adriangbot commented on PR #26250: URL: https://github.com/apache/datafusion/pull/26250#issuecomment-6284631919
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/26250#issuecomment-6284388928) **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) Comparing cb09c73e3 (cb09c73e3365f5421a21b3f39826beb43e28ce96) to cb09c73e3 [diff](https://github.com/apache/datafusion/compare/cb09c73e3365f5421a21b3f39826beb43e28ce96..cb09c73e3365f5421a21b3f39826beb43e28ce96) <details><summary>Run configuration</summary> ```yaml run benchmark clickbench_partitioned env: DATAFUSION_EXECUTION_PARQUET_SCHEMA_FORCE_VIEW_TYPES: "true" DATAFUSION_RUNTIME_MEMORY_LIMIT: "32G" baseline: ref: "cb09c73e3" env: DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "false" changed: ref: "cb09c73e3" env: DATAFUSION_EXECUTION_PARQUET_ENABLE_RLE_TO_DICTIONARY: "true" ``` </details> <details><summary>CPU Details (lscpu)</summary> ``` Architecture: aarch64 CPU op-mode(s): 64-bit Byte Order: Little Endian CPU(s): 16 On-line CPU(s) list: 0-15 Vendor ID: ARM Model name: Neoverse-V2 Model: 1 Thread(s) per core: 1 Core(s) per cluster: 16 Socket(s): - Cluster(s): 1 Stepping: r0p1 BogoMIPS: 2000.00 Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti L1d cache: 1 MiB (16 instances) L1i cache: 1 MiB (16 instances) L2 cache: 32 MiB (16 instances) L3 cache: 80 MiB (1 instance) NUMA node(s): 1 NUMA node0 CPU(s): 0-15 Vulnerability Gather data sampling: Not affected Vulnerability Indirect target selection: Not affected Vulnerability Itlb multihit: Not affected Vulnerability L1tf: Not affected Vulnerability Mds: Not affected Vulnerability Meltdown: Not affected Vulnerability Mmio stale data: Not affected Vulnerability Reg file data sampling: Not affected Vulnerability Retbleed: Not affected Vulnerability Spec rstack overflow: Not affected Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl Vulnerability Spectre v1: Mitigation; __user pointer sanitization Vulnerability Spectre v2: Mitigation; CSV2, BHB Vulnerability Srbds: Not affected Vulnerability Tsa: Not affected Vulnerability Tsx async abort: Not affected Vulnerability Vmscape: Not affected ``` </details> <details><summary>Details</summary> <p> ``` Comparing HEAD and dict-encoding-bench -------------------- Benchmark clickbench_partitioned.json -------------------- βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββ³ββββββββββββββββ β Query β HEAD β dict-encoding-bench β Change β β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ© β QQuery 0 β 1.23 ms β 1.15 ms β +1.06x faster β β QQuery 1 β 11.86 ms β 12.12 ms β no change β β QQuery 2 β 37.34 ms β 37.10 ms β no change β β QQuery 3 β 31.66 ms β 32.17 ms β no change β β QQuery 4 β 245.32 ms β 247.78 ms β no change β β QQuery 5 β 282.83 ms β 599.41 ms β 2.12x