81 ms │ no change │
│ QQuery 26 │ 101.02 / 103.14 ±1.70 / 105.88 ms │ 100.35 / 102.19 ±1.51 / 103.62 ms │ no change │
│ QQuery 27 │ 848.14 / 850.72 ±3.05 / 855.08 ms │ 849.64 / 855.13 ±4.28 / 861.07 ms │ no change │
│ QQuery 28 │ 3257.25 / 3285.19 ±19.45 / 3313.21 ms │ 3261.89 / 3282.59 ±19.09 / 3312.22 ms │ no change │
│ QQuery 29 │ 51.35 / 53.90 ±2.55 / 58.70 ms │ 49.55 / 54.72 ±4.80 / 62.29 ms │ no change │
│ QQuery 30 │ 358.44 / 361.26 ±3.76 / 368.48 ms │ 357.33 / 366.27 ±7.79 / 378.41 ms │ no change │
│ QQuery 31 │ 350.66 / 377.47 ±14.44 / 393.73 ms │ 354.09 / 377.71 ±14.90 / 393.95 ms │ no change │
│ QQuery 32 │ 1021.98 / 1039.66 ±20.18 / 1077.97 ms │ 1016.63 / 1062.21 ±53.52 / 1163.09 ms │ no change │
│ QQuery 33 │ 1447.36 / 1466.65 ±12.96 / 1480.68 ms │ 1453.09 / 1460.66 ±9.34 / 1478.74 ms │ no change │
│ QQuery 34 │ 1446.41 / 1514.07 ±44.05 / 1564.54 ms │ 1427.78 / 1444.54 ±9.23 / 1456.18 ms │ no change │
│ QQuery 35 │ 381.50 / 386.22 ±2.68 / 388.81 ms │ 382.77 / 388.08 ±3.43 / 392.61 ms │ no change │
│ QQuery 36 │ 111.86 / 119.77 ±4.03 / 122.67 ms │ 111.34 / 120.69 ±5.55 / 128.55 ms │ no change │
│ QQuery 37 │ 47.71 / 48.42 ±0.65 / 49.20 ms │ 48.02 / 49.74 ±1.47 / 52.19 ms │ no change │
│ QQuery 38 │ 75.04 / 76.63 ±1.21 / 78.55 ms │ 76.62 / 77.58 ±0.68 / 78.30 ms │ no change │
│ QQuery 39 │ 214.36 / 222.46 ±6.01 / 232.56 ms │ 201.98 / 211.22 ±6.98 / 220.66 ms │ +1.05x faster │
│ QQuery 40 │ 24.16 / 26.13 ±1.36 / 28.19 ms │ 24.02 / 25.49 ±1.08 / 26.85 ms │ no change │
│ QQuery 41 │ 20.73 / 21.67 ±1.05 / 23.52 ms │ 20.05 / 21.24 ±1.19 / 22.67 ms │ no change │
│ QQuery 42 │ 18.92 / 19.61 ±0.77 / 20.87 ms │ 18.73 / 19.56 ±0.63 / 20.68 ms │ no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary ┃ ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD) │ 22423.43ms │
│ Total Time (reorder-by-groupby-keys) │ 22404.60ms │
│ Average Time (HEAD) │ 521.48ms │
│ Average Time (reorder-by-groupby-keys) │ 521.04ms │
│ Queries Faster │ 1 │
│ Queries Slower │ 2 │
│ Queries with No Change │ 40 │
│ Queries with Failure │ 0 │
└────────────────────────────────────────┴────────────┘
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned — base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 113.4s |
| Peak memory | 40.3 GiB |
| Avg memory | 32.1 GiB |
| CPU user | 1066.5s |
| CPU sys | 86.6s |
| Peak spill | 0 B |
**clickbench_partitioned — branch**
| Metric | Value |
|--------|-------|
| Wall time | 113.2s |
| Peak memory | 40.6 GiB |
| Avg memory | 29.1 GiB |
| CPU user | 1064.8s |
| CPU sys | 86.0s |
| 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4237332206
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4237311314)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4237311314-1156-rgsvt 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (beb832507495d703815cfef410bcb5256bb4c15e) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..beb832507495d703815cfef410bcb5256bb4c15e) using: tpcds
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236450538
🤖 Criterion benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236437119)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236437119-1147-2gxtm 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (6c143f75da587a8f623dd7b92ef85d5df975ca4f) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..6c143f75da587a8f623dd7b92ef85d5df975ca4f)
BENCH_NAME=clickbench
BENCH_COMMAND=cargo bench --features=parquet --bench clickbench
BENCH_FILTER=
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236904580
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236786870)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<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 reorder-by-groupby-keys
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query ┃ HEAD ┃ reorder-by-groupby-keys ┃ Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1 │ 6.59 / 7.02 ±0.79 / 8.61 ms │ 6.63 / 7.08 ±0.75 / 8.56 ms │ no change │
│ QQuery 2 │ 144.28 / 144.73 ±0.29 / 145.17 ms │ 145.77 / 146.73 ±0.58 / 147.55 ms │ no change │
│ QQuery 3 │ 113.11 / 114.19 ±1.20 / 116.46 ms │ 112.36 / 113.03 ±0.60 / 113.72 ms │ no change │
│ QQuery 4 │ 1269.59 / 1305.04 ±24.48 / 1341.33 ms │ 1284.78 / 1315.81 ±17.37 / 1335.00 ms │ no change │
│ QQuery 5 │ 171.37 / 173.66 ±1.85 / 176.86 ms │ 174.78 / 177.06 ±1.98 / 179.89 ms │ no change │
│ QQuery 6 │ 837.13 / 855.78 ±13.30 / 872.65 ms │ 846.27 / 867.61 ±17.75 / 892.36 ms │ no change │
│ QQuery 7 │ 340.87 / 344.22 ±2.43 / 347.13 ms │ 346.29 / 348.18 ±1.53 / 350.01 ms │ no change │
│ QQuery 8 │ 115.21 / 116.17 ±1.15 / 118.21 ms │ 115.43 / 117.26 ±1.78 / 120.28 ms │ no change │
│ QQuery 9 │ 100.32 / 102.57 ±2.35 / 106.88 ms │ 102.18 / 105.74 ±2.74 / 108.27 ms │ no change │
│ QQuery 10 │ 105.59 / 107.10 ±1.28 / 108.58 ms │ 104.84 / 106.60 ±0.98 / 107.50 ms │ no change │
