You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@datafusion.apache.org by "jayzhan211 (via GitHub)" <gi...@apache.org> on 2026/10/16 23:53:22 UTC
[PR] fix: compact sparse string views in FilterExec coalescing [datafusion]
jayzhan211 opened a new pull request, #26436:
URL: https://github.com/apache/datafusion/pull/26436
## Which issue does this PR close?
- Fixes a regression from #25503 (not released yet, so it would ship in 56.0.0).
## Rationale for this change
Since #25503, a selective filter over `Utf8View`/`BinaryView` columns keeps the whole input string buffers alive. ClickBench Q21-Q23 use about 2-3x more memory and Q22/Q23 run 1.6-1.7x slower with the default `schema_force_view_types = true`.
#25503 made `FilterExec` call `BatchCoalescer::push_batch_with_filter` instead of `filter_record_batch` + `push_batch`. When at most 1/16 of the rows are selected, arrow-select 60 takes a sparse path that filters the views but reuses the source data buffers as they are (`InProgressByteViewArray::copy_rows_by_filter_from` calls `append_views_and_update_buffer_index(..., false)`). The old path went through `push_batch`, which sees that the filtered array uses less than half of its buffers (`need_gc`) and copies the selected strings into compact buffers. So each output batch now pins every data page it was filtered from, and operators that hold batches (aggregation, sort, TopK) hold all of them. arrow-rs main has the same code.
Repro (ClickBench `hits_partitioned`, default settings):
```sql
SET datafusion.execution.parquet.binary_as_string = true;
CREATE EXTERNAL TABLE hits STORED AS PARQUET LOCATION 'benchmarks/data/hits_partitioned';
SELECT * FROM hits WHERE "URL" LIKE '%google%' ORDER BY "EventTime" LIMIT 10; -- Q23
```
## What changes are included in this PR?
`LimitedBatchCoalescer::push_batch_with_filter` sends batches that have a `Utf8View` or `BinaryView` column with data buffers through `filter_record_batch` + `push_batch`, where arrow compacts sparse string views. All other batches keep the fused path from #25503.
This changes nothing for view batches when the filter is not sparse: arrow already materializes the filter and calls `push_batch` for them. So the only path that changes is the sparse one for view columns, which goes back to its behavior before #25503.
## What is the testing strategy for this PR?
- New unit test `test_push_batch_with_filter_compacts_string_views`: a 1/100 filter over 1600 long strings. Without the change the output keeps all 114,688 input string bytes; with it, less than a quarter.
- Benchmarks, release-nonlto, same commit (5f2158269e) with and without this PR, plus #25503 reverted for reference; median of 3 interleaved runs, peak RSS:
| Query | main | this PR | #25503 reverted |
|---|---|---|---|
| ClickBench Q21 | 0.60 s, 2.1 GB | 0.59 s, 1.1 GB | 0.62 s, 1.2 GB |
| ClickBench Q22 | 2.14 s, 4.1 GB | 1.40 s, 1.5 GB | 1.26 s, 1.4 GB |
| ClickBench Q23 | 7.10 s, 6.9 GB | 4.35 s, 3.4 GB | 4.33 s, 3.4 GB |
| ClickBench total (43 queries) | 29.8 s | 25.5 s | 25.9 s |
| TPC-H SF10 total (22 queries) | 6.89 s | 6.88 s | 6.93 s |
Other ClickBench queries differ only within run-to-run noise. TPC-H keeps the #25503 result.
## Are there any user-facing changes?
No.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6240597362
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/26436#issuecomment-6240443625)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing fix/filter-sparse-view-gc (7c633954f6e5f3792b9ca2eff964a97609cda729) to 5f21582 (merge-base) [diff](https://github.com/apache/datafusion/compare/5f2158269e24ac45e2740dcfe6e79a6f69f28330..7c633954f6e5f3792b9ca2eff964a97609cda729)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpcds
```
</details>
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and fix_filter-sparse-view-gc
--------------------
Benchmark tpcds_sf1.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β fix_filter-sparse-view-gc β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 5.13 ms β 5.16 ms β no change β
β QQuery 2 β 80.48 ms β 78.62 ms β no change β
β QQuery 3 β 28.03 ms β 28.11 ms β no change β
β QQuery 4 β 420.12 ms β 412.83 ms β no change β
β QQuery 5 β 48.83 ms β 48.60 ms β no change β
β QQuery 6 β 34.04 ms β 33.99 ms β no change β
β QQuery 7 β 70.29 ms β 71.02 ms β no change β
β QQuery 8 β 34.98 ms β 35.47 ms β no change β
β QQuery 9 β 51.85 ms β 51.41 ms β no change β
β QQuery 10 β 59.01 ms β 59.61 ms β no change β
β QQuery 11 β 255.95 ms β 262.36 ms β no change β
β QQuery 12 β 27.50 ms β 27.93 ms β no change β
β QQuery 13 β 111.42 ms β 110.63 ms β no change β
β QQuery 14 β 379.26 ms β 377.06 ms β no change β
β QQuery 15 β 51.73 ms β 51.05 ms β no change β
β QQuery 16 β 5.76 ms β 5.95 ms β no change β
β QQuery 17 β 76.54 ms β 76.32 ms β no change β
β QQuery 18 β 100.48 ms β 101.76 ms β no change β
β QQuery 19 β 40.49 ms β 40.05 ms β no change β