slower β β QQuery 6 β 1.22 ms β 1.20 ms β no change β β QQuery 7 β 13.45 ms β 13.33 ms β no change β β QQuery 8 β 348.49 ms β 351.78 ms β no change β β QQuery 9 β 480.64 ms β 492.88 ms β no change β β QQuery 10 β 64.95 ms β 70.39 ms β 1.08x slower β β QQuery 11 β 75.31 ms β 81.03 ms β 1.08x slower β β QQuery 12 β 267.66 ms β 536.52 ms β 2.00x slower β β QQuery 13 β 362.10 ms β 818.04 ms β 2.26x slower β β QQuery 14 β 284.13 ms β 667.75 ms β 2.35x slower β β QQuery 15 β 292.77 ms β 298.15 ms β no change β β QQuery 16 β 646.19 ms β 1190.40 ms β 1.84x slower β β QQuery 17 β 640.67 ms β 1158.69 ms β 1.81x slower β β QQuery 18 β 1296.07 ms β 2020.58 ms β 1.56x slower β β QQuery 19 β 27.56 ms β 27.70 ms β no change β β QQuery 20 β 521.46 ms β 865.54 ms β 1.66x slower β β QQuery 21 β 518.40 ms β 1053.65 ms β 2.03x slower β β QQuery 22 β 1026.85 ms β 2185.39 ms β 2.13x slower β β QQuery 23 β 3133.98 ms β 4975.82 ms β 1.59x slower β β QQuery 24 β 42.79 ms β 78.47 ms β 1.83x slower β β QQuery 25 β 106.54 ms β 227.54 ms β 2.14x slower β β QQuery 26 β 40.53 ms β 78.13 ms β 1.93x slower β β QQuery 27 β 513.93 ms β 931.37 ms β 1.81x slower β β QQuery 28 β 2910.74 ms β 3384.85 ms β 1.16x slower β β QQuery 29 β 40.51 ms β 40.24 ms β no change β β QQuery 30 β 318.41 ms β 423.06 ms β 1.33x slower β β QQuery 31 β 290.12 ms β 403.19 ms β 1.39x slower β β QQuery 32 β 1144.66 ms β 1093.44 ms β no change β β QQuery 33 β 1639.75 ms β 2599.95 ms β 1.59x slower β β QQuery 34 β 1667.80 ms β 2653.98 ms β 1.59x slower β β QQuery 35 β 302.27 ms β 309.25 ms β no change β β QQuery 36 β 72.38 ms β 113.03 ms β 1.56x slower β β QQuery 37 β 35.70 ms β 58.48 ms β 1.64x slower β β QQuery 38 β 42.90 ms β 71.52 ms β 1.67x slower β β QQuery 39 β 144.82 ms β 233.77 ms β 1.61x slower β β QQuery 40 β 14.77 ms β 14.54 ms β no change β β QQuery 41 β 13.96 ms β 14.00 ms β no change β β QQuery 42 β 13.71 ms β 13.64 ms β no change β βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββ΄ββββββββββββββββ ββββββββββββββββββββββββββββββββββββββ³βββββββββββββ β Benchmark Summary β β β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ© β Total Time (HEAD) β 19968.44ms β β Total Time (dict-encoding-bench) β 30481.05ms β β Average Time (HEAD) β 464.38ms β β Average Time (dict-encoding-bench) β 708.86ms β β Queries Faster β 1 β β Queries Slower β 26 β β Queries with No Change β 16 β β Queries with Failure β 0 β ββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ Distribution per query (min / mean Β±stddev / max): Comparing HEAD and dict-encoding-bench -------------------- Benchmark clickbench_partitioned.json -------------------- βββββββββββββ³βββββββββββββββββββββββββββββββββββββββββ³βββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ β Query β HEAD β dict-encoding-bench