│ QQuery 11 │ 868.04 / 883.96 ±11.68 / 897.49 ms │ 873.25 / 888.44 ±8.30 / 896.71 ms │ no change │
│ QQuery 12 │ 43.86 / 44.89 ±1.12 / 46.77 ms │ 44.03 / 46.24 ±2.13 / 49.51 ms │ no change │
│ QQuery 13 │ 399.26 / 403.22 ±3.92 / 410.56 ms │ 396.96 / 402.81 ±3.77 / 408.50 ms │ no change │
│ QQuery 14 │ 1002.76 / 1006.56 ±3.43 / 1012.17 ms │ 985.35 / 1002.60 ±13.16 / 1022.62 ms │ no change │
│ QQuery 15 │ 14.87 / 15.91 ±0.77 / 17.21 ms │ 15.18 / 15.99 ±1.00 / 17.96 ms │ no change │
│ QQuery 16 │ 7.00 / 7.17 ±0.16 / 7.41 ms │ 7.01 / 7.58 ±0.47 / 8.20 ms │ 1.06x slower │
│ QQuery 17 │ 225.27 / 227.91 ±1.67 / 230.07 ms │ 224.97 / 226.08 ±0.83 / 226.99 ms │ no change │
│ QQuery 18 │ 125.60 / 127.72 ±1.21 / 129.06 ms │ 124.53 / 126.37 ±1.12 / 127.71 ms │ no change │
│ QQuery 19 │ 154.15 / 156.19 ±2.01 / 159.41 ms │ 153.93 / 155.22 ±1.33 / 157.75 ms │ no change │
│ QQuery 20 │ 13.40 / 13.78 ±0.47 / 14.69 ms │ 13.88 / 13.96 ±0.09 / 14.12 ms │ no change │
│ QQuery 21 │ 18.93 / 19.18 ±0.24 / 19.52 ms │ 18.96 / 19.82 ±1.26 / 22.32 ms │ no change │
│ QQuery 22 │ 477.95 / 482.52 ±3.65 / 487.44 ms │ 484.37 / 487.63 ±1.68 / 489.23 ms │ no change │
│ QQuery 23 │ 856.53 / 864.08 ±6.08 / 871.73 ms │ 863.41 / 869.79 ±4.24 / 876.20 ms │ no change │
│ QQuery 24 │ 377.02 / 380.14 ±2.73 / 384.73 ms │ 377.75 / 381.21 ±3.72 / 385.91 ms │ no change │
│ QQuery 25 │ 336.90 / 339.88 ±2.72 / 344.23 ms │ 338.35 / 340.07 ±1.57 / 342.90 ms │ no change │
│ QQuery 26 │ 81.10 / 81.94 ±0.54 / 82.61 ms │ 80.80 / 82.38 ±1.14 / 83.74 ms │ no change │
│ QQuery 27 │ 7.01 / 7.41 ±0.59 / 8.58 ms │ 6.74 / 7.11 ±0.30 / 7.54 ms │ no change │
│ QQuery 28 │ 147.64 / 149.26 ±2.14 / 153.41 ms │ 147.28 / 148.90 ±1.66 / 151.66 ms │ no change │
│ QQuery 29 │ 277.12 / 280.38 ±2.08 / 282.46 ms │ 277.36 / 279.11 ±1.14 / 280.56 ms │ no change │
│ QQuery 30 │ 43.44 / 45.29 ±1.78 / 48.47 ms │ 43.57 / 45.21 ±1.24 / 46.56 ms │ no change │
│ QQuery 31 │ 170.11 / 172.10 ±1.72 / 174.79 ms │ 169.40 / 170.48 ±0.99 / 171.87 ms │ no change │
│ QQuery 32 │ 56.03 / 57.57 ±1.29 / 59.25 ms │ 55.83 / 57.94 ±1.21 / 59.40 ms │ no change │
│ QQuery 33 │ 138.36 / 140.93 ±1.62 / 142.63 ms │ 140.50 / 141.28 ±0.88 / 142.75 ms │ no change │
│ QQuery 34 │ 6.87 / 7.12 ±0.25 / 7.56 ms │ 6.89 / 7.13 ±0.22 / 7.52 ms │ no change │
│ QQuery 35 │ 106.24 / 107.10 ±0.81 / 108.58 ms │ 106.46 / 107.24 ±0.84 / 108.86 ms │ no change │
│ QQuery 36 │ 6.64 / 7.43 ±1.11 / 9.64 ms │ 6.58 / 6.87 ±0.27 / 7.31 ms │ +1.08x faster │
│ QQuery 37 │ 8.02 / 8.52 ±0.36 / 9.08 ms │ 8.07 / 8.26 ±0.10 / 8.35 ms │ no change │
│ QQuery 38 │ 81.74 / 85.54 ±3.85 / 91.37 ms │ 83.79 / 86.19 ±3.24 / 92.38 ms │ no change │
│ QQuery 39 │ 122.30 / 125.89 ±2.87 / 128.90 ms │ 123.21 / 125.96 ±1.75 / 127.85 ms │ no change │
│ QQuery 40 │ 111.50 / 116.28 ±5.93 / 127.89 ms │ 109.04 / 113.46 ±5.48 / 123.63 ms │ no change │
│ QQuery 41 │ 14.19 / 15.49 ±1.24 / 17.45 ms │ 13.91 / 15.06 ±0.89 / 16.31 ms │ no change │
│ QQuery 42 │ 107.92 / 108.54 ±0.60 / 109.58 ms │ 107.06 / 108.87 ±0.96 / 109.74 ms │ no change │
│ QQuery 43 │ 6.06 / 6.15 ±0.12 / 6.39 ms │ 5.83 / 5.93 ±0.10 / 6.13 ms │ no change │
│ QQuery 44 │ 11.30 / 11.71 ±0.31 / 12.25 ms │ 11.10 / 11.61 ±0.41 / 12.28 ms │ no change │
│ QQuery 45 │ 51.08 / 51.85 ±0.52 / 52.55 ms │ 49.65 / 50.54 ±0.95 / 52.15 ms │ no change │
│ QQuery 46 │ 8.35 / 9.11 ±0.65 / 10.23 ms │ 8.30 / 8.60 ±0.22 / 8.89 ms │ +1.06x faster │
│ QQuery 47 │ 686.59 / 688.81 ±1.61 / 691.05 ms │ 688.66 / 691.35 ±2.62 / 695.13 ms │ no change │
│ QQuery 48 │ 280.27 / 291.45 ±5.78 / 296.95 ms │ 283.29 / 290.83 ±4.80 / 294.80 ms │ no change │
│ QQuery 49 │ 249.82 / 252.93 ±2.22 / 256.30 ms │ 251.25 / 253.15 ±1.55 / 255.02 ms │ no change │
│ QQuery 50 │ 220.14 / 224.24 ±3.04 / 227.79 ms │ 218.86 / 225.91 ±5.06 / 231.57 ms │ no change │
│ QQuery 51 │ 179.79 / 182.34 ±2.31 / 186.49 ms │ 178.45 / 183.48 ±3.54 / 186.47 ms │ no change │
│ QQuery 52 │ 106.62 / 107.98 ±1.02 / 109.23 ms │ 108.39 / 108.98 ±0.46 / 109.46 ms │ no change │
│ QQuery 53 │ 102.10 / 103.80 ±1.40 / 105.60 ms │ 101.05 / 102.50 ±1.17 / 104.52 ms │ no change │
│ QQuery 54 │ 144.70 / 146.20 ±0.98 / 147.68 ms │ 142.77 / 145.56 ±1.44 / 146.78 ms │ no change │
│ QQuery 55 │ 107.26 / 107.95 ±0.57 / 108.84 ms │ 105.34 / 106.83 ±1.08 / 108.57 ms │ no change │
│ QQuery 56 │ 140.33 / 141.66 ±1.28 / 143.23 ms │ 138.95 / 141.17 ±1.47 / 142.66 ms │ no change │
│ QQuery 57 │ 171.23 / 173.40 ±1.70 / 176.41 ms │ 170.15 / 173.72 ±3.05 / 179.26 ms │ no change │
│ QQuery 58 │ 295.93 / 306.35 ±10.47 / 326.49 ms │ 287.36 / 305.92 ±15.53 / 325.17 ms │ no change │
│ QQuery 59 │ 199.74 / 200.02 ±0.33 / 200.66 ms │ 198.41 / 200.33 ±1.66 / 203.07 ms │ no change │
│ QQuery 60 │ 143.43 / 144.39 ±0.57 / 144.99 ms │ 141.78 / 142.52 ±0.62 / 143.46 ms │ no change │
