subject:"Re\: \[PR\] \[Experiment\] Adaptive filter pushdown \[datafusion\]"

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-18 Thread via GitHub



alamb commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4481183612

   I agree being able to turn off runtime adaptivity seems like an important 
knob for people who know something about their data or want their execution to 
be more predictable


-- 
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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangb commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4466054708

   Yes that makes sense. Setting `filter_confidence_z` to something very high 
would keep adaptation from happening (I think).


-- 
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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



asolimando commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4465650148

   @adriangb do you foresee a moment where this is not gated by a config knob?
   
   I have in mind what happens between partial and final aggregation phases, 
where partial bails out after 100k rows if NDV is too high (groups number is 
proportional to the size of the input).
   
   Runtime adaptivity is a real need and often a good idea, but in some 
pathological cases there can be false positives: by chance all your distinct 
values are at the beginning, or your selectivity for the first x row 
groups/files is low, but that finally doesn't reflect the real distribution, 
and you'd rather trust global statistics from files/catalog otherwise, possibly 
confirming with a feedback loop from runtime stats.
   
   It would be great to gate these runtime adaptivity techniques or at least 
expose the thresholds programmatically to adjust their sensitivity based on 
stats for your workload.
   
   WDYT?


-- 
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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangb commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4464041143

   ## Split into a reviewable stack
   
   This experiment has been broken into 4 stacked PRs so each piece can be 
reviewed (and where possible, merged) on its own. Each builds on the previous; 
the diffs are cumulative against `main`, but every PR adds exactly **one new 
commit** — review that commit.
   
   1. **#22234 — `OptionalFilterPhysicalExpr` + proto** (+400)
  A transparent `PhysicalExpr` wrapper marking a filter as 
droppable-without-affecting-correctness, plus proto round-trip support. Purely 
additive, no caller — inert until something reads the marker.
   
   2. **#22235 — Per-conjunct pruning statistics** (+500/-18)
  `PruningPredicate::try_new_tagged_conjuncts` / `prune_per_conjunct` and 
the row-group / page-index variants surface per-conjunct effectiveness as a 
free side effect of the pruning pass. Existing untagged paths unchanged.
   
   3. **#22236 — `SelectivityTracker` cost model** (+2973)
  The cross-file cost model that partitions filter conjuncts into row-level 
/ post-scan / dropped buckets, with ~45 unit tests and a benchmark. Not yet 
wired into the scan.
   
   4. **#22237 — Adaptive parquet scan integration** (+1823/-624)
  Wires it all together: `AdaptiveParquetStream`, re-partitioning at 
row-group boundaries, integration with the fully-matched run splitting from 
#21637, the hash-join `OptionalFilterPhysicalExpr` wrap, and config knobs.
   
   ### Notes
   
   - Each layer compiles and passes clippy (`-D warnings`) independently.
   - **PRs 1–3 have no external dependency** and can merge on their own merits.
   - **PR 4** pins a custom `arrow-rs` branch for the push-decoder 
`StrategySwap` APIs — it cannot merge upstream until those APIs land in a 
released `arrow-rs`.
   
   This PR remains as the integration reference / discussion thread.


-- 
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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463469911

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463167109)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark clickbench_partitioned.json
   
   
┏━━━┳━━━┳━━━┳┓
   ┃ Query ┃  HEAD ┃ 
exp_r6-pruningpredicate-rates ┃ Change ┃
   
┡━━━╇━━━╇━━━╇┩
   │ QQuery 0  │  1.18 / 4.55 ±6.65 / 17.85 ms │  1.23 / 4.64 
±6.77 / 18.17 ms │  no change │
   │ QQuery 1  │12.53 / 12.72 ±0.13 / 12.88 ms │13.95 / 14.15 
±0.15 / 14.30 ms │   1.11x slower │
   │ QQuery 2  │35.55 / 35.73 ±0.19 / 36.00 ms │35.95 / 36.40 
±0.26 / 36.74 ms │  no change │
   │ QQuery 3  │30.63 / 31.18 ±0.49 / 32.08 ms │30.74 / 32.07 
±2.22 / 36.49 ms │  no change │
   │ QQuery 4  │ 229.35 / 232.73 ±3.16 / 237.83 ms │ 226.18 / 230.52 
±3.52 / 235.78 ms │  no change │
   │ QQuery 5  │ 276.08 / 278.25 ±1.63 / 280.41 ms │ 273.86 / 276.18 
±2.00 / 279.55 ms │  no change │
   │ QQuery 6  │   6.43 / 7.23 ±0.64 / 8.26 ms │   5.04 / 5.43 
±0.30 / 5.94 ms │  +1.33x faster │
   │ QQuery 7  │13.82 / 13.87 ±0.06 / 13.98 ms │14.88 / 15.10 
±0.13 / 15.24 ms │   1.09x slower │
   │ QQuery 8  │ 321.74 / 325.66 ±3.17 / 329.95 ms │ 322.20 / 327.14 
±2.51 / 329.03 ms │  no change │
   │ QQuery 9  │ 438.05 / 445.22 ±4.52 / 451.29 ms │ 444.27 / 453.00 
±6.64 / 463.39 ms │  no change │
   │ QQuery 10 │70.56 / 71.67 ±0.91 / 72.69 ms │69.99 / 71.01 
±0.82 / 72.28 ms │  no change │
   │ QQuery 11 │81.40 / 81.89 ±0.28 / 82.25 ms │81.30 / 84.87 
±4.81 / 94.35 ms │  no change │
   │ QQuery 12 │ 267.52 / 272.78 ±3.60 / 277.34 ms │ 258.50 / 263.44 
±3.28 / 267.84 ms │  no change │
   │ QQuery 13 │ 381.73 / 393.06 ±8.76 / 402.14 ms │ 401.17 / 407.61 
±4.68 / 414.83 ms │  no change │
   │ QQuery 14 │ 281.88 / 284.29 ±1.70 / 286.36 ms │ 274.30 / 278.15 
±3.35 / 282.64 ms │

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463436093

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463167109)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark tpcds_sf1.json
   