β QQuery 20 β 34.17 ms β 34.39 ms β no change β
β QQuery 21 β 16.40 ms β 16.33 ms β no change β
β QQuery 22 β 61.46 ms β 62.24 ms β no change β
β QQuery 23 β 298.15 ms β 297.59 ms β no change β
β QQuery 24 β 185.07 ms β 184.60 ms β no change β
β QQuery 25 β 105.42 ms β 104.69 ms β no change β
β QQuery 26 β 43.99 ms β 44.24 ms β no change β
β QQuery 27 β 5.48 ms β 5.54 ms β no change β
β QQuery 28 β 59.43 ms β 54.85 ms β +1.08x faster β
β QQuery 29 β 91.99 ms β 92.52 ms β no change β
β QQuery 30 β 30.85 ms β 30.98 ms β no change β
β QQuery 31 β 104.67 ms β 105.15 ms β no change β
β QQuery 32 β 19.02 ms β 19.05 ms β no change β
β QQuery 33 β 35.04 ms β 34.70 ms β no change β
β QQuery 34 β 9.55 ms β 9.71 ms β no change β
β QQuery 35 β 69.74 ms β 70.04 ms β no change β
β QQuery 36 β 5.44 ms β 5.41 ms β no change β
β QQuery 37 β 6.39 ms β 6.36 ms β no change β
β QQuery 38 β 58.69 ms β 58.93 ms β no change β
β QQuery 39 β 70.45 ms β 70.38 ms β no change β
β QQuery 40 β 21.95 ms β 22.10 ms β no change β
β QQuery 41 β 11.69 ms β 11.75 ms β no change β
β QQuery 42 β 22.95 ms β 22.87 ms β no change β
β QQuery 43 β 4.59 ms β 4.61 ms β no change β
β QQuery 44 β 8.26 ms β 8.23 ms β no change β
β QQuery 45 β 34.41 ms β 34.56 ms β no change β
β QQuery 46 β 10.95 ms β 10.99 ms β no change β
β QQuery 47 β 213.57 ms β 209.93 ms β no change β
β QQuery 48 β 88.52 ms β 88.54 ms β no change β
β QQuery 49 β 67.19 ms β 68.02 ms β no change β
β QQuery 50 β 56.64 ms β 57.06 ms β no change β
β QQuery 51 β 91.12 ms β 91.23 ms β no change β
β QQuery 52 β 23.09 ms β 23.15 ms β no change β
β QQuery 53 β 28.14 ms β 27.97 ms β no change β
β QQuery 54 β 51.52 ms β 51.55 ms β no change β
β QQuery 55 β 22.67 ms β 22.62 ms β no change β
β QQuery 56 β 36.56 ms β 36.38 ms β no change β
β QQuery 57 β 163.03 ms β 162.63 ms β no change β
β QQuery 58 β 104.91 ms β 103.98 ms β no change β
β QQuery 59 β 114.58 ms β 113.44 ms β no change β
β QQuery 60 β 36.56 ms β 36.97 ms β no change β
β QQuery 61 β 10.18 ms β 10.23 ms β no change β
β QQuery 62 β 43.05 ms β 42.55 ms β no change β
β QQuery 63 β 28.25 ms β 28.33 ms β no change β
β QQuery 64 β 348.80 ms β 349.48 ms β no change β
β QQuery 65 β 130.16 ms β 126.73 ms β no change β
β QQuery 66 β 73.90 ms β 73.28 ms β no change β
β QQuery 67 β 237.95 ms β 234.98 ms β no change β
β QQuery 68 β 11.12 ms β 11.39 ms β no change β
β QQuery 69 β 54.34 ms β 54.10 ms β no change β
β QQuery 70 β 102.00 ms β 102.04 ms β no change β
β QQuery 71 β 33.64 ms β 33.58 ms β no change β
β QQuery 72 β 1686.98 ms β 1584.13 ms β +1.06x faster β
β QQuery 73 β 9.40 ms β 9.59 ms β no change β
β QQuery 74 β 147.53 ms β 148.18 ms β no change β
β QQuery 75 β 135.68 ms β 136.45 ms β no change β
β QQuery 76 β 33.22 ms β 33.28 ms β no change β
β QQuery 77 β 57.93 ms β 57.68 ms β no change β
β QQuery 78 β 157.83 ms β 157.88 ms β no change β
β QQuery 79 β 64.99 ms β 64.94 ms β no change β
β QQuery 80 β 92.85 ms β 91.59 ms β no change β
β QQuery 81 β 24.47 ms β 24.39 ms β no change β
β QQuery 82 β 15.67 ms β 15.79 ms β no change β
β QQuery 83 β 31.54 ms β 31.04 ms β no change β
β QQuery 84 β 28.75 ms β 29.01 ms β no change β
β QQuery 85 β 97.01 ms β 96.36 ms β no change β
β QQuery 86 β 24.37 ms β 24.30 ms β no change β
β QQuery 87 β 58.52 ms β 59.06 ms β no change β
β QQuery 88 β 57.49 ms β 57.64 ms β no change β
β QQuery 89 β 34.14 ms β 34.42 ms β no change β
β QQuery 90 β 15.33 ms β 15.61 ms β no change β
β QQuery 91 β 39.39 ms β 39.66 ms β no change β
β QQuery 92 β 28.17 ms β 27.72 ms β no change β
β QQuery 93 β 48.30 ms β 47.73 ms β no change β
β QQuery 94 β 36.72 ms β 36.51 ms β no change β
β QQuery 95 β 77.90 ms β 77.93 ms β no change β
β QQuery 96 β 23.02 ms β 22.96 ms β no change β
β QQuery 97 β 50.17 ms β 49.61 ms β no change β
β QQuery 98 β 42.04 ms β 41.82 ms β no change β
β QQuery 99 β 64.30 ms β 63.46 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββββββ³ββββββββββββ
β Benchmark Summary β β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 8663.26ms β
β Total Time (fix_filter-sparse-view-gc) β 8539.62ms β
β Average Time (HEAD) β 87.51ms β
β Average Time (fix_filter-sparse-view-gc) β 86.26ms β
β Queries Faster β 2 β
β Queries Slower β 0 β
β Queries with No Change β 97 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and fix_filter-sparse-view-gc
--------------------
Benchmark tpcds_sf1.json
--------------------
βββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β fix_filter-sparse-view-gc β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 5.13 / 5.68 Β±1.05 / 7.79 ms β 5.16 / 5.70 Β±1.02 / 7.75 ms β no change β
β QQuery 2 β 80.48 / 80.75 Β±0.19 / 80.97 ms β 78.62 / 78.94 Β±0.24 / 79.37 ms β no change β
β QQuery 3 β 28.03 / 28.27 Β±0.19 / 28.47 ms β 28.11 / 28.30 Β±0.15 / 28.55 ms β no change β
β QQuery 4 β 420.12 / 425.21 Β±5.32 / 431.92 ms β 412.83 / 424.22 Β±6.80 / 433.78 ms β no change β
β QQuery 5 β 48.83 / 49.18 Β±0.23 / 49.52 ms β 48.60 / 49.09 Β±0.38 / 49.66 ms β no change β