│ QQuery 61 │ 12.93 / 13.23 ±0.27 / 13.64 ms │ 13.00 / 13.27 ±0.30 / 13.67 ms │ no change │
│ QQuery 62 │ 890.72 / 928.37 ±31.81 / 971.13 ms │ 869.72 / 890.52 ±19.70 / 914.33 ms │ no change │
│ QQuery 63 │ 103.88 / 107.31 ±3.35 / 112.98 ms │ 101.49 / 105.63 ±2.63 / 109.45 ms │ no change │
│ QQuery 64 │ 679.51 / 683.78 ±3.65 / 688.57 ms │ 669.81 / 678.29 ±5.19 / 684.52 ms │ no change │
│ QQuery 65 │ 250.48 / 253.04 ±2.64 / 257.07 ms │ 246.53 / 248.36 ±1.85 / 251.63 ms │ no change │
│ QQuery 66 │ 236.34 / 254.47 ±11.61 / 271.97 ms │ 247.09 / 263.11 ±9.19 / 274.32 ms │ no change │
│ QQuery 67 │ 304.84 / 312.70 ±5.86 / 322.86 ms │ 302.55 / 311.98 ±6.76 / 323.13 ms │ no change │
│ QQuery 68 │ 8.50 / 9.83 ±0.81 / 10.68 ms │ 9.41 / 10.78 ±1.00 / 12.38 ms │ 1.10x slower │
│ QQuery 69 │ 101.29 / 103.02 ±1.66 / 106.16 ms │ 100.70 / 101.87 ±0.65 / 102.71 ms │ no change │
│ QQuery 70 │ 328.12 / 355.65 ±17.39 / 381.96 ms │ 343.32 / 356.93 ±10.54 / 373.42 ms │ no change │
│ QQuery 71 │ 136.01 / 136.36 ±0.27 / 136.81 ms │ 133.09 / 136.14 ±3.30 / 142.14 ms │ no change │
│ QQuery 72 │ 605.33 / 616.57 ±7.56 / 625.08 ms │ 602.24 / 615.69 ±9.64 / 631.08 ms │ no change │
│ QQuery 73 │ 7.49 / 8.34 ±0.82 / 9.72 ms │ 6.48 / 7.87 ±1.01 / 9.23 ms │ +1.06x faster │
│ QQuery 74 │ 536.94 / 542.79 ±6.35 / 555.07 ms │ 545.91 / 554.08 ±7.28 / 564.74 ms │ no change │
│ QQuery 75 │ 275.43 / 278.19 ±3.54 / 284.82 ms │ 276.32 / 276.90 ±0.61 / 277.88 ms │ no change │
│ QQuery 76 │ 132.31 / 133.34 ±1.01 / 135.22 ms │ 130.91 / 131.72 ±0.65 / 132.76 ms │ no change │
│ QQuery 77 │ 186.59 / 189.21 ±2.06 / 191.86 ms │ 187.58 / 189.50 ±2.07 / 192.95 ms │ no change │
│ QQuery 78 │ 337.66 / 343.06 ±2.95 / 346.53 ms │ 334.22 / 339.50 ±3.36 / 343.95 ms │ no change │
│ QQuery 79 │ 228.44 / 232.16 ±2.01 / 234.55 ms │ 229.79 / 232.10 ±1.86 / 235.21 ms │ no change │
│ QQuery 80 │ 321.59 / 324.20 ±2.62 / 329.10 ms │ 321.43 / 322.53 ±0.71 / 323.59 ms │ no change │
│ QQuery 81 │ 25.78 / 26.67 ±0.67 / 27.68 ms │ 25.09 / 26.34 ±1.38 / 28.97 ms │ no change │
│ QQuery 82 │ 196.84 / 202.34 ±2.98 / 205.62 ms │ 197.35 / 199.49 ±2.15 / 203.57 ms │ no change │
│ QQuery 83 │ 38.48 / 39.42 ±0.85 / 40.67 ms │ 37.86 / 38.31 ±0.29 / 38.73 ms │ no change │
│ QQuery 84 │ 47.22 / 48.36 ±0.58 / 48.83 ms │ 47.73 / 48.19 ±0.38 / 48.64 ms │ no change │
│ QQuery 85 │ 146.03 / 147.25 ±1.32 / 149.55 ms │ 143.02 / 147.14 ±2.18 / 148.97 ms │ no change │
│ QQuery 86 │ 38.64 / 39.13 ±0.26 / 39.41 ms │ 38.89 / 39.27 ±0.42 / 39.92 ms │ no change │
│ QQuery 87 │ 84.58 / 87.62 ±3.04 / 92.70 ms │ 83.82 / 87.05 ±3.25 / 92.88 ms │ no change │
│ QQuery 88 │ 99.42 / 100.49 ±1.19 / 102.65 ms │ 98.56 / 99.81 ±0.83 / 100.88 ms │ no change │
│ QQuery 89 │ 118.23 / 119.79 ±0.88 / 120.64 ms │ 116.34 / 117.65 ±1.28 / 119.51 ms │ no change │
│ QQuery 90 │ 23.26 / 23.62 ±0.30 / 23.93 ms │ 23.19 / 23.89 ±0.83 / 25.44 ms │ no change │
│ QQuery 91 │ 60.77 / 62.64 ±1.30 / 64.78 ms │ 62.48 / 63.51 ±1.37 / 65.99 ms │ no change │
│ QQuery 92 │ 57.15 / 58.23 ±1.14 / 60.08 ms │ 56.47 / 57.68 ±1.19 / 59.83 ms │ no change │
│ QQuery 93 │ 184.30 / 185.77 ±0.89 / 187.00 ms │ 184.36 / 186.26 ±1.27 / 187.80 ms │ no change │
│ QQuery 94 │ 60.43 / 61.07 ±0.44 / 61.63 ms │ 61.24 / 61.78 ±0.44 / 62.58 ms │ no change │
│ QQuery 95 │ 127.21 / 128.00 ±0.69 / 129.16 ms │ 126.44 / 127.49 ±0.93 / 128.62 ms │ no change │
│ QQuery 96 │ 69.13 / 73.89 ±2.56 / 76.59 ms │ 72.66 / 73.17 ±0.38 / 73.75 ms │ no change │
│ QQuery 97 │ 125.08 / 127.04 ±1.87 / 129.99 ms │ 122.37 / 124.69 ±1.60 / 126.91 ms │ no change │
│ QQuery 98 │ 151.54 / 154.40 ±1.92 / 156.93 ms │ 150.80 / 153.96 ±2.26 / 157.05 ms │ no change │
│ QQuery 99 │ 10805.99 / 10827.79 ±16.23 / 10856.43 ms │ 10741.67 / 10778.98 ±27.63 / 10827.59 ms │ no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary ┃ ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD) │ 31249.82ms │
│ Total Time (reorder-by-groupby-keys) │ 31190.69ms │
│ Average Time (HEAD) │ 315.65ms │
│ Average Time (reorder-by-groupby-keys) │ 315.06ms │
│ Queries Faster │ 3 │
│ Queries Slower │ 2 │
│ Queries with No Change │ 94 │
│ Queries with Failure │ 0 │
└────────────────────────────────────────┴────────────┘
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds — base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 156.6s |
| Peak memory | 5.2 GiB |
| Avg memory | 4.4 GiB |
| CPU user | 256.7s |
| CPU sys | 17.9s |
| Peak spill | 0 B |
**tpcds — branch**
| Metric | Value |
|--------|-------|
| Wall time | 156.2s |
| Peak memory | 5.2 GiB |
| Avg memory | 4.4 GiB |
| CPU user | 256.9s |
| CPU sys | 17.4s |
| 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4237121306
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236980088)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<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 reorder-by-groupby-keys