   
┏━━━┳━━━┳┳━━━┓
   ┃ Query ┃  HEAD ┃  
exp_r6-pruningpredicate-rates ┃Change ┃
   
┡━━━╇━━━╇╇━━━┩
   │ QQuery 1  │   6.31 / 6.87 ±0.83 / 8.52 ms │5.99 / 6.52 
±0.79 / 8.09 ms │ +1.05x faster │
   │ QQuery 2  │82.22 / 83.32 ±0.65 / 84.01 ms │ 48.87 / 49.59 
±0.46 / 50.27 ms │ +1.68x faster │
   │ QQuery 3  │29.82 / 30.16 ±0.21 / 30.37 ms │ 30.35 / 30.78 
±0.34 / 31.36 ms │ no change │
   │ QQuery 4  │ 543.15 / 550.06 ±3.83 / 554.81 ms │  505.82 / 507.37 ±1.88 
/ 510.87 ms │ +1.08x faster │
   │ QQuery 5  │53.77 / 54.12 ±0.36 / 54.82 ms │ 54.52 / 55.26 
±0.56 / 55.94 ms │ no change │
   │ QQuery 6  │36.64 / 36.98 ±0.28 / 37.36 ms │ 36.20 / 36.40 
±0.17 / 36.63 ms │ no change │
   │ QQuery 7  │ 113.22 / 115.19 ±1.78 / 118.10 ms │  119.12 / 120.28 ±1.11 
/ 122.15 ms │ no change │
   │ QQuery 8  │39.93 / 40.10 ±0.24 / 40.58 ms │ 18.88 / 18.97 
±0.07 / 19.05 ms │ +2.11x faster │
   │ QQuery 9  │54.41 / 56.48 ±1.61 / 59.35 ms │ 52.52 / 55.03 
±1.83 / 58.17 ms │ no change │
   │ QQuery 10 │83.91 / 84.85 ±1.06 / 86.87 ms │ 87.19 / 87.37 
±0.23 / 87.81 ms │ no change │
   │ QQuery 11 │ 343.01 / 347.89 ±3.82 / 354.56 ms │  327.47 / 330.87 ±2.61 
/ 335.55 ms │ no change │
   │ QQuery 12 │30.04 / 30.23 ±0.17 / 30.52 ms │ 31.96 / 32.28 
±0.25 / 32.58 ms │  1.07x slower │
   │ QQuery 13 │ 131.72 / 132.36 ±0.49 / 132.94 ms │  158.62 / 159.53 ±1.27 
/ 162.05 ms │  1.21x slower │
   │ QQuery 14 │ 526.67 / 528.54 ±1.42 / 530.47 ms │  527.74 / 530.20 ±1.60 
/ 532.34 ms │ no change │
   │ QQuery 15 │63.04 / 63.54 ±0.58 / 64.65 ms │ 26.76 / 27.02 
±0.17 / 27.22 ms │ +2.35x faster │
   │ QQuery 16 │   7.16 / 7.45 ±0.35 / 8.10 ms │6.61 /

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463419392

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463165828)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark tpch_sf1.json
   
   
┏━━━┳━━━┳┳━━━┓
   ┃ Query ┃  HEAD ┃  
exp_r6-pruningpredicate-rates ┃Change ┃
   
┡━━━╇━━━╇╇━━━┩
   │ QQuery 1  │39.02 / 40.17 ±0.90 / 41.38 ms │ 37.78 / 39.10 ±1.63 / 
41.63 ms │ no change │
   │ QQuery 2  │24.32 / 24.63 ±0.37 / 25.36 ms │ 23.38 / 23.79 ±0.64 / 
25.04 ms │ no change │
   │ QQuery 3  │46.66 / 47.03 ±0.23 / 47.25 ms │ 34.67 / 35.65 ±0.92 / 
37.28 ms │ +1.32x faster │
   │ QQuery 4  │21.68 / 21.80 ±0.09 / 21.95 ms │ 17.74 / 18.33 ±0.66 / 
19.53 ms │ +1.19x faster │
   │ QQuery 5  │66.51 / 67.60 ±1.13 / 69.40 ms │ 45.02 / 46.23 ±0.85 / 
47.64 ms │ +1.46x faster │
   │ QQuery 6  │34.37 / 35.10 ±0.80 / 36.55 ms │ 13.86 / 14.49 ±0.69 / 
15.55 ms │ +2.42x faster │
   │ QQuery 7  │56.78 / 56.93 ±0.11 / 57.04 ms │ 39.79 / 40.81 ±1.38 / 
43.54 ms │ +1.39x faster │
   │ QQuery 8  │77.66 / 78.02 ±0.41 / 78.83 ms │ 50.14 / 50.39 ±0.20 / 
50.67 ms │ +1.55x faster │
   │ QQuery 9  │ 103.44 / 104.01 ±0.31 / 104.35 ms │ 63.04 / 63.71 ±0.36 / 
64.04 ms │ +1.63x faster │
   │ QQuery 10 │73.88 / 74.29 ±0.35 / 74.92 ms │ 63.45 / 64.33 ±1.28 / 
66.82 ms │ +1.15x faster │
   │ QQuery 11 │15.16 / 15.30 ±0.12 / 15.48 ms │ 13.60 / 13.99 ±0.41 / 
14.74 ms │ +1.09x faster │
   │ QQuery 12 │44.32 / 44.82 ±0.35 / 45.25 ms │ 25.24 / 26.07 ±0.93 / 
27.76 ms │ +1.72x faster │
   │ QQuery 13 │46.09 / 46.76 ±0.36 / 47.09 ms │ 35.25 / 36.58 ±1.36 / 
38.99 ms │ +1.28x faster │
   │ QQuery 14 │38.86 / 39.12 ±0.18 / 39.40 ms │ 29.21 / 30.37 ±1.18 / 
32.20 ms │ +1.29x faster │
   │ QQuery 15 │40.83 / 41.36 ±0.69 / 42.70 ms │ 30.75 / 30.92 ±0.18 / 
31.24 ms │ +1.34x faster │
   │ QQuery 16 │22.80 / 22.87 ±0.08 / 23.01 ms │ 18.32 / 18.54 ±0.12 / 
18.68 ms │ +1.23x faster │
   │ QQuery 17 │ 149.76 / 153.08 ±2.03 / 155.51 ms │ 77.97 / 78.47 ±0.51 / 
79.40 ms │ +1.95x faster │
   │ QQuer

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463417691

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463167109)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark tpch_sf1.json
   