β QQuery 6 β 34.04 / 34.55 Β±0.32 / 34.89 ms β 33.99 / 37.17 Β±4.96 / 47.06 ms β 1.08x slower β
β QQuery 7 β 70.29 / 71.46 Β±0.62 / 71.98 ms β 71.02 / 71.40 Β±0.29 / 71.87 ms β no change β
β QQuery 8 β 34.98 / 35.67 Β±0.59 / 36.75 ms β 35.47 / 35.63 Β±0.14 / 35.83 ms β no change β
β QQuery 9 β 51.85 / 52.91 Β±0.80 / 54.16 ms β 51.41 / 52.70 Β±0.75 / 53.47 ms β no change β
β QQuery 10 β 59.01 / 61.22 Β±2.71 / 66.56 ms β 59.61 / 60.64 Β±1.52 / 63.62 ms β no change β
β QQuery 11 β 255.95 / 260.08 Β±2.96 / 264.54 ms β 262.36 / 267.66 Β±4.69 / 273.91 ms β no change β
β QQuery 12 β 27.50 / 28.09 Β±0.62 / 29.29 ms β 27.93 / 28.33 Β±0.35 / 28.98 ms β no change β
β QQuery 13 β 111.42 / 113.53 Β±3.08 / 119.63 ms β 110.63 / 113.35 Β±2.50 / 117.71 ms β no change β
β QQuery 14 β 379.26 / 383.42 Β±4.81 / 390.47 ms β 377.06 / 384.52 Β±4.47 / 390.48 ms β no change β
β QQuery 15 β 51.73 / 52.93 Β±1.52 / 55.88 ms β 51.05 / 52.21 Β±1.27 / 54.68 ms β no change β
β QQuery 16 β 5.76 / 5.97 Β±0.27 / 6.50 ms β 5.95 / 6.07 Β±0.18 / 6.44 ms β no change β
β QQuery 17 β 76.54 / 76.76 Β±0.32 / 77.38 ms β 76.32 / 77.72 Β±0.93 / 79.11 ms β no change β
β QQuery 18 β 100.48 / 104.29 Β±3.66 / 110.90 ms β 101.76 / 102.69 Β±0.96 / 104.50 ms β no change β
β QQuery 19 β 40.49 / 40.92 Β±0.44 / 41.66 ms β 40.05 / 40.34 Β±0.46 / 41.23 ms β no change β
β QQuery 20 β 34.17 / 34.66 Β±0.30 / 35.03 ms β 34.39 / 34.66 Β±0.29 / 35.12 ms β no change β
β QQuery 21 β 16.40 / 16.64 Β±0.25 / 17.07 ms β 16.33 / 16.53 Β±0.19 / 16.77 ms β no change β
β QQuery 22 β 61.46 / 62.76 Β±0.86 / 63.92 ms β 62.24 / 63.26 Β±0.68 / 64.12 ms β no change β
β QQuery 23 β 298.15 / 302.72 Β±4.15 / 309.21 ms β 297.59 / 302.90 Β±3.85 / 307.17 ms β no change β
β QQuery 24 β 185.07 / 187.35 Β±2.80 / 192.71 ms β 184.60 / 186.34 Β±1.12 / 188.14 ms β no change β
β QQuery 25 β 105.42 / 106.10 Β±0.84 / 107.65 ms β 104.69 / 105.63 Β±0.93 / 107.38 ms β no change β
β QQuery 26 β 43.99 / 44.33 Β±0.19 / 44.54 ms β 44.24 / 44.75 Β±0.39 / 45.27 ms β no change β
β QQuery 27 β 5.48 / 6.68 Β±1.94 / 10.54 ms β 5.54 / 5.78 Β±0.20 / 6.16 ms β +1.15x faster β
β QQuery 28 β 59.43 / 59.90 Β±0.28 / 60.25 ms β 54.85 / 58.86 Β±2.01 / 60.10 ms β no change β
β QQuery 29 β 91.99 / 93.26 Β±1.18 / 95.18 ms β 92.52 / 93.95 Β±1.79 / 97.48 ms β no change β
β QQuery 30 β 30.85 / 31.24 Β±0.23 / 31.46 ms β 30.98 / 31.47 Β±0.48 / 32.27 ms β no change β
β QQuery 31 β 104.67 / 106.49 Β±2.38 / 111.20 ms β 105.15 / 106.08 Β±1.16 / 108.33 ms β no change β
β QQuery 32 β 19.02 / 19.42 Β±0.26 / 19.78 ms β 19.05 / 19.30 Β±0.23 / 19.68 ms β no change β
β QQuery 33 β 35.04 / 35.78 Β±0.84 / 37.39 ms β 34.70 / 35.51 Β±0.48 / 36.20 ms β no change β
β QQuery 34 β 9.55 / 10.12 Β±0.68 / 11.37 ms β 9.71 / 10.15 Β±0.57 / 11.19 ms β no change β
β QQuery 35 β 69.74 / 70.63 Β±0.89 / 72.09 ms β 70.04 / 70.66 Β±0.77 / 72.17 ms β no change β
β QQuery 36 β 5.44 / 5.58 Β±0.21 / 6.00 ms β 5.41 / 5.58 Β±0.24 / 6.06 ms β no change β
β QQuery 37 β 6.39 / 6.50 Β±0.11 / 6.71 ms β 6.36 / 6.53 Β±0.15 / 6.77 ms β no change β
β QQuery 38 β 58.69 / 59.43 Β±0.66 / 60.67 ms β 58.93 / 59.24 Β±0.23 / 59.52 ms β no change β
β QQuery 39 β 70.45 / 71.54 Β±1.35 / 74.11 ms β 70.38 / 71.75 Β±1.71 / 75.13 ms β no change β
β QQuery 40 β 21.95 / 22.34 Β±0.43 / 23.13 ms β 22.10 / 22.31 Β±0.19 / 22.60 ms β no change β
β QQuery 41 β 11.69 / 12.60 Β±1.42 / 15.44 ms β 11.75 / 11.88 Β±0.12 / 12.12 ms β +1.06x faster β
β QQuery 42 β 22.95 / 23.08 Β±0.17 / 23.38 ms β 22.87 / 23.52 Β±0.98 / 25.47 ms β no change β
β QQuery 43 β 4.59 / 4.74 Β±0.23 / 5.19 ms β 4.61 / 4.71 Β±0.13 / 4.97 ms β no change β
β QQuery 44 β 8.26 / 8.42 Β±0.15 / 8.66 ms β 8.23 / 8.37 Β±0.08 / 8.45 ms β no change β
β QQuery 45 β 34.41 / 35.39 Β±0.73 / 36.04 ms β 34.56 / 35.18 Β±0.59 / 36.22 ms β no change β
β QQuery 46 β 10.95 / 11.15 Β±0.14 / 11.38 ms β 10.99 / 11.20 Β±0.22 / 11.58 ms β no change β
β QQuery 47 β 213.57 / 216.10 Β±2.03 / 219.16 ms β 209.93 / 213.00 Β±2.03 / 215.22 ms β no change β
β QQuery 48 β 88.52 / 89.67 Β±1.09 / 91.68 ms β 88.54 / 89.28 Β±0.48 / 89.94 ms β no change β
β QQuery 49 β 67.19 / 68.98 Β±1.74 / 72.19 ms β 68.02 / 68.45 Β±0.55 / 69.52 ms β no change β
β QQuery 50 β 56.64 / 56.83 Β±0.29 / 57.40 ms β 57.06 / 57.39 Β±0.32 / 58.00 ms β no change β
β QQuery 51 β 91.12 / 93.21 Β±2.42 / 97.93 ms β 91.23 / 92.62 Β±1.01 / 94.11 ms β no change β
β QQuery 52 β 23.09 / 23.57 Β±0.26 / 23.84 ms β 23.15 / 23.98 Β±1.58 / 27.15 ms β no change β
β QQuery 53 β 28.14 / 28.41 Β±0.19 / 28.72 ms β 27.97 / 28.35 Β±0.23 / 28.57 ms β no change β
β QQuery 54 β 51.52 / 51.88 Β±0.24 / 52.28 ms β 51.55 / 51.98 Β±0.25 / 52.21 ms β no change β
β QQuery 55 β 22.67 / 22.81 Β±0.12 / 22.98 ms β 22.62 / 22.80 Β±0.12 / 22.95 ms β no change β
β QQuery 56 β 36.56 / 37.55 Β±1.19 / 39.76 ms β 36.38 / 36.72 Β±0.22 / 37.02 ms β no change β
β QQuery 57 β 163.03 / 165.08 Β±2.78 / 170.54 ms β 162.63 / 163.78 Β±0.79 / 165.04 ms β no change β