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query ┃ HEAD ┃ reorder-by-groupby-keys ┃ Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1 │ 6.70 / 7.19 ±0.82 / 8.82 ms │ 6.58 / 7.08 ±0.87 / 8.81 ms │ no change │
│ QQuery 2 │ 144.50 / 145.62 ±0.79 / 146.67 ms │ 145.45 / 147.08 ±1.21 / 148.80 ms │ no change │
│ QQuery 3 │ 114.00 / 115.19 ±0.76 / 115.92 ms │ 113.51 / 114.29 ±0.59 / 115.14 ms │ no change │
│ QQuery 4 │ 1358.59 / 1395.24 ±21.94 / 1416.44 ms │ 1389.17 / 1427.79 ±21.69 / 1455.17 ms │ no change │
│ QQuery 5 │ 174.30 / 175.22 ±0.65 / 176.04 ms │ 174.25 / 175.66 ±1.03 / 177.24 ms │ no change │
│ QQuery 6 │ 871.77 / 878.63 ±6.52 / 889.71 ms │ 887.05 / 902.84 ±17.73 / 937.10 ms │ no change │
│ QQuery 7 │ 341.85 / 345.17 ±3.01 / 349.04 ms │ 342.76 / 346.23 ±2.98 / 349.94 ms │ no change │
│ QQuery 8 │ 118.27 / 119.21 ±1.09 / 121.19 ms │ 116.12 / 118.72 ±1.50 / 120.72 ms │ no change │
│ QQuery 9 │ 101.59 / 104.90 ±2.67 / 108.21 ms │ 101.70 / 104.55 ±2.44 / 108.19 ms │ no change │
│ QQuery 10 │ 106.00 / 107.11 ±0.85 / 108.37 ms │ 106.99 / 107.89 ±0.94 / 109.30 ms │ no change │
│ QQuery 11 │ 974.73 / 992.75 ±12.47 / 1009.86 ms │ 968.85 / 1005.19 ±28.79 / 1041.30 ms │ no change │
│ QQuery 12 │ 45.75 / 47.36 ±1.37 / 49.19 ms │ 46.33 / 47.52 ±1.27 / 49.86 ms │ no change │
│ QQuery 13 │ 402.74 / 407.01 ±3.72 / 413.88 ms │ 400.75 / 403.62 ±2.31 / 406.42 ms │ no change │
│ QQuery 14 │ 994.27 / 1008.12 ±9.12 / 1021.99 ms │ 998.12 / 1011.08 ±7.38 / 1019.97 ms │ no change │
│ QQuery 15 │ 15.72 / 16.69 ±0.89 / 18.10 ms │ 15.76 / 16.81 ±0.98 / 18.61 ms │ no change │
│ QQuery 16 │ 7.25 / 7.67 ±0.59 / 8.79 ms │ 7.34 / 7.88 ±0.45 / 8.57 ms │ no change │
│ QQuery 17 │ 229.17 / 230.93 ±1.63 / 233.52 ms │ 230.43 / 232.51 ±1.08 / 233.37 ms │ no change │
│ QQuery 18 │ 126.51 / 127.98 ±1.43 / 130.36 ms │ 128.18 / 129.01 ±0.66 / 130.00 ms │ no change │
│ QQuery 19 │ 157.64 / 160.17 ±1.38 / 161.35 ms │ 154.16 / 157.50 ±2.07 / 160.44 ms │ no change │
│ QQuery 20 │ 14.79 / 15.09 ±0.21 / 15.40 ms │ 14.12 / 14.63 ±0.48 / 15.41 ms │ no change │
│ QQuery 21 │ 20.04 / 20.49 ±0.51 / 21.45 ms │ 20.06 / 20.63 ±0.86 / 22.30 ms │ no change │
│ QQuery 22 │ 493.66 / 494.82 ±0.78 / 495.84 ms │ 494.13 / 501.97 ±6.12 / 507.65 ms │ no change │
│ QQuery 23 │ 908.57 / 916.05 ±5.50 / 925.70 ms │ 892.52 / 907.02 ±9.70 / 920.67 ms │ no change │
│ QQuery 24 │ 388.39 / 391.76 ±2.09 / 394.46 ms │ 383.20 / 386.62 ±2.73 / 391.14 ms │ no change │
│ QQuery 25 │ 342.89 / 346.54 ±3.21 / 350.21 ms │ 338.80 / 342.19 ±2.27 / 345.77 ms │ no change │
│ QQuery 26 │ 82.22 / 83.81 ±1.80 / 87.13 ms │ 80.47 / 82.10 ±1.15 / 83.37 ms │ no change │
│ QQuery 27 │ 7.40 / 8.12 ±0.60 / 8.90 ms │ 6.83 / 7.28 ±0.30 / 7.70 ms │ +1.11x faster │
│ QQuery 28 │ 151.36 / 153.46 ±1.52 / 154.88 ms │ 149.12 / 150.41 ±1.65 / 153.51 ms │ no change │
│ QQuery 29 │ 283.38 / 286.69 ±2.29 / 289.95 ms │ 281.43 / 282.05 ±0.51 / 282.84 ms │ no change │
│ QQuery 30 │ 44.03 / 45.18 ±0.75 / 46.03 ms │ 43.15 / 45.37 ±1.27 / 46.40 ms │ no change │
│ QQuery 31 │ 173.79 / 174.97 ±1.17 / 177.17 ms │ 169.24 / 171.11 ±1.14 / 172.57 ms │ no change │
│ QQuery 32 │ 58.25 / 59.06 ±0.69 / 59.90 ms │ 57.20 / 57.78 ±0.67 / 58.69 ms │ no change │
│ QQuery 33 │ 141.16 / 142.83 ±1.32 / 144.46 ms │ 139.56 / 141.91 ±1.61 / 143.31 ms │ no change │
│ QQuery 34 │ 6.89 / 7.17 ±0.19 / 7.43 ms │ 7.10 / 7.26 ±0.18 / 7.59 ms │ no change │
│ QQuery 35 │ 108.14 / 109.97 ±0.93 / 110.61 ms │ 107.21 / 108.38 ±0.95 / 109.33 ms │ no change │
│ QQuery 36 │ 7.23 / 7.41 ±0.19 / 7.76 ms │ 6.63 / 6.78 ±0.15 / 7.01 ms │ +1.09x faster │
│ QQuery 37 │ 8.91 / 9.16 ±0.24 / 9.50 ms │ 8.49 / 8.86 ±0.42 / 9.49 ms │ no change │
│ QQuery 38 │ 85.34 / 88.62 ±3.41 / 94.71 ms │ 84.07 / 89.33 ±3.08 / 93.33 ms │ no change │
│ QQuery 39 │ 128.32 / 130.27 ±1.48 / 132.51 ms │ 123.72 / 126.00 ±1.50 / 127.78 ms │ no change │
│ QQuery 40 │ 114.74 / 119.31 ±5.61 / 130.26 ms │ 107.56 / 113.37 ±6.70 / 126.17 ms │ no change │
│ QQuery 41 │ 14.85 / 15.96 ±0.78 / 17.01 ms │ 14.97 / 15.81 ±0.84 / 17.37 ms │ no change │
│ QQuery 42 │ 110.18 / 111.33 ±0.82 / 112.29 ms │ 108.06 / 109.34 ±0.96 / 110.49 ms │ no change │
│ QQuery 43 │ 6.07 / 6.34 ±0.19 / 6.66 ms │ 6.23 / 6.40 ±0.20 / 6.77 ms │ no change │
│ QQuery 44 │ 12.57 / 13.64 ±1.18 / 15.69 ms │ 12.12 / 12.67 ±0.35 / 13.20 ms │ +1.08x faster │
│ QQuery 45 │ 51.77 / 52.68 ±0.90 / 54.25 ms │ 51.64 / 51.96 ±0.57 / 53.09 ms │ no change │
│ QQuery 46 │ 8.82 / 9.22 ±0.53 / 10.17 ms │ 8.55 / 8.79 ±0.16 / 8.92 ms │ no change │
│ QQuery 47 │ 764.23 / 777.68 ±12.15 / 796.75 ms │ 734.98 / 739.31 ±5.07 / 749.17 ms │ no change │
│ QQuery 48 │ 297.95 / 300.30 ±1.83 / 303.17 ms │ 289.36 / 297.59 ±5.56 / 306.33 ms │ no change │
│ QQuery 49 │ 255.53 / 258.72 ±2.49 / 262.28 ms │ 251.56 / 254.46 ±1.90 / 256.66 ms │ no change │