   
┏━━━┳┳┳━━━┓
   ┃ Query ┃   HEAD ┃  
exp_r6-pruningpredicate-rates ┃Change ┃
   
┡━━━╇╇╇━━━┩
   │ QQuery 1  │ 38.44 / 39.21 ±0.72 / 40.56 ms │ 37.29 / 37.98 ±1.02 / 40.00 
ms │ no change │
   │ QQuery 2  │ 20.36 / 20.61 ±0.20 / 20.94 ms │ 23.44 / 23.94 ±0.61 / 25.13 
ms │  1.16x slower │
   │ QQuery 3  │ 34.25 / 35.62 ±1.17 / 37.10 ms │ 34.22 / 35.46 ±1.22 / 37.06 
ms │ no change │
   │ QQuery 4  │ 16.80 / 17.07 ±0.18 / 17.37 ms │ 17.53 / 17.89 ±0.61 / 19.11 
ms │ no change │
   │ QQuery 5  │ 44.09 / 44.72 ±0.39 / 45.18 ms │ 43.57 / 45.50 ±1.11 / 46.99 
ms │ no change │
   │ QQuery 6  │ 16.08 / 16.45 ±0.32 / 16.96 ms │ 13.42 / 13.93 ±0.81 / 15.54 
ms │ +1.18x faster │
   │ QQuery 7  │ 48.12 / 50.84 ±1.83 / 52.96 ms │ 39.30 / 40.28 ±1.10 / 42.13 
ms │ +1.26x faster │
   │ QQuery 8  │ 45.00 / 45.25 ±0.13 / 45.33 ms │ 49.36 / 49.76 ±0.55 / 50.83 
ms │  1.10x slower │
   │ QQuery 9  │ 50.73 / 51.33 ±0.78 / 52.83 ms │ 62.17 / 62.89 ±0.45 / 63.38 
ms │  1.23x slower │
   │ QQuery 10 │ 63.28 / 63.41 ±0.11 / 63.60 ms │ 62.88 / 63.51 ±0.88 / 65.23 
ms │ no change │
   │ QQuery 11 │ 13.49 / 14.37 ±1.29 / 16.90 ms │ 13.15 / 13.92 ±0.98 / 15.84 
ms │ no change │
   │ QQuery 12 │ 24.95 / 25.37 ±0.52 / 26.37 ms │ 24.41 / 24.89 ±0.37 / 25.42 
ms │ no change │
   │ QQuery 13 │ 35.48 / 36.47 ±0.56 / 37.03 ms │ 35.85 / 37.67 ±3.02 / 43.69 
ms │ no change │
   │ QQuery 14 │ 25.32 / 25.45 ±0.09 / 25.57 ms │ 28.41 / 30.60 ±2.21 / 34.44 
ms │  1.20x slower │
   │ QQuery 15 │ 30.68 / 31.12 ±0.66 / 32.41 ms │ 30.07 / 30.38 ±0.25 / 30.77 
ms │ no change │
   │ QQuery 16 │ 15.02 / 15.76 ±0.86 / 17.37 ms │ 18.14 / 18.26 ±0.09 / 18.35 
ms │  1.16x slower │
   │ QQuery 17 │ 73.47 / 74.82 ±1.31 / 76.49 ms │ 76.20 / 76.77 ±0.40 / 77.26 
ms │ no change │
   │ QQuery 18 │ 68.15 / 69.86 ±1.60 / 72.71 ms │ 56.16 / 56.56 ±0.46

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463380911

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463165828)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark clickbench_partitioned.json
   
   
┏━━━┳━━━┳━━━┳━━━┓
   ┃ Query ┃  HEAD ┃ 
exp_r6-pruningpredicate-rates ┃Change ┃
   
┡━━━╇━━━╇━━━╇━━━┩
   │ QQuery 0  │  1.24 / 4.76 ±6.88 / 18.53 ms │  1.23 / 4.75 
±6.89 / 18.53 ms │ no change │
   │ QQuery 1  │12.92 / 13.45 ±0.43 / 14.19 ms │13.76 / 14.21 
±0.31 / 14.58 ms │  1.06x slower │
   │ QQuery 2  │35.78 / 36.08 ±0.33 / 36.67 ms │36.05 / 36.47 
±0.39 / 37.14 ms │ no change │
   │ QQuery 3  │31.21 / 31.69 ±0.66 / 32.94 ms │30.84 / 31.12 
±0.21 / 31.39 ms │ no change │
   │ QQuery 4  │ 239.42 / 245.23 ±4.05 / 251.95 ms │ 237.35 / 243.14 
±2.91 / 244.96 ms │ no change │
   │ QQuery 5  │ 286.77 / 290.06 ±1.91 / 292.70 ms │ 285.52 / 287.78 
±2.23 / 291.67 ms │ no change │
   │ QQuery 6  │   6.05 / 6.91 ±1.38 / 9.66 ms │   5.42 / 5.66 
±0.17 / 5.94 ms │ +1.22x faster │
   │ QQuery 7  │17.01 / 17.21 ±0.13 / 17.42 ms │15.38 / 16.55 
±2.18 / 20.91 ms │ no change │
   │ QQuery 8  │ 345.95 / 348.04 ±2.48 / 351.95 ms │ 344.20 / 346.27 
±1.67 / 348.36 ms │ no change │
   │ QQuery 9  │ 464.82 / 472.21 ±3.97 / 476.05 ms │ 473.14 / 476.69 
±3.51 / 482.55 ms │ no change │
   │ QQuery 10 │   97.88 / 99.37 ±1.49 / 101.86 ms │70.64 / 72.12 
±1.39 / 74.58 ms │ +1.38x faster │
   │ QQuery 11 │ 108.23 / 109.05 ±0.90 / 110.79 ms │82.45 / 83.00 
±0.51 / 83.90 ms │ +1.31x faster │
   │ QQuery 12 │ 318.00 / 322.96 ±3.51 / 327.11 ms │ 275.07 / 281.84 
±5.07 / 288.12 ms │ +1.15x faster │
   │ QQuery 13 │447.33 / 459.79 ±11.12 / 475.00 ms │ 424.66 / 434.97 
±7.85 / 448.38 ms │ +1.06x faster │
   │ QQuery 14 │ 338.23 / 345.11 ±5.73 / 355.29 ms │ 286.42 / 289.97 
±3.11 / 294.05 ms │ +1.19x faster │
   │

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463350547

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463165828)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark tpcds_sf1.json
   