β QQuery 58 β 104.91 / 106.86 Β±1.52 / 108.72 ms β 103.98 / 105.38 Β±1.16 / 107.40 ms β no change β
β QQuery 59 β 114.58 / 115.12 Β±0.47 / 115.84 ms β 113.44 / 114.45 Β±1.35 / 117.12 ms β no change β
β QQuery 60 β 36.56 / 37.37 Β±0.54 / 38.06 ms β 36.97 / 37.49 Β±0.85 / 39.17 ms β no change β
β QQuery 61 β 10.18 / 10.82 Β±0.96 / 12.72 ms β 10.23 / 10.32 Β±0.14 / 10.59 ms β no change β
β QQuery 62 β 43.05 / 43.29 Β±0.18 / 43.60 ms β 42.55 / 43.28 Β±1.08 / 45.42 ms β no change β
β QQuery 63 β 28.25 / 28.65 Β±0.58 / 29.79 ms β 28.33 / 28.58 Β±0.16 / 28.81 ms β no change β
β QQuery 64 β 348.80 / 353.87 Β±5.35 / 363.49 ms β 349.48 / 354.15 Β±5.50 / 364.74 ms β no change β
β QQuery 65 β 130.16 / 131.71 Β±0.93 / 132.82 ms β 126.73 / 130.97 Β±2.54 / 134.75 ms β no change β
β QQuery 66 β 73.90 / 75.45 Β±1.65 / 78.60 ms β 73.28 / 74.79 Β±1.57 / 77.65 ms β no change β
β QQuery 67 β 237.95 / 243.34 Β±4.39 / 248.01 ms β 234.98 / 240.98 Β±4.14 / 246.65 ms β no change β
β QQuery 68 β 11.12 / 11.47 Β±0.38 / 12.18 ms β 11.39 / 11.54 Β±0.16 / 11.84 ms β no change β
β QQuery 69 β 54.34 / 55.11 Β±0.86 / 56.71 ms β 54.10 / 55.01 Β±0.89 / 56.57 ms β no change β
β QQuery 70 β 102.00 / 104.82 Β±3.79 / 112.18 ms β 102.04 / 105.21 Β±4.08 / 113.03 ms β no change β
β QQuery 71 β 33.64 / 34.11 Β±0.45 / 34.81 ms β 33.58 / 34.06 Β±0.57 / 35.01 ms β no change β
β QQuery 72 β 1686.98 / 1702.76 Β±10.57 / 1719.97 ms β 1584.13 / 1663.16 Β±39.98 / 1691.15 ms β no change β
β QQuery 73 β 9.40 / 9.78 Β±0.46 / 10.65 ms β 9.59 / 9.88 Β±0.27 / 10.23 ms β no change β
β QQuery 74 β 147.53 / 150.92 Β±2.46 / 155.07 ms β 148.18 / 151.19 Β±2.79 / 156.15 ms β no change β
β QQuery 75 β 135.68 / 138.79 Β±4.43 / 147.56 ms β 136.45 / 137.84 Β±1.15 / 139.60 ms β no change β
β QQuery 76 β 33.22 / 33.41 Β±0.13 / 33.54 ms β 33.28 / 33.45 Β±0.14 / 33.64 ms β no change β
β QQuery 77 β 57.93 / 58.27 Β±0.28 / 58.65 ms β 57.68 / 58.24 Β±0.45 / 58.94 ms β no change β
β QQuery 78 β 157.83 / 160.52 Β±2.47 / 164.58 ms β 157.88 / 160.08 Β±2.54 / 164.72 ms β no change β
β QQuery 79 β 64.99 / 66.09 Β±1.08 / 68.03 ms β 64.94 / 65.75 Β±0.93 / 67.53 ms β no change β
β QQuery 80 β 92.85 / 96.12 Β±4.05 / 103.97 ms β 91.59 / 93.51 Β±1.05 / 94.75 ms β no change β
β QQuery 81 β 24.47 / 24.72 Β±0.21 / 25.00 ms β 24.39 / 24.74 Β±0.22 / 24.93 ms β no change β
β QQuery 82 β 15.67 / 15.90 Β±0.14 / 16.07 ms β 15.79 / 16.08 Β±0.38 / 16.82 ms β no change β
β QQuery 83 β 31.54 / 32.71 Β±1.80 / 36.30 ms β 31.04 / 31.43 Β±0.26 / 31.84 ms β no change β
β QQuery 84 β 28.75 / 28.98 Β±0.13 / 29.15 ms β 29.01 / 29.14 Β±0.10 / 29.26 ms β no change β
β QQuery 85 β 97.01 / 98.22 Β±1.62 / 101.39 ms β 96.36 / 99.99 Β±3.80 / 105.53 ms β no change β
β QQuery 86 β 24.37 / 24.59 Β±0.28 / 25.12 ms β 24.30 / 24.60 Β±0.31 / 25.16 ms β no change β
β QQuery 87 β 58.52 / 59.10 Β±0.33 / 59.45 ms β 59.06 / 59.42 Β±0.40 / 60.15 ms β no change β
β QQuery 88 β 57.49 / 58.80 Β±1.54 / 61.81 ms β 57.64 / 58.00 Β±0.24 / 58.34 ms β no change β
β QQuery 89 β 34.14 / 34.89 Β±0.54 / 35.81 ms β 34.42 / 36.62 Β±3.69 / 43.97 ms β no change β
β QQuery 90 β 15.33 / 15.61 Β±0.24 / 15.94 ms β 15.61 / 15.95 Β±0.29 / 16.48 ms β no change β
β QQuery 91 β 39.39 / 39.81 Β±0.23 / 40.10 ms β 39.66 / 40.12 Β±0.38 / 40.61 ms β no change β
β QQuery 92 β 28.17 / 29.37 Β±1.50 / 32.32 ms β 27.72 / 28.57 Β±0.65 / 29.64 ms β no change β
β QQuery 93 β 48.30 / 49.60 Β±1.35 / 52.16 ms β 47.73 / 49.57 Β±2.35 / 54.20 ms β no change β
β QQuery 94 β 36.72 / 37.53 Β±0.91 / 39.20 ms β 36.51 / 36.98 Β±0.38 / 37.67 ms β no change β
β QQuery 95 β 77.90 / 78.31 Β±0.38 / 78.94 ms β 77.93 / 78.81 Β±0.99 / 80.06 ms β no change β
β QQuery 96 β 23.02 / 23.73 Β±1.22 / 26.16 ms β 22.96 / 23.22 Β±0.21 / 23.50 ms β no change β
β QQuery 97 β 50.17 / 51.21 Β±0.66 / 51.96 ms β 49.61 / 50.72 Β±0.99 / 52.03 ms β no change β
β QQuery 98 β 42.04 / 43.82 Β±1.79 / 47.05 ms β 41.82 / 42.82 Β±0.64 / 43.75 ms β no change β
β QQuery 99 β 64.30 / 64.56 Β±0.25 / 65.01 ms β 63.46 / 63.96 Β±0.34 / 64.50 ms β no change β
βββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββββββ³ββββββββββββ
β Benchmark Summary β β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 8793.92ms β
β Total Time (fix_filter-sparse-view-gc) β 8741.20ms β
β Average Time (HEAD) β 88.83ms β
β Average Time (fix_filter-sparse-view-gc) β 88.29ms β
β Queries Faster β 2 β
β Queries Slower β 1 β
β Queries with No Change β 96 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 45.0s |
| Peak memory | 1.9 GiB |
| Avg memory | 1.3 GiB |
| CPU user | 188.3s |
| CPU sys | 5.0s |
| Peak spill | 0 B |
**tpcds β branch**
| Metric | Value |
|--------|-------|
| Wall time | 45.0s |
| Peak memory | 1.9 GiB |
| Avg memory | 1.3 GiB |
| CPU user | 186.4s |
| CPU sys | 5.0s |
| 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] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "jayzhan211 (via GitHub)" <gi...@apache.org>.