│ QQuery 50 │ 227.32 / 231.07 ±3.82 / 237.89 ms │ 218.89 / 228.78 ±5.28 / 234.67 ms │ no change │
│ QQuery 51 │ 184.53 / 185.90 ±1.21 / 187.65 ms │ 180.21 / 183.11 ±2.34 / 186.31 ms │ no change │
│ QQuery 52 │ 108.44 / 110.13 ±1.60 / 112.59 ms │ 109.35 / 110.25 ±0.66 / 111.08 ms │ no change │
│ QQuery 53 │ 104.34 / 104.84 ±0.92 / 106.68 ms │ 102.34 / 103.07 ±0.92 / 104.87 ms │ no change │
│ QQuery 54 │ 147.20 / 148.34 ±1.19 / 150.51 ms │ 143.71 / 145.92 ±1.31 / 147.22 ms │ no change │
│ QQuery 55 │ 108.57 / 109.47 ±0.76 / 110.80 ms │ 106.66 / 108.18 ±1.13 / 109.64 ms │ no change │
│ QQuery 56 │ 140.17 / 141.64 ±0.98 / 142.99 ms │ 139.74 / 142.19 ±2.21 / 146.04 ms │ no change │
│ QQuery 57 │ 173.49 / 175.27 ±1.04 / 176.39 ms │ 175.35 / 177.91 ±2.07 / 180.79 ms │ no change │
│ QQuery 58 │ 291.31 / 298.63 ±6.33 / 306.21 ms │ 288.99 / 294.07 ±4.46 / 299.83 ms │ no change │
│ QQuery 59 │ 199.04 / 201.11 ±2.34 / 205.06 ms │ 196.01 / 198.76 ±1.85 / 201.69 ms │ no change │
│ QQuery 60 │ 145.36 / 147.19 ±1.86 / 150.41 ms │ 142.99 / 145.70 ±1.43 / 147.27 ms │ no change │
│ QQuery 61 │ 13.46 / 13.71 ±0.21 / 14.07 ms │ 13.10 / 13.45 ±0.26 / 13.80 ms │ no change │
│ QQuery 62 │ 930.61 / 972.23 ±29.76 / 1021.65 ms │ 908.21 / 940.05 ±24.68 / 976.36 ms │ no change │
│ QQuery 63 │ 105.09 / 106.22 ±1.21 / 108.49 ms │ 103.86 / 107.19 ±1.87 / 109.25 ms │ no change │
│ QQuery 64 │ 689.63 / 692.29 ±2.42 / 696.74 ms │ 686.16 / 690.48 ±3.93 / 696.00 ms │ no change │
│ QQuery 65 │ 259.00 / 263.32 ±4.13 / 269.76 ms │ 253.27 / 257.56 ±3.75 / 263.50 ms │ no change │
│ QQuery 66 │ 230.68 / 254.14 ±11.88 / 262.15 ms │ 250.81 / 260.20 ±5.33 / 266.65 ms │ no change │
│ QQuery 67 │ 322.78 / 324.56 ±1.82 / 328.05 ms │ 312.75 / 318.49 ±6.50 / 331.05 ms │ no change │
│ QQuery 68 │ 10.54 / 11.75 ±0.73 / 12.79 ms │ 8.81 / 10.51 ±0.98 / 11.48 ms │ +1.12x faster │
│ QQuery 69 │ 101.72 / 103.07 ±0.96 / 104.47 ms │ 102.01 / 103.34 ±0.84 / 104.45 ms │ no change │
│ QQuery 70 │ 329.00 / 353.08 ±16.38 / 370.81 ms │ 338.93 / 351.76 ±11.60 / 372.40 ms │ no change │
│ QQuery 71 │ 137.80 / 139.54 ±1.36 / 141.13 ms │ 134.15 / 136.29 ±1.29 / 138.06 ms │ no change │
│ QQuery 72 │ 611.61 / 624.25 ±10.48 / 637.14 ms │ 602.37 / 622.03 ±13.84 / 640.93 ms │ no change │
│ QQuery 73 │ 7.22 / 8.52 ±1.17 / 10.49 ms │ 7.19 / 8.65 ±1.83 / 12.21 ms │ no change │
│ QQuery 74 │ 624.91 / 628.18 ±2.00 / 630.63 ms │ 602.57 / 612.26 ±6.14 / 621.53 ms │ no change │
│ QQuery 75 │ 279.10 / 280.79 ±1.71 / 284.03 ms │ 276.16 / 277.77 ±1.73 / 279.99 ms │ no change │
│ QQuery 76 │ 133.19 / 134.26 ±0.98 / 136.00 ms │ 131.89 / 133.93 ±1.88 / 136.59 ms │ no change │
│ QQuery 77 │ 189.56 / 191.19 ±1.27 / 192.84 ms │ 188.74 / 190.69 ±1.58 / 193.36 ms │ no change │
│ QQuery 78 │ 345.94 / 349.24 ±1.90 / 351.13 ms │ 342.32 / 346.40 ±2.67 / 350.45 ms │ no change │
│ QQuery 79 │ 240.39 / 242.43 ±1.47 / 244.41 ms │ 229.49 / 235.30 ±3.40 / 239.78 ms │ no change │
│ QQuery 80 │ 323.10 / 324.61 ±1.51 / 326.77 ms │ 319.87 / 322.32 ±1.89 / 324.92 ms │ no change │
│ QQuery 81 │ 26.58 / 27.48 ±1.08 / 29.60 ms │ 26.80 / 27.50 ±1.03 / 29.55 ms │ no change │
│ QQuery 82 │ 196.81 / 202.47 ±3.19 / 205.82 ms │ 198.73 / 200.73 ±1.85 / 204.25 ms │ no change │
│ QQuery 83 │ 39.23 / 40.26 ±0.96 / 41.90 ms │ 39.23 / 39.96 ±0.82 / 41.55 ms │ no change │
│ QQuery 84 │ 48.19 / 49.44 ±1.39 / 51.83 ms │ 48.49 / 49.30 ±0.62 / 50.34 ms │ no change │
│ QQuery 85 │ 149.48 / 151.34 ±1.81 / 154.69 ms │ 149.52 / 150.82 ±1.84 / 154.44 ms │ no change │
│ QQuery 86 │ 40.32 / 41.16 ±0.68 / 41.94 ms │ 39.39 / 40.06 ±0.43 / 40.65 ms │ no change │
│ QQuery 87 │ 85.28 / 88.14 ±3.15 / 94.24 ms │ 84.75 / 88.57 ±3.45 / 94.22 ms │ no change │
│ QQuery 88 │ 101.48 / 101.97 ±0.56 / 102.98 ms │ 100.13 / 101.00 ±0.78 / 102.30 ms │ no change │
│ QQuery 89 │ 119.55 / 121.97 ±1.38 / 123.49 ms │ 118.20 / 120.40 ±1.92 / 123.24 ms │ no change │
│ QQuery 90 │ 24.65 / 25.06 ±0.34 / 25.67 ms │ 24.29 / 25.15 ±1.01 / 26.97 ms │ no change │
│ QQuery 91 │ 63.67 / 64.43 ±0.49 / 65.08 ms │ 62.32 / 63.73 ±1.03 / 64.86 ms │ no change │
│ QQuery 92 │ 58.46 / 59.02 ±0.46 / 59.56 ms │ 57.73 / 59.77 ±2.11 / 63.69 ms │ no change │
│ QQuery 93 │ 186.11 / 188.92 ±1.64 / 190.81 ms │ 187.92 / 189.79 ±1.40 / 191.52 ms │ no change │
│ QQuery 94 │ 60.97 / 62.46 ±0.96 / 63.83 ms │ 61.24 / 62.60 ±1.04 / 64.45 ms │ no change │
│ QQuery 95 │ 127.44 / 128.84 ±0.78 / 129.79 ms │ 127.74 / 129.35 ±0.94 / 130.18 ms │ no change │
│ QQuery 96 │ 73.90 / 75.48 ±1.48 / 78.23 ms │ 74.33 / 75.45 ±1.21 / 76.99 ms │ no change │
│ QQuery 97 │ 126.69 / 128.80 ±1.45 / 130.95 ms │ 125.57 / 126.69 ±0.92 / 127.76 ms │ no change │
│ QQuery 98 │ 152.38 / 155.28 ±2.54 / 159.05 ms │ 155.71 / 157.28 ±1.60 / 160.27 ms │ no change │