   
┏━━━┳━━┳━━━┳━━━┓
   ┃ Query ┃ HEAD ┃ 
exp_r6-pruningpredicate-rates ┃Change ┃
   
┡━━━╇━━╇━━━╇━━━┩
   │ QQuery 1  │  5.90 / 6.51 ±0.86 / 8.22 ms │   6.05 / 6.53 ±0.80 
/ 8.12 ms │ no change │
   │ QQuery 2  │   46.04 / 46.62 ±0.48 / 47.50 ms │49.51 / 50.03 ±0.45 
/ 50.75 ms │  1.07x slower │
   │ QQuery 3  │   28.05 / 28.20 ±0.15 / 28.44 ms │30.27 / 30.58 ±0.22 
/ 30.92 ms │  1.08x slower │
   │ QQuery 4  │300.65 / 302.69 ±1.65 / 305.02 ms │ 513.97 / 519.01 ±4.46 / 
526.33 ms │  1.71x slower │
   │ QQuery 5  │   78.88 / 80.72 ±2.29 / 85.20 ms │54.60 / 55.39 ±0.49 
/ 56.01 ms │ +1.46x faster │
   │ QQuery 6  │   29.95 / 30.53 ±0.39 / 31.07 ms │36.61 / 37.17 ±0.40 
/ 37.71 ms │  1.22x slower │
   │ QQuery 7  │141.56 / 143.90 ±2.80 / 149.31 ms │ 118.80 / 121.83 ±2.64 / 
125.71 ms │ +1.18x faster │
   │ QQuery 8  │   20.77 / 21.07 ±0.19 / 21.37 ms │18.77 / 18.90 ±0.11 
/ 19.08 ms │ +1.11x faster │
   │ QQuery 9  │125.99 / 136.17 ±5.18 / 140.11 ms │53.90 / 54.84 ±0.68 
/ 55.68 ms │ +2.48x faster │
   │ QQuery 10 │116.80 / 119.01 ±2.57 / 124.03 ms │87.24 / 87.49 ±0.21 
/ 87.71 ms │ +1.36x faster │
   │ QQuery 11 │203.80 / 205.77 ±1.94 / 209.23 ms │ 337.89 / 340.34 ±1.47 / 
342.45 ms │  1.65x slower │
   │ QQuery 12 │   22.87 / 22.94 ±0.09 / 23.09 ms │32.09 / 32.41 ±0.25 
/ 32.79 ms │  1.41x slower │
   │ QQuery 13 │146.95 / 148.12 ±0.90 / 149.44 ms │ 157.47 / 159.43 ±1.21 / 
161.24 ms │  1.08x slower │
   │ QQuery 14 │526.29 / 530.92 ±4.75 / 539.17 ms │ 532.32 / 533.89 ±0.93 / 
535.08 ms │ no change │
   │ QQuery 15 │   19.47 / 19.78 ±0.25 / 20.13 ms │27.15 / 27.67 ±0.31 
/ 28.02 ms │  1.40x slower │
   │ QQuery 16 │  6.61 / 6.75 ±0.14 / 7.02 ms │   6.70 / 6.83 ±0.14 
/ 7.02 ms │ no change

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463319079

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463167109)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4463167109-158-9gwt6 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (5b0c4911de7afdf0a67507847453c00298fac303) to main 
[diff](https://github.com/apache/datafusion/compare/c5a8ee12d51eaeea59e95ab4ebb32270e81c88ab..5b0c4911de7afdf0a67507847453c00298fac303)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463318200

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463167109)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4463167109-160-p9xhw 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (5b0c4911de7afdf0a67507847453c00298fac303) to main 
[diff](https://github.com/apache/datafusion/compare/c5a8ee12d51eaeea59e95ab4ebb32270e81c88ab..5b0c4911de7afdf0a67507847453c00298fac303)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463318796

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463167109)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4463167109-159-wt6s5 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (5b0c4911de7afdf0a67507847453c00298fac303) to main 
[diff](https://github.com/apache/datafusion/compare/c5a8ee12d51eaeea59e95ab4ebb32270e81c88ab..5b0c4911de7afdf0a67507847453c00298fac303)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463318530

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463165828)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4463165828-157-kl4cp 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (5b0c4911de7afdf0a67507847453c00298fac303) to main 
[diff](https://github.com/apache/datafusion/compare/c5a8ee12d51eaeea59e95ab4ebb32270e81c88ab..5b0c4911de7afdf0a67507847453c00298fac303)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463232125

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463165828)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4463165828-156-lh97d 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (5b0c4911de7afdf0a67507847453c00298fac303) to main 
[diff](https://github.com/apache/datafusion/compare/c5a8ee12d51eaeea59e95ab4ebb32270e81c88ab..5b0c4911de7afdf0a67507847453c00298fac303)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463205895

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4463165828)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4463165828-155-jl9jn 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (5b0c4911de7afdf0a67507847453c00298fac303) to main 
[diff](https://github.com/apache/datafusion/compare/c5a8ee12d51eaeea59e95ab4ebb32270e81c88ab..5b0c4911de7afdf0a67507847453c00298fac303)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangb commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463167109

   run benchmarks
   
   ```yaml
   baseline:
   ref: main
   env:
  DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: false
  DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: false
   changed:
   ref: HEAD
   env:
  DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: true
  DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: true
   ```


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-15 Thread via GitHub



adriangb commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4463165828

   run benchmarks
   
   ```yaml
   baseline:
   ref: main
   env:
  DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: true
  DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: true
   changed:
   ref: HEAD
   env:
  DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: true
  DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: true
   ```


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448140023

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448016719)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark clickbench_partitioned.json
   