jayzhan211 commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6290293796
> One optional idea: arrow only keeps the source buffers on its sparse path (at most 1/16 selected), so the fallback could also check `filter.true_count() * 16 <= filter.len()` and keep the fused path for dense filters.
Thanks for the review! I looked into this, and I don't think the extra check would change what runs for dense filters. When more than 1/16 of the rows are selected, arrow's `push_batch_with_filter` already filters the batch and then calls `push_batch` ([arrow-select 60.0.0, coalesce.rs L652-L665](https://github.com/apache/arrow-rs/blob/60.0.0/arrow-select/src/coalesce.rs#L652-L665)). That's the same thing the fallback here does, so `WHERE "SearchPhrase" <> ''` runs the same code with or without the check.
I think the Q13 and Q14 numbers are runner noise. Q4, Q5 and Q18 have no `WHERE` clause, so they never reach this code, but they moved by about the same amount (7-11% slower) in the same run. On my machine, Q13 was 440-478 ms on main and 435-469 ms with this PR across three interleaved runs.
I'd also rather not copy arrow's private 1/16 threshold here. If arrow changes it, the two checks could drift apart and the leak could come back for filters in between. Since this fallback goes away once apache/arrow-rs#11635 is fixed, I'd keep it as is. Happy to add the check if you still prefer it, though.
--
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] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "jayzhan211 (via GitHub)" <gi...@apache.org>.
jayzhan211 commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6226980567
@Rich-T-kid could you take a look at this? If it makes sense to you, we could merge this first, then fix the root cause in arrow-rs (apache/arrow-rs#11635), and remove the change in this PR once the arrow fix is released.
--
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] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6240654493
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/26436#issuecomment-6240443625)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing fix/filter-sparse-view-gc (7c633954f6e5f3792b9ca2eff964a97609cda729) to 5f21582 (merge-base) [diff](https://github.com/apache/datafusion/compare/5f2158269e24ac45e2740dcfe6e79a6f69f28330..7c633954f6e5f3792b9ca2eff964a97609cda729)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
```
</details>
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and fix_filter-sparse-view-gc
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β fix_filter-sparse-view-gc β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.23 ms β 1.25 ms β no change β
β QQuery 1 β 11.66 ms β 11.90 ms β no change β
β QQuery 2 β 36.37 ms β 36.44 ms β no change β
β QQuery 3 β 31.42 ms β 32.17 ms β no change β
β QQuery 4 β 241.19 ms β 267.96 ms β 1.11x slower β
β QQuery 5 β 275.38 ms β 301.55 ms β 1.10x slower β
β QQuery 6 β 1.32 ms β 1.41 ms β 1.07x slower β
β QQuery 7 β 13.13 ms β 13.44 ms β no change β
β QQuery 8 β 380.80 ms β 366.23 ms β no change β
β QQuery 9 β 523.57 ms β 516.09 ms β no change β
β QQuery 10 β 65.80 ms β 66.98 ms β no change β
β QQuery 11 β 77.45 ms β 77.87 ms β no change β
β QQuery 12 β 284.18 ms β 278.00 ms β no change β
β QQuery 13 β 377.60 ms β 402.81 ms β 1.07x slower β
β QQuery 14 β 284.59 ms β 304.99 ms β 1.07x slower β
β QQuery 15 β 307.23 ms β 299.06 ms β no change β
β QQuery 16 β 680.67 ms β 630.35 ms β +1.08x faster β
β QQuery 17 β 683.80 ms β 664.47 ms β no change β
β QQuery 18 β 1279.87 ms β 1407.89 ms β 1.10x slower β
β QQuery 19 β 27.50 ms β 28.73 ms β no change β
β QQuery 20 β 526.27 ms β 528.20 ms β no change β
β QQuery 21 β 508.37 ms β 518.17 ms β no change β
β QQuery 22 β 1025.42 ms β 1001.87 ms β no change β
β QQuery 23 β 3394.65 ms β 3238.15 ms β no change β
β QQuery 24 β 42.86 ms β 41.24 ms β no change β
β QQuery 25 β 107.65 ms β 107.06 ms β no change β
β QQuery 26 β 41.82 ms β 41.19 ms β no change β
β QQuery 27 β 522.41 ms β 531.17 ms β no change β
β QQuery 28 β 2907.86 ms β 2938.60 ms β no change β
β QQuery 29 β 39.80 ms β 40.23 ms β no change β
β QQuery 30 β 325.22 ms β 335.32 ms β no change β
β QQuery 31 β 287.74 ms β 303.62 ms β 1.06x slower β
β QQuery 32 β 1332.01 ms β 962.56 ms β +1.38x faster β
β QQuery 33 β 1798.06 ms β 1682.05 ms β +1.07x faster β
β QQuery 34 β 1778.79 ms β 1669.51 ms β +1.07x faster β
β QQuery 35 β 357.90 ms β 331.03 ms β +1.08x faster β
β QQuery 36 β 71.73 ms β 70.49 ms β no change β
β QQuery 37 β 37.83 ms β 37.46 ms β no change β
β QQuery 38 β 42.56 ms β 41.11 ms β no change β
β QQuery 39 β 157.79 ms β 153.37 ms β no change β
β QQuery 40 β 15.44 ms β 14.91 ms β no change β
β QQuery 41 β 14.59 ms β 14.16 ms β no change β
β QQuery 42 β 14.23 ms β 13.74 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 20935.77ms β
β Total Time (fix_filter-sparse-view-gc) β 20324.79ms β
β Average Time (HEAD) β 486.88ms β
β Average Time (fix_filter-sparse-view-gc) β 472.67ms β
β Queries Faster β 5 β
β Queries Slower β 7 β
β Queries with No Change β 31 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and fix_filter-sparse-view-gc
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³βββββββββββββββββββββββββββββββββββββββββ³βββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β fix_filter-sparse-view-gc β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.23 / 4.42 Β±6.25 / 16.91 ms β 1.25 / 4.54 Β±6.37 / 17.28 ms β no change β
β QQuery 1 β 11.66 / 12.06 Β±0.21 / 12.29 ms β 11.90 / 12.13 Β±0.23 / 12.46 ms β no change β