│ QQuery 99 │ 10804.33 / 10852.56 ±25.46 / 10878.54 ms │ 10783.59 / 10814.13 ±20.32 / 10838.76 ms │ no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary ┃ ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD) │ 31984.85ms │
│ Total Time (reorder-by-groupby-keys) │ 31829.52ms │
│ Average Time (HEAD) │ 323.08ms │
│ Average Time (reorder-by-groupby-keys) │ 321.51ms │
│ Queries Faster │ 4 │
│ Queries Slower │ 0 │
│ Queries with No Change │ 95 │
│ Queries with Failure │ 0 │
└────────────────────────────────────────┴────────────┘
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds — base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 160.3s |
| Peak memory | 5.5 GiB |
| Avg memory | 4.5 GiB |
| CPU user | 264.4s |
| CPU sys | 17.5s |
| Peak spill | 0 B |
**tpcds — branch**
| Metric | Value |
|--------|-------|
| Wall time | 159.4s |
| Peak memory | 5.6 GiB |
| Avg memory | 4.7 GiB |
| CPU user | 263.0s |
| CPU sys | 17.1s |
| 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4237126988
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236980088)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<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 reorder-by-groupby-keys
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query ┃ HEAD ┃ reorder-by-groupby-keys ┃ Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0 │ 1.28 / 4.73 ±6.67 / 18.07 ms │ 1.25 / 4.62 ±6.52 / 17.66 ms │ no change │
│ QQuery 1 │ 14.56 / 14.97 ±0.21 / 15.13 ms │ 14.34 / 14.88 ±0.28 / 15.13 ms │ no change │
│ QQuery 2 │ 43.37 / 43.88 ±0.33 / 44.39 ms │ 43.20 / 43.87 ±0.36 / 44.22 ms │ no change │
│ QQuery 3 │ 43.29 / 44.56 ±0.99 / 46.25 ms │ 42.47 / 44.62 ±1.53 / 46.82 ms │ no change │
│ QQuery 4 │ 289.84 / 306.80 ±18.77 / 341.47 ms │ 288.89 / 312.34 ±18.60 / 334.32 ms │ no change │
│ QQuery 5 │ 343.42 / 354.00 ±10.53 / 373.18 ms │ 336.62 / 356.46 ±17.24 / 388.01 ms │ no change │
│ QQuery 6 │ 5.22 / 6.67 ±1.61 / 9.61 ms │ 5.12 / 6.58 ±0.97 / 7.80 ms │ no change │
│ QQuery 7 │ 17.48 / 17.78 ±0.27 / 18.09 ms │ 17.24 / 17.71 ±0.46 / 18.47 ms │ no change │
│ QQuery 8 │ 412.09 / 441.72 ±17.31 / 461.75 ms │ 423.80 / 446.54 ±15.37 / 465.94 ms │ no change │
│ QQuery 9 │ 637.31 / 667.00 ±19.74 / 686.46 ms │ 635.94 / 680.30 ±24.50 / 710.29 ms │ no change │
│ QQuery 10 │ 94.82 / 97.35 ±3.50 / 104.15 ms │ 92.79 / 96.10 ±3.25 / 102.31 ms │ no change │
│ QQuery 11 │ 101.65 / 103.05 ±1.51 / 105.76 ms │ 103.05 / 106.40 ±2.89 / 111.28 ms │ no change │
│ QQuery 12 │ 334.40 / 359.12 ±17.23 / 378.43 ms │ 339.15 / 361.45 ±13.98 / 378.21 ms │ no change │
│ QQuery 13 │ 497.12 / 503.54 ±5.83 / 512.89 ms │ 448.04 / 471.87 ±24.12 / 505.89 ms │ +1.07x faster │
│ QQuery 14 │ 342.51 / 355.48 ±14.81 / 383.50 ms │ 346.90 / 361.13 ±13.77 / 380.60 ms │ no change │
│ QQuery 15 │ 351.58 / 380.27 ±28.37 / 433.00 ms │ 360.30 / 388.60 ±25.96 / 435.29 ms │ no change │
│ QQuery 16 │ 720.74 / 751.33 ±23.00 / 783.01 ms │ 728.23 / 747.31 ±21.67 / 774.50 ms │ no change │
│ QQuery 17 │ 710.22 / 741.30 ±18.23 / 766.65 ms │ 731.01 / 744.34 ±9.70 / 758.50 ms │ no change │
│ QQuery 18 │ 1412.05 / 1508.58 ±49.48 / 1544.74 ms │ 1499.27 / 1530.60 ±30.63 / 1573.36 ms │ no change │
│ QQuery 19 │ 39.15 / 53.67 ±28.08 / 109.83 ms │ 36.49 / 39.12 ±2.84 / 44.29 ms │ +1.37x faster │
│ QQuery 20 │ 715.15 / 741.68 ±21.66 / 770.06 ms │ 729.02 / 738.92 ±13.20 / 765.09 ms │ no change │
│ QQuery 21 │ 761.25 / 781.63 ±12.34 / 796.07 ms │ 765.85 / 770.44 ±3.58 / 775.36 ms │ no change │
│ QQuery 22 │ 1138.07 / 1157.36 ±11.07 / 1169.13 ms │ 1140.09 / 1154.88 ±13.63 / 1173.98 ms │ no change │
│ QQuery 23 │ 3169.09 / 3224.13 ±36.93 / 3278.95 ms │ 3142.24 / 3222.97 ±46.19 / 3284.43 ms │ no change │
│ QQuery 24 │ 101.00 / 103.16 ±1.65 / 104.68 ms │ 99.04 / 107.30 ±5.16 / 114.87 ms │ no change │
│ QQuery 25 │ 138.28 / 143.32 ±2.92 / 146.15 ms │ 138.98 / 141.26 ±2.44 / 145.95 ms │ no change │
│ QQuery 26 │ 100.65 / 106.72 ±3.55 / 111.71 ms │ 98.64 / 106.32 ±4.19 / 111.12 ms │ no change │
│ QQuery 27 │ 852.44 / 857.38 ±2.64 / 859.48 ms │ 859.85 / 866.74 ±6.09 / 875.91 ms │ no change │
│ QQuery 28 │ 3281.08 / 3327.60 ±27.98 / 3364.11 ms │ 3306.47 / 3345.50 ±20.15 / 3363.08 ms │ no change │
│ QQuery 29 │ 50.22 / 56.37 ±5.39 / 65.88 ms │ 51.59 / 53.22 ±1.92 / 56.99 ms │ +1.06x faster │
│ QQuery 30 │ 361.03 / 372.00 ±12.43 / 395.10 ms │ 346.79 / 364.89 ±11.73 / 379.48 ms │ no change │
│ QQuery 31 │ 373.46 / 387.57 ±14.76 / 413.03 ms │ 365.72 / 382.61 ±17.52 / 415.28 ms │ no change │
│ QQuery 32 │ 1081.80 / 1113.80 ±29.61 / 1166.67 ms │ 1104.25 / 1250.81 ±97.38 / 1352.88 ms │ 1.12x slower │
│ QQuery 33 │ 1455.50 / 1488.09 ±29.06 / 1538.84 ms │ 1444.85 / 1490.95 ±49.62 / 1585.33 ms │ no change │
│ QQuery 34 │ 1544.65 / 1577.64 ±22.02 / 1609.81 ms │ 1444.94 / 1560.07 ±60.66 / 1625.88 ms │ no change │
│ QQuery 35 │ 453.78 / 462.31 ±5.30 / 468.06 ms │ 457.62 / 470.29 ±11.01 / 484.50 ms │ no change │