   
┏━━━┳━━━┳━━━┳┓
   ┃ Query ┃  HEAD ┃ 
exp_r6-pruningpredicate-rates ┃ Change ┃
   
┡━━━╇━━━╇━━━╇┩
   │ QQuery 0  │  1.19 / 4.69 ±6.90 / 18.48 ms │  1.22 / 4.73 
±6.88 / 18.48 ms │  no change │
   │ QQuery 1  │12.42 / 12.83 ±0.22 / 13.07 ms │13.41 / 14.25 
±0.46 / 14.80 ms │   1.11x slower │
   │ QQuery 2  │36.24 / 37.00 ±0.60 / 37.75 ms │36.37 / 36.60 
±0.21 / 36.97 ms │  no change │
   │ QQuery 3  │30.76 / 31.57 ±0.85 / 32.81 ms │31.43 / 31.69 
±0.32 / 32.30 ms │  no change │
   │ QQuery 4  │ 232.92 / 234.43 ±1.42 / 236.13 ms │230.89 / 237.03 
±10.40 / 257.69 ms │  no change │
   │ QQuery 5  │ 292.00 / 299.53 ±5.98 / 307.25 ms │ 274.20 / 277.03 
±2.47 / 280.12 ms │  +1.08x faster │
   │ QQuery 6  │  6.75 / 8.25 ±1.64 / 11.42 ms │   5.38 / 5.68 
±0.32 / 6.22 ms │  +1.45x faster │
   │ QQuery 7  │13.80 / 14.13 ±0.24 / 14.54 ms │14.88 / 15.18 
±0.21 / 15.42 ms │   1.07x slower │
   │ QQuery 8  │316.80 / 340.57 ±13.96 / 352.30 ms │ 316.98 / 329.85 
±7.90 / 339.90 ms │  no change │
   │ QQuery 9  │453.46 / 479.04 ±13.61 / 493.93 ms │452.07 / 481.24 
±16.67 / 501.98 ms │  no change │
   │ QQuery 10 │69.99 / 73.96 ±4.54 / 82.08 ms │70.14 / 71.41 
±0.68 / 72.06 ms │  no change │
   │ QQuery 11 │80.81 / 82.19 ±1.51 / 85.00 ms │81.85 / 83.54 
±2.51 / 88.52 ms │  no change │
   │ QQuery 12 │ 274.03 / 280.63 ±6.02 / 289.99 ms │ 275.91 / 279.80 
±3.07 / 284.47 ms │  no change │
   │ QQuery 13 │ 389.59 / 398.66 ±6.66 / 405.44 ms │ 414.38 / 422.87 
±5.95 / 431.68 ms │   1.06x slower │
   │ QQuery 14 │278.11 / 290.70 ±10.17 / 304.98 ms │ 280.89 / 290.75 
±7.10 / 298.47 ms │

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448135053

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448018436)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark clickbench_partitioned.json
   
   
┏━━━┳━━━┳━━━┳━━━┓
   ┃ Query ┃  HEAD ┃ 
exp_r6-pruningpredicate-rates ┃Change ┃
   
┡━━━╇━━━╇━━━╇━━━┩
   │ QQuery 0  │  1.19 / 4.64 ±6.78 / 18.20 ms │  1.20 / 4.60 
±6.73 / 18.06 ms │ no change │
   │ QQuery 1  │12.71 / 12.78 ±0.05 / 12.84 ms │14.26 / 14.56 
±0.22 / 14.79 ms │  1.14x slower │
   │ QQuery 2  │35.29 / 35.56 ±0.22 / 35.89 ms │37.96 / 38.21 
±0.28 / 38.76 ms │  1.07x slower │
   │ QQuery 3  │30.38 / 30.86 ±0.51 / 31.67 ms │31.00 / 31.50 
±0.54 / 32.55 ms │ no change │
   │ QQuery 4  │ 229.44 / 232.64 ±2.60 / 236.74 ms │ 228.21 / 232.00 
±2.79 / 235.24 ms │ no change │
   │ QQuery 5  │ 279.19 / 280.84 ±1.56 / 283.61 ms │ 274.44 / 276.89 
±1.86 / 279.97 ms │ no change │
   │ QQuery 6  │   5.90 / 6.19 ±0.25 / 6.64 ms │   5.38 / 5.47 
±0.09 / 5.63 ms │ +1.13x faster │
   │ QQuery 7  │16.06 / 16.25 ±0.15 / 16.50 ms │16.03 / 16.23 
±0.14 / 16.47 ms │ no change │
   │ QQuery 8  │ 310.30 / 314.14 ±2.90 / 317.95 ms │ 307.67 / 313.97 
±3.62 / 318.38 ms │ no change │
   │ QQuery 9  │ 445.85 / 450.72 ±4.28 / 456.55 ms │ 447.67 / 453.13 
±2.91 / 455.99 ms │ no change │
   │ QQuery 10 │94.14 / 95.36 ±1.38 / 97.98 ms │70.07 / 72.23 
±2.62 / 77.23 ms │ +1.32x faster │
   │ QQuery 11 │ 104.35 / 105.78 ±0.90 / 106.94 ms │82.44 / 83.14 
±0.57 / 83.87 ms │ +1.27x faster │
   │ QQuery 12 │ 305.56 / 311.39 ±3.64 / 315.88 ms │ 259.35 / 264.03 
±3.35 / 268.57 ms │ +1.18x faster │
   │ QQuery 13 │ 417.77 / 428.69 ±8.81 / 443.96 ms │ 401.15 / 404.05 
±3.79 / 411.20 ms │ +1.06x faster │
   │ QQuery 14 │ 315.76 / 320.19 ±3.81 / 324.58 ms │ 271.41 / 274.72 
±2.74 / 279.53 ms │ +1.17x faster │
   │

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448113343

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448018436)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark tpch_sf1.json
   
   
┏━━━┳━━━┳┳━━━┓
   ┃ Query ┃  HEAD ┃  
exp_r6-pruningpredicate-rates ┃Change ┃
   
┡━━━╇━━━╇╇━━━┩
   │ QQuery 1  │39.53 / 40.30 ±1.28 / 42.84 ms │ 37.62 / 39.16 ±1.10 / 
40.25 ms │ no change │
   │ QQuery 2  │24.19 / 24.83 ±0.75 / 26.07 ms │ 23.25 / 24.13 ±0.59 / 
24.67 ms │ no change │
   │ QQuery 3  │47.34 / 47.55 ±0.18 / 47.85 ms │ 34.44 / 35.26 ±1.23 / 
37.70 ms │ +1.35x faster │
   │ QQuery 4  │21.78 / 22.21 ±0.58 / 23.35 ms │ 17.75 / 18.11 ±0.32 / 
18.64 ms │ +1.23x faster │
   │ QQuery 5  │66.07 / 67.36 ±0.77 / 68.45 ms │ 43.13 / 44.66 ±1.25 / 
46.30 ms │ +1.51x faster │
   │ QQuery 6  │34.65 / 35.16 ±0.57 / 36.12 ms │ 13.68 / 14.21 ±0.58 / 
15.35 ms │ +2.47x faster │
   │ QQuery 7  │56.51 / 57.56 ±1.17 / 59.45 ms │ 39.59 / 40.87 ±0.94 / 
42.05 ms │ +1.41x faster │
   │ QQuery 8  │78.32 / 79.52 ±1.29 / 81.14 ms │ 50.19 / 50.73 ±0.46 / 
51.51 ms │ +1.57x faster │
   │ QQuery 9  │ 104.11 / 105.11 ±0.94 / 106.40 ms │ 63.54 / 64.56 ±0.63 / 
65.54 ms │ +1.63x faster │
   │ QQuery 10 │74.19 / 74.55 ±0.49 / 75.48 ms │ 63.29 / 64.05 ±0.50 / 
64.74 ms │ +1.16x faster │
   │ QQuery 11 │14.94 / 15.11 ±0.15 / 15.33 ms │ 13.40 / 14.03 ±0.92 / 
15.84 ms │ +1.08x faster │
   │ QQuery 12 │44.77 / 46.22 ±1.55 / 48.18 ms │ 25.37 / 26.10 ±0.87 / 
27.79 ms │ +1.77x faster │
   │ QQuery 13 │45.79 / 46.34 ±0.29 / 46.62 ms │ 36.10 / 36.87 ±0.58 / 
37.67 ms │ +1.26x faster │
   │ QQuery 14 │39.05 / 39.26 ±0.15 / 39.52 ms │ 29.64 / 30.16 ±0.51 / 
31.10 ms │ +1.30x faster │
   │ QQuery 15 │41.12 / 41.96 ±0.63 / 42.67 ms │ 30.79 / 31.26 ±0.58 / 
32.38 ms │ +1.34x faster │
   │ QQuery 16 │22.64 / 22.73 ±0.09 / 22.88 ms │ 18.20 / 18.43 ±0.26 / 
18.91 ms │ +1.23x faster │
   │ QQuery 17 │ 152.80 / 157.34 ±2.41 / 159.54 ms │ 77.26 / 78.81 ±1.93 / 
82.58 ms │ +2.00x faster │
   │ QQuer