β QQuery 2 β 36.37 / 36.70 Β±0.26 / 37.08 ms β 36.44 / 36.96 Β±0.37 / 37.57 ms β no change β
β QQuery 3 β 31.42 / 31.98 Β±0.65 / 33.24 ms β 32.17 / 33.00 Β±0.53 / 33.55 ms β no change β
β QQuery 4 β 241.19 / 245.43 Β±3.06 / 249.92 ms β 267.96 / 272.89 Β±3.72 / 277.52 ms β 1.11x slower β
β QQuery 5 β 275.38 / 285.30 Β±5.61 / 292.60 ms β 301.55 / 305.03 Β±3.09 / 310.19 ms β 1.07x slower β
β QQuery 6 β 1.32 / 1.47 Β±0.26 / 1.98 ms β 1.41 / 1.59 Β±0.20 / 1.97 ms β 1.08x slower β
β QQuery 7 β 13.13 / 15.77 Β±4.56 / 24.87 ms β 13.44 / 13.81 Β±0.20 / 14.03 ms β +1.14x faster β
β QQuery 8 β 380.80 / 382.56 Β±1.54 / 384.70 ms β 366.23 / 375.71 Β±6.14 / 382.01 ms β no change β
β QQuery 9 β 523.57 / 533.89 Β±8.37 / 542.51 ms β 516.09 / 526.84 Β±7.63 / 535.81 ms β no change β
β QQuery 10 β 65.80 / 67.20 Β±0.86 / 68.37 ms β 66.98 / 67.74 Β±0.52 / 68.32 ms β no change β
β QQuery 11 β 77.45 / 78.63 Β±0.68 / 79.42 ms β 77.87 / 79.30 Β±1.18 / 81.11 ms β no change β
β QQuery 12 β 284.18 / 291.97 Β±5.72 / 299.40 ms β 278.00 / 290.67 Β±9.51 / 303.73 ms β no change β
β QQuery 13 β 377.60 / 393.26 Β±14.58 / 414.30 ms β 402.81 / 417.30 Β±12.76 / 436.30 ms β 1.06x slower β
β QQuery 14 β 284.59 / 301.36 Β±18.05 / 336.18 ms β 304.99 / 313.42 Β±7.39 / 323.56 ms β no change β
β QQuery 15 β 307.23 / 320.58 Β±8.22 / 329.59 ms β 299.06 / 319.23 Β±18.71 / 353.83 ms β no change β
β QQuery 16 β 680.67 / 696.78 Β±10.35 / 712.58 ms β 630.35 / 641.75 Β±7.16 / 650.54 ms β +1.09x faster β
β QQuery 17 β 683.80 / 692.27 Β±8.64 / 706.92 ms β 664.47 / 675.57 Β±7.30 / 683.36 ms β no change β
β QQuery 18 β 1279.87 / 1329.63 Β±38.21 / 1385.83 ms β 1407.89 / 1442.04 Β±27.63 / 1473.07 ms β 1.08x slower β
β QQuery 19 β 27.50 / 28.15 Β±0.57 / 28.83 ms β 28.73 / 32.19 Β±5.56 / 43.17 ms β 1.14x slower β
β QQuery 20 β 526.27 / 542.14 Β±14.93 / 564.25 ms β 528.20 / 536.29 Β±6.39 / 545.48 ms β no change β
β QQuery 21 β 508.37 / 514.50 Β±5.17 / 522.89 ms β 518.17 / 526.77 Β±5.24 / 531.79 ms β no change β
β QQuery 22 β 1025.42 / 1035.40 Β±7.80 / 1044.01 ms β 1001.87 / 1022.71 Β±14.58 / 1037.36 ms β no change β
β QQuery 23 β 3394.65 / 3681.53 Β±288.32 / 4196.36 ms β 3238.15 / 3292.79 Β±33.36 / 3333.65 ms β +1.12x faster β
β QQuery 24 β 42.86 / 56.96 Β±24.53 / 105.94 ms β 41.24 / 43.99 Β±2.48 / 48.48 ms β +1.29x faster β
β QQuery 25 β 107.65 / 109.41 Β±0.93 / 110.21 ms β 107.06 / 108.55 Β±1.29 / 110.80 ms β no change β
β QQuery 26 β 41.82 / 66.99 Β±47.52 / 161.96 ms β 41.19 / 41.57 Β±0.26 / 41.95 ms β +1.61x faster β
β QQuery 27 β 522.41 / 534.83 Β±8.05 / 543.27 ms β 531.17 / 538.99 Β±6.35 / 545.56 ms β no change β
β QQuery 28 β 2907.86 / 2946.25 Β±38.08 / 3008.58 ms β 2938.60 / 2972.98 Β±23.11 / 3008.74 ms β no change β
β QQuery 29 β 39.80 / 40.32 Β±0.45 / 41.08 ms β 40.23 / 40.55 Β±0.22 / 40.75 ms β no change β
β QQuery 30 β 325.22 / 335.23 Β±10.36 / 352.33 ms β 335.32 / 349.67 Β±10.38 / 365.63 ms β no change β
β QQuery 31 β 287.74 / 305.02 Β±13.82 / 324.61 ms β 303.62 / 318.69 Β±9.68 / 327.56 ms β no change β
β QQuery 32 β 1332.01 / 1458.11 Β±133.79 / 1707.55 ms β 962.56 / 1098.85 Β±104.16 / 1222.83 ms β +1.33x faster β
β QQuery 33 β 1798.06 / 1831.83 Β±37.81 / 1891.34 ms β 1682.05 / 1700.29 Β±14.24 / 1719.41 ms β +1.08x faster β
β QQuery 34 β 1778.79 / 1833.35 Β±33.01 / 1879.46 ms β 1669.51 / 1735.93 Β±119.92 / 1975.08 ms β +1.06x faster β
β QQuery 35 β 357.90 / 446.13 Β±129.08 / 701.89 ms β 331.03 / 407.42 Β±129.57 / 666.15 ms β +1.10x faster β
β QQuery 36 β 71.73 / 76.28 Β±3.44 / 81.01 ms β 70.49 / 82.93 Β±13.37 / 108.42 ms β 1.09x slower β
β QQuery 37 β 37.83 / 53.88 Β±24.43 / 101.88 ms β 37.46 / 38.51 Β±0.69 / 39.52 ms β +1.40x faster β
β QQuery 38 β 42.56 / 49.92 Β±7.59 / 64.32 ms β 41.11 / 45.33 Β±2.31 / 47.60 ms β +1.10x faster β
β QQuery 39 β 157.79 / 167.53 Β±7.02 / 178.37 ms β 153.37 / 163.49 Β±6.69 / 172.24 ms β no change β
β QQuery 40 β 15.44 / 19.01 Β±5.66 / 30.21 ms β 14.91 / 16.85 Β±2.47 / 21.54 ms β +1.13x faster β
β QQuery 41 β 14.59 / 16.65 Β±2.52 / 21.58 ms β 14.16 / 14.31 Β±0.10 / 14.47 ms β +1.16x faster β
β QQuery 42 β 14.23 / 14.52 Β±0.21 / 14.85 ms β 13.74 / 13.90 Β±0.12 / 14.04 ms β no change β
βββββββββββββ΄βββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 21885.18ms β
β Total Time (fix_filter-sparse-view-gc) β 20973.08ms β
β Average Time (HEAD) β 508.96ms β
β Average Time (fix_filter-sparse-view-gc) β 487.75ms β
β Queries Faster β 13 β
β Queries Slower β 7 β
β Queries with No Change β 23 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 110.0s |
| Peak memory | 17.9 GiB |
| Avg memory | 6.0 GiB |
| CPU user | 1063.3s |
| CPU sys | 114.6s |
| Peak spill | 0 B |
**clickbench_partitioned β branch**
| Metric | Value |
|--------|-------|
| Wall time | 110.0s |
| Peak memory | 12.3 GiB |
| Avg memory | 4.6 GiB |
| CPU user | 1061.2s |
| CPU sys | 83.4s |
| 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] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6240509415
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/26436#issuecomment-6240443625)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c6240443625-3552-5bpxs 6.12.94+ #1 SMP Fri Aug 21 08:00:16 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 fix/filter-sparse-view-gc (7c633954f6e5f3792b9ca2eff964a97609cda729) to 5f21582 (merge-base) [diff](https://github.com/apache/datafusion/compare/5f2158269e24ac45e2740dcfe6e79a6f69f28330..7c633954f6e5f3792b9ca2eff964a97609cda729)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpch
```
</details>
Results will be posted here when complete
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6240509136
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/26436#issuecomment-6240443625)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c6240443625-3551-nzmvb 6.12.94+ #1 SMP Fri Aug 21 08:00:16 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 fix/filter-sparse-view-gc (7c633954f6e5f3792b9ca2eff964a97609cda729) to 5f21582 (merge-base) [diff](https://github.com/apache/datafusion/compare/5f2158269e24ac45e2740dcfe6e79a6f69f28330..7c633954f6e5f3792b9ca2eff964a97609cda729)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpcds
```
</details>
Results will be posted here when complete
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6240577561
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/26436#issuecomment-6240443625)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
Comparing fix/filter-sparse-view-gc (7c633954f6e5f3792b9ca2eff964a97609cda729) to 5f21582 (merge-base) [diff](https://github.com/apache/datafusion/compare/5f2158269e24ac45e2740dcfe6e79a6f69f28330..7c633954f6e5f3792b9ca2eff964a97609cda729)
<details><summary>Run configuration</summary>
```yaml
run benchmark tpch
```
</details>
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 16
On-line CPU(s) list: 0-15
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 16
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 1 MiB (16 instances)
L1i cache: 1 MiB (16 instances)
L2 cache: 32 MiB (16 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-15
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and fix_filter-sparse-view-gc
--------------------
Benchmark tpch_sf1.json
--------------------
βββββββββββββ³βββββββββββ³ββββββββββββββββββββββββββββ³ββββββββββββ
β Query β HEAD β fix_filter-sparse-view-gc β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 40.30 ms β 40.60 ms β no change β
β QQuery 2 β 18.05 ms β 18.06 ms β no change β
β QQuery 3 β 28.29 ms β 27.81 ms β no change β
β QQuery 4 β 17.16 ms β 17.09 ms β no change β
β QQuery 5 β 34.74 ms β 35.03 ms β no change β
β QQuery 6 β 13.76 ms β 13.58 ms β no change β
β QQuery 7 β 41.70 ms β 40.91 ms β no change β
β QQuery 8 β 41.07 ms β 40.69 ms β no change β
β QQuery 9 β 50.21 ms β 50.00 ms β no change β
β QQuery 10 β 40.33 ms β 40.07 ms β no change β
β QQuery 11 β 13.03 ms β 12.93 ms β no change β
β QQuery 12 β 17.85 ms β 17.83 ms β no change β
β QQuery 13 β 39.21 ms β 38.91 ms β no change β
β QQuery 14 β 24.40 ms β 24.36 ms β no change β
β QQuery 15 β 30.77 ms β 30.60 ms β no change β
β QQuery 16 β 13.56 ms β 13.46 ms β no change β
β QQuery 17 β 72.70 ms β 72.58 ms β no change β
β QQuery 18 β 44.64 ms β 45.30 ms β no change β
β QQuery 19 β 31.85 ms β 31.52 ms β no change β
β QQuery 20 β 31.31 ms β 31.49 ms β no change β
β QQuery 21 β 54.58 ms β 54.17 ms β no change β
β QQuery 22 β 13.54 ms β 13.30 ms β no change β
βββββββββββββ΄βββββββββββ΄ββββββββββββββββββββββββββββ΄ββββββββββββ
ββββββββββββββββββββββββββββββββββββββββββββ³βββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 713.05ms β
β Total Time (fix_filter-sparse-view-gc) β 710.28ms β
β Average Time (HEAD) β 32.41ms β
β Average Time (fix_filter-sparse-view-gc) β 32.29ms β
β Queries Faster β 0 β
β Queries Slower β 0 β
β Queries with No Change β 22 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββββββ΄βββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and fix_filter-sparse-view-gc
--------------------
Benchmark tpch_sf1.json
--------------------
βββββββββββββ³βββββββββββββββββββββββββββββββββ³βββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β fix_filter-sparse-view-gc β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 1 β 40.30 / 41.27 Β±0.96 / 42.74 ms β 40.60 / 42.38 Β±1.29 / 44.50 ms β no change β
β QQuery 2 β 18.05 / 18.46 Β±0.38 / 19.08 ms β 18.06 / 18.35 Β±0.26 / 18.83 ms β no change β
β QQuery 3 β 28.29 / 28.56 Β±0.17 / 28.79 ms β 27.81 / 28.31 Β±0.26 / 28.55 ms β no change β
β QQuery 4 β 17.16 / 17.27 Β±0.11 / 17.47 ms β 17.09 / 17.53 Β±0.53 / 18.57 ms β no change β
β QQuery 5 β 34.74 / 35.22 Β±0.56 / 36.30 ms β 35.03 / 35.26 Β±0.26 / 35.73 ms β no change β
β QQuery 6 β 13.76 / 14.64 Β±0.74 / 15.57 ms β 13.58 / 13.87 Β±0.49 / 14.85 ms β +1.06x faster β
β QQuery 7 β 41.70 / 42.82 Β±1.00 / 44.61 ms β 40.91 / 42.05 Β±0.71 / 42.85 ms β no change β
β QQuery 8 β 41.07 / 42.48 Β±2.01 / 46.44 ms β 40.69 / 41.48 Β±1.11 / 43.67 ms β no change β
β QQuery 9 β 50.21 / 51.29 Β±0.75 / 52.42 ms β 50.00 / 50.56 Β±0.49 / 51.38 ms β no change β
β QQuery 10 β 40.33 / 40.95 Β±0.97 / 42.88 ms β 40.07 / 40.78 Β±1.26 / 43.29 ms β no change β
β QQuery 11 β 13.03 / 13.34 Β±0.24 / 13.76 ms β 12.93 / 13.19 Β±0.22 / 13.57 ms β no change β
β QQuery 12 β 17.85 / 18.33 Β±0.32 / 18.72 ms β 17.83 / 18.24 Β±0.35 / 18.67 ms β no change β
β QQuery 13 β 39.21 / 40.48 Β±1.31 / 42.53 ms β 38.91 / 40.02 Β±1.30 / 42.48 ms β no change β
β QQuery 14 β 24.40 / 24.46 Β±0.09 / 24.64 ms β 24.36 / 24.62 Β±0.25 / 24.97 ms β no change β
β QQuery 15 β 30.77 / 31.18 Β±0.35 / 31.67 ms β 30.60 / 30.79 Β±0.18 / 31.11 ms β no change β
β QQuery 16 β 13.56 / 13.68 Β±0.09 / 13.81 ms β 13.46 / 13.65 Β±0.19 / 13.98 ms β no change β
β QQuery 17 β 72.70 / 74.70 Β±1.87 / 77.65 ms β 72.58 / 73.69 Β±1.17 / 75.85 ms β no change β
β QQuery 18 β 44.64 / 45.87 Β±1.10 / 47.86 ms β 45.30 / 45.86 Β±0.57 / 46.87 ms β no change β
β QQuery 19 β 31.85 / 32.11 Β±0.24 / 32.54 ms β 31.52 / 32.46 Β±1.29 / 34.94 ms β no change β