│ QQuery 36 │ 119.90 / 124.84 ±3.52 / 129.98 ms │ 119.02 / 128.43 ±5.14 / 132.81 ms │ no change │
│ QQuery 37 │ 48.13 / 50.19 ±1.32 / 52.25 ms │ 51.22 / 52.82 ±1.95 / 55.82 ms │ 1.05x slower │
│ QQuery 38 │ 76.60 / 78.82 ±1.30 / 80.14 ms │ 77.74 / 79.86 ±1.83 / 81.90 ms │ no change │
│ QQuery 39 │ 216.72 / 229.17 ±9.14 / 243.79 ms │ 209.64 / 227.17 ±13.81 / 246.54 ms │ no change │
│ QQuery 40 │ 23.18 / 26.44 ±2.45 / 28.84 ms │ 24.96 / 26.84 ±1.20 / 28.61 ms │ no change │
│ QQuery 41 │ 20.70 / 24.05 ±2.41 / 28.08 ms │ 21.15 / 22.06 ±0.84 / 23.64 ms │ +1.09x faster │
│ QQuery 42 │ 20.39 / 21.31 ±0.97 / 23.18 ms │ 19.82 / 20.76 ±0.71 / 21.64 ms │ no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary ┃ ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD) │ 23211.33ms │
│ Total Time (reorder-by-groupby-keys) │ 23359.96ms │
│ Average Time (HEAD) │ 539.80ms │
│ Average Time (reorder-by-groupby-keys) │ 543.25ms │
│ Queries Faster │ 4 │
│ Queries Slower │ 2 │
│ Queries with No Change │ 37 │
│ Queries with Failure │ 0 │
└────────────────────────────────────────┴────────────┘
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned — base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 117.3s |
| Peak memory | 45.8 GiB |
| Avg memory | 30.4 GiB |
| CPU user | 1103.3s |
| CPU sys | 87.1s |
| Peak spill | 0 B |
**clickbench_partitioned — branch**
| Metric | Value |
|--------|-------|
| Wall time | 117.9s |
| Peak memory | 40.2 GiB |
| Avg memory | 30.2 GiB |
| CPU user | 1105.5s |
| CPU sys | 91.4s |
| 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236608988
🤖 Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236474741)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<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 reorder-by-groupby-keys
--------------------
Benchmark clickbench_1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query ┃ HEAD ┃ reorder-by-groupby-keys ┃ Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0 │ 0.73 / 1.10 ±0.67 / 2.43 ms │ 0.74 / 1.10 ±0.63 / 2.36 ms │ no change │
│ QQuery 1 │ 13.57 / 14.13 ±0.98 / 16.09 ms │ 13.73 / 14.05 ±0.37 / 14.77 ms │ no change │
│ QQuery 2 │ 40.76 / 41.06 ±0.18 / 41.29 ms │ 41.16 / 41.26 ±0.10 / 41.39 ms │ no change │
│ QQuery 3 │ 37.53 / 38.43 ±1.13 / 40.41 ms │ 38.01 / 38.32 ±0.26 / 38.62 ms │ no change │
│ QQuery 4 │ 266.72 / 280.05 ±14.56 / 307.55 ms │ 268.36 / 273.93 ±4.96 / 283.10 ms │ no change │
│ QQuery 5 │ 455.62 / 464.89 ±6.40 / 473.16 ms │ 453.37 / 462.20 ±8.01 / 474.59 ms │ no change │
│ QQuery 6 │ 6.05 / 6.74 ±0.64 / 7.93 ms │ 6.12 / 6.35 ±0.27 / 6.76 ms │ +1.06x faster │
│ QQuery 7 │ 15.39 / 15.81 ±0.27 / 16.20 ms │ 15.68 / 15.93 ±0.13 / 16.02 ms │ no change │
│ QQuery 8 │ 391.61 / 402.82 ±7.71 / 412.11 ms │ 393.74 / 403.02 ±6.32 / 410.58 ms │ no change │
│ QQuery 9 │ 619.47 / 626.15 ±4.00 / 630.72 ms │ 622.97 / 631.98 ±5.67 / 638.24 ms │ no change │
│ QQuery 10 │ 90.16 / 92.52 ±2.64 / 97.60 ms │ 91.10 / 94.95 ±3.10 / 100.42 ms │ no change │
│ QQuery 11 │ 103.55 / 104.97 ±1.83 / 108.50 ms │ 103.47 / 104.64 ±1.29 / 106.97 ms │ no change │
│ QQuery 12 │ 464.34 / 475.31 ±6.65 / 482.49 ms │ 463.97 / 472.33 ±5.85 / 481.38 ms │ no change │
│ QQuery 13 │ 512.56 / 525.03 ±8.13 / 535.07 ms │ 523.35 / 536.53 ±6.90 / 543.31 ms │ no change │
│ QQuery 14 │ 461.30 / 474.90 ±8.83 / 485.73 ms │ 466.10 / 477.74 ±9.06 / 490.66 ms │ no change │
│ QQuery 15 │ 343.66 / 352.13 ±9.20 / 365.77 ms │ 339.62 / 353.15 ±11.35 / 372.03 ms │ no change │
│ QQuery 16 │ 713.75 / 737.51 ±15.14 / 759.17 ms │ 703.86 / 715.03 ±12.76 / 737.37 ms │ no change │
│ QQuery 17 │ 710.53 / 719.52 ±4.97 / 724.92 ms │ 701.00 / 714.06 ±8.98 / 728.47 ms │ no change │
│ QQuery 18 │ 1464.52 / 1481.91 ±21.83 / 1524.78 ms │ 1416.12 / 1437.62 ±10.92 / 1445.06 ms │ no change │
│ QQuery 19 │ 35.92 / 45.84 ±14.67 / 74.68 ms │ 38.83 / 52.71 ±22.98 / 98.42 ms │ 1.15x slower │
│ QQuery 20 │ 625.14 / 639.10 ±19.70 / 677.21 ms │ 623.54 / 644.04 ±25.98 / 691.05 ms │ no change │
│ QQuery 21 │ 706.10 / 710.00 ±3.43 / 716.10 ms │ 706.75 / 715.98 ±12.56 / 740.29 ms │ no change │
│ QQuery 22 │ 1355.79 / 1365.99 ±6.39 / 1375.03 ms │ 1356.20 / 1361.39 ±3.95 / 1367.21 ms │ no change │
│ QQuery 23 │ 3751.15 / 3770.73 ±24.55 / 3811.09 ms │ 3733.55 / 3749.05 ±9.81 / 3764.23 ms │ no change │
│ QQuery 24 │ 212.64 / 216.05 ±2.37 / 219.13 ms │ 211.80 / 219.53 ±6.42 / 230.94 ms │ no change │
│ QQuery 25 │ 181.76 / 184.47 ±1.61 / 186.82 ms │ 182.94 / 184.37 ±1.33 / 186.46 ms │ no change │
│ QQuery 26 │ 212.24 / 214.73 ±1.96 / 218.03 ms │ 213.92 / 217.43 ±2.86 / 222.58 ms │ no change │
│ QQuery 27 │ 745.43 / 749.74 ±4.52 / 758.03 ms │ 742.89 / 746.06 ±2.37 / 748.92 ms │ no change │
│ QQuery 28 │ 3447.37 / 3475.93 ±15.56 / 3491.92 ms │ 3435.77 / 3468.55 ±19.35 / 3489.32 ms │ no change │