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448120994

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448018436)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark tpcds_sf1.json
   
   
┏━━━┳━━━┳┳━━━┓
   ┃ Query ┃  HEAD ┃  
exp_r6-pruningpredicate-rates ┃Change ┃
   
┡━━━╇━━━╇╇━━━┩
   │ QQuery 1  │   5.98 / 6.41 ±0.80 / 8.02 ms │6.12 / 6.57 
±0.86 / 8.29 ms │ no change │
   │ QQuery 2  │46.28 / 46.52 ±0.25 / 46.98 ms │ 50.10 / 50.48 
±0.41 / 51.21 ms │  1.09x slower │
   │ QQuery 3  │27.56 / 28.14 ±0.36 / 28.59 ms │ 30.29 / 30.67 
±0.34 / 31.31 ms │  1.09x slower │
   │ QQuery 4  │ 293.30 / 298.18 ±4.40 / 306.03 ms │  505.73 / 510.70 ±3.43 
/ 515.42 ms │  1.71x slower │
   │ QQuery 5  │79.70 / 79.84 ±0.13 / 80.07 ms │ 54.26 / 56.69 
±2.34 / 61.14 ms │ +1.41x faster │
   │ QQuery 6  │30.27 / 30.41 ±0.14 / 30.66 ms │ 36.21 / 36.95 
±0.40 / 37.31 ms │  1.21x slower │
   │ QQuery 7  │ 141.08 / 142.48 ±1.30 / 144.71 ms │  119.65 / 120.29 ±0.49 
/ 121.12 ms │ +1.18x faster │
   │ QQuery 8  │20.73 / 21.03 ±0.19 / 21.24 ms │ 18.68 / 18.88 
±0.18 / 19.18 ms │ +1.11x faster │
   │ QQuery 9  │ 133.61 / 136.18 ±2.30 / 139.19 ms │ 55.11 / 56.30 
±0.72 / 57.35 ms │ +2.42x faster │
   │ QQuery 10 │ 116.96 / 120.16 ±5.30 / 130.73 ms │ 86.16 / 86.81 
±0.40 / 87.34 ms │ +1.38x faster │
   │ QQuery 11 │ 200.03 / 201.54 ±1.24 / 203.48 ms │  326.03 / 332.81 ±3.98 
/ 337.76 ms │  1.65x slower │
   │ QQuery 12 │21.95 / 22.19 ±0.26 / 22.70 ms │ 32.26 / 32.65 
±0.28 / 33.11 ms │  1.47x slower │
   │ QQuery 13 │ 146.03 / 148.39 ±3.45 / 155.20 ms │  159.81 / 160.61 ±0.48 
/ 161.02 ms │  1.08x slower │
   │ QQuery 14 │ 529.94 / 535.59 ±3.41 / 539.58 ms │  532.77 / 535.23 ±1.94 
/ 538.42 ms │ no change │
   │ QQuery 15 │19.28 / 19.84 ±0.31 / 20.14 ms │ 27.38 / 27.98 
±0.56 / 29.04 ms │  1.41x slower │
   │ QQuery 16 │   6.50 / 6.66 ±0.20 / 7.06 ms │6.55 /

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448121304

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448016719)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark tpcds_sf1.json
   
   
┏━━━┳┳━━━┳━━━┓
   ┃ Query ┃   HEAD ┃ 
exp_r6-pruningpredicate-rates ┃Change ┃
   
┡━━━╇╇━━━╇━━━┩
   │ QQuery 1  │6.40 / 6.91 ±0.90 / 8.70 ms │   6.08 / 6.62 
±0.88 / 8.37 ms │ no change │
   │ QQuery 2  │ 84.07 / 84.22 ±0.12 / 84.38 ms │49.48 / 49.79 
±0.38 / 50.53 ms │ +1.69x faster │
   │ QQuery 3  │ 29.82 / 30.08 ±0.20 / 30.32 ms │29.94 / 30.32 
±0.25 / 30.71 ms │ no change │
   │ QQuery 4  │  533.43 / 542.87 ±6.07 / 550.19 ms │ 495.50 / 498.16 ±2.70 
/ 503.23 ms │ +1.09x faster │
   │ QQuery 5  │ 53.30 / 53.94 ±0.53 / 54.84 ms │54.70 / 55.08 
±0.40 / 55.79 ms │ no change │
   │ QQuery 6  │ 35.76 / 36.24 ±0.35 / 36.73 ms │35.88 / 36.14 
±0.19 / 36.40 ms │ no change │
   │ QQuery 7  │  112.53 / 113.74 ±1.54 / 116.74 ms │ 117.84 / 118.96 ±0.59 
/ 119.42 ms │ no change │
   │ QQuery 8  │ 39.86 / 40.24 ±0.34 / 40.70 ms │18.58 / 19.29 
±0.98 / 21.23 ms │ +2.09x faster │
   │ QQuery 9  │ 55.10 / 56.44 ±1.41 / 58.86 ms │52.63 / 56.06 
±2.79 / 60.49 ms │ no change │
   │ QQuery 10 │ 83.07 / 84.37 ±1.73 / 87.72 ms │86.57 / 86.86 
±0.24 / 87.17 ms │ no change │
   │ QQuery 11 │  332.71 / 337.05 ±2.77 / 341.16 ms │ 324.20 / 326.64 ±2.07 
/ 330.12 ms │ no change │
   │ QQuery 12 │ 29.15 / 29.92 ±0.57 / 30.89 ms │31.82 / 32.26 
±0.24 / 32.51 ms │  1.08x slower │
   │ QQuery 13 │  130.63 / 131.05 ±0.26 / 131.46 ms │ 158.47 / 159.13 ±0.73 
/ 160.56 ms │  1.21x slower │
   │ QQuery 14 │  520.48 / 523.82 ±2.94 / 528.09 ms │ 532.21 / 533.44 ±1.04 
/ 534.98 ms │ no change │
   │ QQuery 15 │ 62.53 / 62.83 ±0.28 / 63.33 ms │27.70 / 27.93 
±0.31 / 28.55 ms │ +2.25x faster │
   │ QQuery 16 │7.08 / 7.33 ±0.27 / 7.85 ms │   6.56 /