β QQuery 20 β 31.31 / 31.65 Β±0.28 / 32.03 ms β 31.49 / 31.96 Β±0.42 / 32.66 ms β no change β
β QQuery 21 β 54.58 / 55.30 Β±0.60 / 56.16 ms β 54.17 / 55.52 Β±0.99 / 57.12 ms β no change β
β QQuery 22 β 13.54 / 13.64 Β±0.12 / 13.87 ms β 13.30 / 13.62 Β±0.22 / 13.88 ms β no change β
βββββββββββββ΄βββββββββββββββββββββββββββββββββ΄βββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββββββ³βββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 727.71ms β
β Total Time (fix_filter-sparse-view-gc) β 724.19ms β
β Average Time (HEAD) β 33.08ms β
β Average Time (fix_filter-sparse-view-gc) β 32.92ms β
β Queries Faster β 1 β
β Queries Slower β 0 β
β Queries with No Change β 21 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββββββ΄βββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpch β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 5.0s |
| Peak memory | 1.1 GiB |
| Avg memory | 490.1 MiB |
| CPU user | 19.9s |
| CPU sys | 1.6s |
| Peak spill | 0 B |
**tpch β branch**
| Metric | Value |
|--------|-------|
| Wall time | 5.0s |
| Peak memory | 1.1 GiB |
| Avg memory | 469.5 MiB |
| CPU user | 19.7s |
| CPU sys | 1.7s |
| 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] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "jayzhan211 (via GitHub)" <gi...@apache.org>.
jayzhan211 merged PR #26436:
URL: https://github.com/apache/datafusion/pull/26436
--
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] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "jayzhan211 (via GitHub)" <gi...@apache.org>.
jayzhan211 commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6290299926
Thanks @zhuqi-lucas @Rich-T-kid !
--
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] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "Rich-T-kid (via GitHub)" <gi...@apache.org>.
Rich-T-kid commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6240443625
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] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6240498621
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/26436#issuecomment-6240443625)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c6240443625-3550-zb4fp 6.12.94+ #1 SMP Fri Aug 21 08:00:16 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 fix/filter-sparse-view-gc (7c633954f6e5f3792b9ca2eff964a97609cda729) to 5f21582 (merge-base) [diff](https://github.com/apache/datafusion/compare/5f2158269e24ac45e2740dcfe6e79a6f69f28330..7c633954f6e5f3792b9ca2eff964a97609cda729)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
```
</details>
Results will be posted here when complete
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] fix: compact sparse string views in FilterExec coalescing [datafusion]
Posted by "codecov-commenter (via GitHub)" <gi...@apache.org>.
codecov-commenter commented on PR #26436:
URL: https://github.com/apache/datafusion/pull/26436#issuecomment-6227257177
## [Codecov](https://app.codecov.io/gh/apache/datafusion/pull/26436?dropdown=coverage&src=pr&el=h1&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache) Report
:x: Patch coverage is `96.96970%` with `1 line` in your changes missing coverage. Please review.
:white_check_mark: Project coverage is 82.89%. Comparing base ([`5f21582`](https://app.codecov.io/gh/apache/datafusion/commit/5f2158269e24ac45e2740dcfe6e79a6f69f28330?dropdown=coverage&el=desc&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache)) to head ([`7c63395`](https://app.codecov.io/gh/apache/datafusion/commit/7c633954f6e5f3792b9ca2eff964a97609cda729?dropdown=coverage&el=desc&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache)).
| [Files with missing lines](https://app.codecov.io/gh/apache/datafusion/pull/26436?dropdown=coverage&src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache) | Patch % | Lines |
|---|---|---|
| [datafusion/physical-plan/src/coalesce/mod.rs](https://app.codecov.io/gh/apache/datafusion/pull/26436?src=pr&el=tree&filepath=datafusion%2Fphysical-plan%2Fsrc%2Fcoalesce%2Fmod.rs&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache#diff-ZGF0YWZ1c2lvbi9waHlzaWNhbC1wbGFuL3NyYy9jb2FsZXNjZS9tb2QucnM=) | 96.96% | [0 Missing and 1 partial :warning: ](https://app.codecov.io/gh/apache/datafusion/pull/26436?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache) |
<details><summary>Additional details and impacted files</summary>
```diff
@@ Coverage Diff @@
## main #26436 +/- ##
=======================================
Coverage 82.89% 82.89%
=======================================
Files 1154 1154
Lines 457620 457653 +33
Branches 457620 457653 +33
=======================================
+ Hits 379352 379383 +31
Misses 55262 55262
- Partials 23006 23008 +2
```
</details>
[:umbrella: View full report in Codecov by Harness](https://app.codecov.io/gh/apache/datafusion/pull/26436?dropdown=coverage&src=pr&el=continue&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache).
:loudspeaker: Have feedback on the report? [Share it here](https://about.codecov.io/codecov-pr-comment-feedback/?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache).
<details><summary> :rocket: New features to boost your workflow: </summary>
- :snowflake: [Test Analytics](https://docs.codecov.com/docs/test-analytics): Detect flaky tests, report on failures, and find test suite problems.
- :package: [JS Bundle Analysis](https://docs.codecov.com/docs/javascript-bundle-analysis): Save yourself from yourself by tracking and limiting bundle sizes in JS merges.
</details>
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org