│ QQuery 29 │ 46.85 / 50.81 ±3.52 / 57.18 ms │ 48.22 / 52.96 ±5.67 / 63.90 ms │ no change │
│ QQuery 30 │ 446.68 / 460.69 ±11.34 / 478.43 ms │ 442.66 / 450.90 ±10.28 / 470.24 ms │ no change │
│ QQuery 31 │ 429.22 / 452.60 ±16.85 / 477.50 ms │ 410.69 / 427.12 ±9.38 / 437.34 ms │ +1.06x faster │
│ QQuery 32 │ 1044.25 / 1063.08 ±23.60 / 1107.04 ms │ 1213.27 / 1248.20 ±36.45 / 1309.03 ms │ 1.17x slower │
│ QQuery 33 │ 1499.19 / 1513.73 ±8.74 / 1523.39 ms │ 1551.20 / 1587.30 ±29.42 / 1625.74 ms │ no change │
│ QQuery 34 │ 1538.08 / 1558.15 ±23.00 / 1599.68 ms │ 1528.42 / 1547.24 ±16.62 / 1569.87 ms │ no change │
│ QQuery 35 │ 369.16 / 371.73 ±2.13 / 374.96 ms │ 374.64 / 377.49 ±1.76 / 379.13 ms │ no change │
│ QQuery 36 │ 126.56 / 131.08 ±2.94 / 135.54 ms │ 127.81 / 144.52 ±24.49 / 193.00 ms │ 1.10x slower │
│ QQuery 37 │ 58.12 / 59.18 ±1.22 / 61.50 ms │ 55.95 / 59.41 ±2.66 / 63.53 ms │ no change │
│ QQuery 38 │ 86.02 / 87.80 ±0.98 / 88.90 ms │ 95.00 / 96.23 ±1.06 / 98.06 ms │ 1.10x slower │
│ QQuery 39 │ 230.33 / 235.62 ±6.05 / 247.00 ms │ 226.75 / 237.48 ±10.27 / 255.72 ms │ no change │
│ QQuery 40 │ 22.06 / 22.89 ±0.68 / 23.81 ms │ 23.21 / 25.40 ±1.78 / 28.00 ms │ 1.11x slower │
│ QQuery 41 │ 20.52 / 21.23 ±0.63 / 22.21 ms │ 20.65 / 21.60 ±0.94 / 22.95 ms │ no change │
│ QQuery 42 │ 19.87 / 20.76 ±1.00 / 22.55 ms │ 20.30 / 21.13 ±1.07 / 23.23 ms │ no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary ┃ ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD) │ 24276.92ms │
│ Total Time (reorder-by-groupby-keys) │ 24450.28ms │
│ Average Time (HEAD) │ 564.58ms │
│ Average Time (reorder-by-groupby-keys) │ 568.61ms │
│ Queries Faster │ 2 │
│ Queries Slower │ 5 │
│ Queries with No Change │ 36 │
│ Queries with Failure │ 0 │
└────────────────────────────────────────┴────────────┘
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_1 — base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 122.5s |
| Peak memory | 37.6 GiB |
| Avg memory | 28.4 GiB |
| CPU user | 1142.3s |
| CPU sys | 90.0s |
| Peak spill | 0 B |
**clickbench_1 — branch**
| Metric | Value |
|--------|-------|
| Wall time | 123.2s |
| Peak memory | 35.3 GiB |
| Avg memory | 27.1 GiB |
| CPU user | 1138.8s |
| CPU sys | 98.8s |
| 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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4237327362
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4237311314)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4237311314-1157-pr4tw 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (beb832507495d703815cfef410bcb5256bb4c15e) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..beb832507495d703815cfef410bcb5256bb4c15e) using: tpch
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236995961
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236980088)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236980088-1152-9qdk2 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (1c67f2912996a4f19a36cb31b43c32455b76873b) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..1c67f2912996a4f19a36cb31b43c32455b76873b) using: clickbench_partitioned
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236995733
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236980088)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236980088-1154-dnxgz 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (1c67f2912996a4f19a36cb31b43c32455b76873b) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..1c67f2912996a4f19a36cb31b43c32455b76873b) using: tpch
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236436817
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236421765)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236421765-1144-xprkn 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (6c143f75da587a8f623dd7b92ef85d5df975ca4f) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..6c143f75da587a8f623dd7b92ef85d5df975ca4f) using: tpch
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236437119
run benchmark clickbench
--
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236482401
🤖 Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21588#issuecomment-4236474741)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4236474741-1148-hwjsx 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 reorder-by-groupby-keys (4e09360b0adfea3b3cbb536b5fdd5d4b702ea73b) to 29c5dd5 (merge-base) [diff](https://github.com/apache/datafusion/compare/29c5dd5b8f7c60663ada64f34e15ecf22a7dc9a1..4e09360b0adfea3b3cbb536b5fdd5d4b702ea73b) using: clickbench_1
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] feat: reorder row groups by grouping key statistics [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21588:
URL: https://github.com/apache/datafusion/pull/21588#issuecomment-4236421765
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