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448113720

   🤖 Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448016719)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   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
   
   
   ```
   Comparing HEAD and exp_r6-pruningpredicate-rates
   
   Benchmark tpch_sf1.json
   
   
┏━━━┳┳┳━━━┓
   ┃ Query ┃   HEAD ┃  
exp_r6-pruningpredicate-rates ┃Change ┃
   
┡━━━╇╇╇━━━┩
   │ QQuery 1  │ 39.29 / 41.24 ±2.68 / 46.24 ms │ 38.32 / 40.10 ±1.75 / 42.38 
ms │ no change │
   │ QQuery 2  │ 21.07 / 21.71 ±0.77 / 23.22 ms │ 23.50 / 23.89 ±0.40 / 24.66 
ms │  1.10x slower │
   │ QQuery 3  │ 36.03 / 38.16 ±1.83 / 40.92 ms │ 34.77 / 36.45 ±1.50 / 38.60 
ms │ no change │
   │ QQuery 4  │ 17.52 / 18.03 ±0.52 / 18.98 ms │ 18.02 / 18.71 ±0.60 / 19.49 
ms │ no change │
   │ QQuery 5  │ 43.72 / 46.02 ±1.44 / 47.59 ms │ 46.17 / 46.33 ±0.13 / 46.56 
ms │ no change │
   │ QQuery 6  │ 16.60 / 16.97 ±0.29 / 17.46 ms │ 13.85 / 14.11 ±0.15 / 14.32 
ms │ +1.20x faster │
   │ QQuery 7  │ 49.96 / 50.66 ±0.45 / 51.17 ms │ 40.37 / 41.42 ±1.40 / 44.09 
ms │ +1.22x faster │
   │ QQuery 8  │ 46.10 / 46.83 ±0.71 / 48.15 ms │ 50.80 / 51.17 ±0.36 / 51.64 
ms │  1.09x slower │
   │ QQuery 9  │ 51.93 / 53.32 ±0.96 / 54.38 ms │ 64.54 / 65.42 ±0.94 / 67.07 
ms │  1.23x slower │
   │ QQuery 10 │ 65.01 / 65.69 ±1.04 / 67.75 ms │ 64.15 / 64.61 ±0.45 / 65.43 
ms │ no change │
   │ QQuery 11 │ 13.86 / 14.46 ±0.70 / 15.81 ms │ 13.56 / 13.88 ±0.26 / 14.27 
ms │ no change │
   │ QQuery 12 │ 25.77 / 26.14 ±0.33 / 26.72 ms │ 25.22 / 25.92 ±1.19 / 28.28 
ms │ no change │
   │ QQuery 13 │ 36.88 / 37.80 ±0.70 / 38.98 ms │ 36.26 / 36.96 ±0.76 / 38.45 
ms │ no change │
   │ QQuery 14 │ 26.28 / 26.48 ±0.11 / 26.57 ms │ 29.73 / 31.86 ±1.63 / 33.62 
ms │  1.20x slower │
   │ QQuery 15 │ 31.91 / 32.88 ±0.72 / 33.64 ms │ 31.15 / 31.46 ±0.26 / 31.77 
ms │ no change │
   │ QQuery 16 │ 15.22 / 15.35 ±0.07 / 15.42 ms │ 18.62 / 19.03 ±0.31 / 19.57 
ms │  1.24x slower │
   │ QQuery 17 │ 77.47 / 79.16 ±1.67 / 82.23 ms │ 80.54 / 82.05 ±1.06 / 83.70 
ms │ no change │
   │ QQuery 18 │ 69.32 / 71.05 ±0.88 / 71.73 ms │ 58.04 / 58.54 ±0.29

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448036917

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448018436)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4448018436-67-df28l 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e) to main 
[diff](https://github.com/apache/datafusion/compare/7f2f78d48b6d3d6aee2ce2fd29910bb4c11b1012..cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448033500

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448018436)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4448018436-66-tr92c 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e) to main 
[diff](https://github.com/apache/datafusion/compare/7f2f78d48b6d3d6aee2ce2fd29910bb4c11b1012..cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448031763

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448016719)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4448016719-63-pzk8r 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e) to main 
[diff](https://github.com/apache/datafusion/compare/7f2f78d48b6d3d6aee2ce2fd29910bb4c11b1012..cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448035196

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448016719)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4448016719-62-t4w2x 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e) to main 
[diff](https://github.com/apache/datafusion/compare/7f2f78d48b6d3d6aee2ce2fd29910bb4c11b1012..cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448034157

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448016719)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4448016719-64-lwm96 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e) to main 
[diff](https://github.com/apache/datafusion/compare/7f2f78d48b6d3d6aee2ce2fd29910bb4c11b1012..cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangbot commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448033550

   🤖 Benchmark running (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/22144#issuecomment-4448018436)
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux 
bench-c4448018436-65-6tf2q 6.12.68+ #1 SMP Wed Apr  1 02:23:28 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 HEAD (cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e) to main 
[diff](https://github.com/apache/datafusion/compare/7f2f78d48b6d3d6aee2ce2fd29910bb4c11b1012..cae5fe60525c82423b6cfcdbfe1a73a98d5c7a0e)
 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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangb commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448018436

   run benchmarks
   
   ```yaml
   baseline:
   ref: main
   env:
  DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: true
  DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: true
   changed:
   ref: HEAD
   env:
  DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: true
  DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: true
   ```


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



adriangb commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4448016719

   run benchmarks
   
   ```yaml
   baseline:
   ref: main
   env:
  DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: false
  DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: false
   changed:
   ref: HEAD
   env:
  DATAFUSION_EXECUTION_PARQUET_PUSHDOWN_FILTERS: true
  DATAFUSION_EXECUTION_PARQUET_REORDER_FILTERS: true
   ```


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



github-actions[bot] commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-204560

   
   Thank you for opening this pull request!
   
   Reviewer note: 
[cargo-semver-checks](https://github.com/obi1kenobi/cargo-semver-checks) 
reported the current version number is not SemVer-compatible with the changes 
in this pull request (compared against the base branch).
   
   
   Details
   
   ```
Cloning apache/main
   Building datafusion v53.1.0 (current)
   error: running cargo-doc on crate 'datafusion' failed with output:
   -
  Compiling proc-macro2 v1.0.106
  Compiling quote v1.0.45
  Compiling unicode-ident v1.0.24
  Compiling libc v0.2.186
   Checking cfg-if v1.0.4
  Compiling shlex v1.3.0
  Compiling autocfg v1.5.0
  Compiling find-msvc-tools v0.1.9
  Compiling libm v0.2.16
  Compiling syn v2.0.117
  Compiling num-traits v0.2.19
   Checking memchr v2.8.0
  Compiling jobserver v0.1.34
  Compiling version_check v0.9.5
  Compiling cc v1.2.62
   Checking bytes v1.11.1
  Compiling zerocopy v0.8.48
  Compiling serde_core v1.0.228
   Checking once_cell v1.21.4
  Compiling getrandom v0.3.4
   Checking itoa v1.0.18
  Compiling zmij v1.0.21
   Checking num-integer v0.1.46
  Compiling serde_json v1.0.149
  Compiling serde v1.0.228
   Checking num-bigint v0.4.6
   Checking iana-time-zone v0.1.65
   Checking equivalent v1.0.2
  Compiling pkg-config v0.3.33
   Checking allocator-api2 v0.2.21
   Checking siphasher v1.0.3
   Checking foldhash v0.2.0
  Compiling synstructure v0.13.2
   Checking hashbrown v0.17.1
   Checking phf_shared v0.12.1
   Checking chrono v0.4.44
  Compiling ahash v0.8.12
  Compiling chrono-tz v0.10.4
   Checking stable_deref_trait v1.2.1
   Checking phf v0.12.1
   Checking num-complex v0.4.6
  Compiling zstd-sys v2.0.16+zstd.1.5.7
   Checking pin-project-lite v0.2.17
   Checking litemap v0.8.2
   Checking futures-core v0.3.32
   Checking writeable v0.6.3
   Checking smallvec v1.15.1
   Checking lexical-util v1.0.7
  Compiling zerocopy-derive v0.8.48
  Compiling serde_derive v1.0.228
  Compiling zerofrom-derive v0.1.7
   Checking zerofrom v0.1.8
  Compiling yoke-derive v0.8.2
   Checking arrow-schema v58.3.0
   Checking half v2.7.1
   Checking arrow-buffer v58.3.0
  Compiling zerovec-derive v0.11.3
   Checking yoke v0.8.2
  Compiling displaydoc v0.2.5
   Checking arrow-data v58.3.0
  Compiling object v0.37.3
  Compiling zstd-safe v7.2.4
   Checking arrow-array v58.3.0
   Checking futures-sink v0.3.32
   Checking zerotrie v0.2.4
   Checking zerovec v0.11.6
  Compiling icu_properties_data v2.2.0
   Checking tinystr v0.8.3
   Checking icu_locale_core v2.2.0
   Checking potential_utf v0.1.5
  Compiling icu_normalizer_data v2.2.0
   Checking utf8_iter v1.0.4
   Checking icu_collections v2.2.0
   Checking icu_provider v2.2.0
  Compiling tokio-macros v2.7.0
   Checking arrow-select v58.3.0
  Compiling semver v1.0.28
  Compiling crc32fast v1.5.0
  Compiling rustc_version v0.4.1
   Checking tokio v1.52.3
   Checking futures-channel v0.3.32
   Checking lexical-write-integer v1.0.6
   Checking lexical-parse-integer v1.0.6
  Compiling futures-macro v0.3.32
  Compiling parking_lot_core v0.9.12
   Checking futures-io v0.3.32
   Checking futures-task v0.3.32
   Checking simd-adler32 v0.3.9
   Checking adler2 v2.0.1
   Checking slab v0.4.12
   Checking bitflags v2.11.1
   Checking futures-util v0.3.32
   Checking miniz_oxide v0.8.9
  Compiling ar_archive_writer v0.5.1
   Checking lexical-parse-float v1.0.6
   Checking lexical-write-float v1.0.6
   Checking icu_normalizer v2.2.0
   Checking icu_properties v2.2.0
  Compiling psm v0.1.31
  Compiling flatbuffers v25.12.19
   Checking aho-corasick v1.1.4
   Checking zstd v0.13.3
   Checking scopeguard v1.2.0
   Checking regex-syntax v0.8.10
   Checking zlib-rs v0.6.3
   Checking unicode-segmentation v1.13.2
   Checking ryu v1.0.23
   Checking unicode-width v0.2.2
   Checking base64 v0.22.1
  Compiling getrandom v0.4.2
   Checking comfy-table v7.2.2
   Checking lock_api v0.4.14
   Checking idna_adapter v1.2.2
   Checking lexical-core v1.0.6
   Checking arrow-ord v58.3.0
   Checking indexmap v2.14.0
   Checking flate2 v1.1.9
   Checking regex-automata v0.4.14
  Compiling stacker v0.1.24
   Checking atoi v2.0.0
   Checking percent-encoding v2.3.2
   Checking alloc-no-stdlib v2.0.4
   Checking twox-hash v2.1.2
  Compiling snap v1.1.1
  Compiling thiserror v2.0.18
   Checking lz4_flex v0.13.1
   Checking alloc-stdlib v0

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

2026-05-13 Thread via GitHub



alamb commented on PR #22144:
URL: https://github.com/apache/datafusion/pull/22144#issuecomment-4443550381

   Perhaps if we had an API such as described here, this would be easier to 
implement
   - 
https://github.com/apache/datafusion/pull/22024#pullrequestreview-4275922067


-- 
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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


-
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

Re: [PR] [Experiment] Adaptive filter pushdown [datafusion]

34 matches

Site Navigation

Mail list logo

Footer information