Skip to content

DRAFT Bench 25811 - #26585

Draft
kosiew wants to merge 2 commits into
apache:mainfrom
kosiew:bench-25811
Draft

DRAFT Bench 25811#26585
kosiew wants to merge 2 commits into
apache:mainfrom
kosiew:bench-25811

Conversation

@kosiew

@kosiew kosiew commented Oct 22, 2026

Copy link
Copy Markdown
Contributor

benchmark for #25811

kosiew and others added 2 commits October 22, 2026 12:29
@kosiew
Add parquet bloom filter pruning benchmarks for multiple queries
930e6a0 
- Add parquet bloom filter pruning benchmarks for three different query types.
- Implement benchmarks for reject_first, reject_last, and survive_all predicates.
- Test pruning performance across memory and slow storage configurations.
@mohitgurav20 @claude @kosiew
perf: evaluate Parquet Bloom filters incrementally during row group p…
dfde3fe 
…runing

Squashed net diff of apache#25811 at 31a19dd
(6c1727f..31a19dd), rebased onto main.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@github-actions github-actions Bot added core Core DataFusion crate datasource Changes to the datasource crate labels Oct 22, 2026
@kosiew

kosiew commented Oct 22, 2026

Copy link
Copy Markdown
Contributor Author

run benchmark parquet_bloom_filter_pruning

baseline:
  ref: 930e6a03c7bff9cba25eccf12e625aa04c63216c

@adriangbot

adriangbot commented Oct 22, 2026

Copy link
Copy Markdown

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c6332775007-3743-pwplq 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu) 
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

Comparing bench-25811 (dfde3fe) to 930e6a0 diff

Run configuration 
run benchmark parquet_bloom_filter_pruning
baseline:
  ref: "930e6a03c7bff9cba25eccf12e625aa04c63216c"

Results will be posted here when complete

File an issue against this benchmark runner

@adriangbot

adriangbot commented Oct 22, 2026

Copy link
Copy Markdown

🤖 Benchmark completed (GKE) | trigger

Instance: c4a-highmem-16 (12 vCPU / 65 GiB)

Comparing bench-25811 (dfde3fe) to 930e6a0 diff

Run configuration 
run benchmark parquet_bloom_filter_pruning
baseline:
  ref: "930e6a03c7bff9cba25eccf12e625aa04c63216c"
CPU Details (lscpu) 
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 

group                                                          HEAD                                   bench-25811
-----                                                          ----                                   -----------
parquet_bloom_filter_pruning/memory/reject_first/columns/2     1.10    740.5±1.58µs        ? ?/sec    1.00    676.2±1.47µs        ? ?/sec
parquet_bloom_filter_pruning/memory/reject_first/columns/32    1.49      6.1±0.05ms        ? ?/sec    1.00      4.1±0.05ms        ? ?/sec
parquet_bloom_filter_pruning/memory/reject_first/columns/8     1.30  1595.5±14.12µs        ? ?/sec    1.00   1226.9±4.34µs        ? ?/sec
parquet_bloom_filter_pruning/memory/reject_last/columns/2      1.00    788.6±1.57µs        ? ?/sec    1.89  1489.5±10.88µs        ? ?/sec
parquet_bloom_filter_pruning/memory/reject_last/columns/32     1.00      7.2±0.25ms        ? ?/sec    43.31   312.4±1.08ms        ? ?/sec
parquet_bloom_filter_pruning/memory/reject_last/columns/8      1.00  1880.5±16.14µs        ? ?/sec    10.87    20.4±0.05ms        ? ?/sec
parquet_bloom_filter_pruning/memory/survive_all/columns/2      1.00      2.3±0.03ms        ? ?/sec    1.26      2.9±0.02ms        ? ?/sec
parquet_bloom_filter_pruning/memory/survive_all/columns/32     1.00     27.6±0.10ms        ? ?/sec    12.08   332.9±0.70ms        ? ?/sec
parquet_bloom_filter_pruning/memory/survive_all/columns/8      1.00      7.5±0.03ms        ? ?/sec    3.39     25.5±0.13ms        ? ?/sec
parquet_bloom_filter_pruning/slow/reject_first/columns/2       1.99    162.2±0.49ms        ? ?/sec    1.00     81.4±0.30ms        ? ?/sec
parquet_bloom_filter_pruning/slow/reject_first/columns/32      30.22      2.6±0.00s        ? ?/sec    1.00     85.5±0.67ms        ? ?/sec
parquet_bloom_filter_pruning/slow/reject_first/columns/8       7.89    646.8±2.91ms        ? ?/sec    1.00     82.0±0.94ms        ? ?/sec
parquet_bloom_filter_pruning/slow/reject_last/columns/2        1.00    162.4±0.82ms        ? ?/sec    1.00    162.9±1.78ms        ? ?/sec
parquet_bloom_filter_pruning/slow/reject_last/columns/32       1.00       2.6±0.01s        ? ?/sec    1.12       2.9±0.01s        ? ?/sec
parquet_bloom_filter_pruning/slow/reject_last/columns/8        1.00    646.6±2.00ms        ? ?/sec    1.04    671.4±3.97ms        ? ?/sec
parquet_bloom_filter_pruning/slow/survive_all/columns/2        1.00    244.9±0.83ms        ? ?/sec    1.00    244.1±1.29ms        ? ?/sec
parquet_bloom_filter_pruning/slow/survive_all/columns/32       1.00       2.7±0.00s        ? ?/sec    1.12       3.0±0.01s        ? ?/sec
parquet_bloom_filter_pruning/slow/survive_all/columns/8        1.00    732.1±2.26ms        ? ?/sec    1.03    753.1±2.78ms        ? ?/sec

Resource Usage

parquet_bloom_filter_pruning — base (merge-base)

Metric Value
Wall time 785.2s
Peak memory 122.1 MiB
Avg memory 37.9 MiB
CPU user 116.5s
CPU sys 1.3s
Peak spill 0 B

parquet_bloom_filter_pruning — branch

Metric Value
Wall time 755.2s
Peak memory 110.4 MiB
Avg memory 35.0 MiB
CPU user 139.1s
CPU sys 1.2s
Peak spill 0 B

File an issue against this benchmark runner

@kosiew kosiew mentioned this pull request Oct 22, 2026
Open
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

core Core DataFusion crate datasource Changes to the datasource crate

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@kosiew @adriangbot @mohitgurav20