β Change β β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ© β QQuery 0 β 1.23 / 4.35 Β±6.08 / 16.51 ms β 1.15 / 4.33 Β±6.19 / 16.72 ms β no change β β QQuery 1 β 11.86 / 12.03 Β±0.10 / 12.13 ms β 12.12 / 12.34 Β±0.16 / 12.57 ms β no change β β QQuery 2 β 37.34 / 37.57 Β±0.21 / 37.82 ms β 37.10 / 37.37 Β±0.26 / 37.83 ms β no change β β QQuery 3 β 31.66 / 32.69 Β±0.85 / 33.82 ms β 32.17 / 32.63 Β±0.41 / 33.33 ms β no change β β QQuery 4 β 245.32 / 249.64 Β±2.58 / 253.38 ms β 247.78 / 253.40 Β±5.51 / 261.27 ms β no change β β QQuery 5 β 282.83 / 285.37 Β±2.03 / 287.45 ms β 599.41 / 609.01 Β±7.21 / 621.64 ms β 2.13x slower β β QQuery 6 β 1.22 / 1.38 Β±0.25 / 1.88 ms β 1.20 / 1.36 Β±0.24 / 1.82 ms β no change β β QQuery 7 β 13.45 / 15.00 Β±3.04 / 21.07 ms β 13.33 / 13.58 Β±0.15 / 13.70 ms β +1.10x faster β β QQuery 8 β 348.49 / 353.24 Β±4.67 / 359.19 ms β 351.78 / 360.02 Β±6.36 / 367.99 ms β no change β β QQuery 9 β 480.64 / 488.56 Β±5.65 / 497.62 ms β 492.88 / 506.44 Β±10.81 / 519.41 ms β no change β β QQuery 10 β 64.95 / 68.85 Β±6.03 / 80.81 ms β 70.39 / 72.93 Β±3.26 / 79.34 ms β 1.06x slower β β QQuery 11 β 75.31 / 75.65 Β±0.26 / 76.02 ms β 81.03 / 82.44 Β±1.37 / 84.65 ms β 1.09x slower β β QQuery 12 β 267.66 / 275.69 Β±6.10 / 284.75 ms β 536.52 / 555.48 Β±14.19 / 578.19 ms β 2.01x slower β β QQuery 13 β 362.10 / 376.49 Β±14.34 / 399.24 ms β 818.04 / 847.70 Β±22.25 / 876.53 ms β 2.25x slower β β QQuery 14 β 284.13 / 291.80 Β±7.90 / 305.81 ms β 667.75 / 675.59 Β±5.32 / 680.46 ms β 2.32x slower β β QQuery 15 β 292.77 / 300.41 Β±4.19 / 304.86 ms β 298.15 / 304.08 Β±5.66 / 314.37 ms β no change β β QQuery 16 β 646.19 / 649.88 Β±2.73 / 653.37 ms β 1190.40 / 1208.27 Β±21.12 / 1248.88 ms β 1.86x slower β β QQuery 17 β 640.67 / 647.39 Β±6.85 / 660.03 ms β 1158.69 / 1182.67 Β±23.08 / 1223.76 ms β 1.83x slower β β QQuery 18 β 1296.07 / 1342.47 Β±31.46 / 1386.72 ms β 2020.58 / 2048.57 Β±16.51 / 2068.62 ms β 1.53x slower β β QQuery 19 β 27.56 / 40.57 Β±13.18 / 63.54 ms β 27.70 / 30.30 Β±4.90 / 40.10 ms β +1.34x faster β β QQuery 20 β 521.46 / 528.66 Β±8.98 / 543.92 ms β 865.54 / 870.27 Β±4.64 / 878.21 ms β 1.65x slower β β QQuery 21 β 518.40 / 522.52 Β±2.59 / 526.18 ms β 1053.65 / 1067.11 Β±8.02 / 1076.66 ms β 2.04x slower β β QQuery 22 β 1026.85 / 1062.04 Β±27.30 / 1106.88 ms β 2185.39 / 2269.76 Β±71.56 / 2398.55 ms β 2.14x slower β β QQuery 23 β 3133.98 / 3294.95 Β±111.86 / 3438.53 ms β 4975.82 / 5302.98 Β±282.16 / 5680.75 ms β 1.61x slower β β QQuery 24 β 42.79 / 124.42 Β±55.10 / 210.59 ms β 78.47 / 220.21 Β±280.15 / 780.51 ms β 1.77x slower β β QQuery 25 β 106.54 / 142.55 Β±70.34 / 283.21 ms β 227.54 / 237.02 Β±14.78 / 266.27 ms β 1.66x slower β β QQuery 26 β 40.53 / 42.61 Β±2.17 / 46.33 ms β 78.13 / 78.85 Β±0.55 / 79.79 ms β 1.85x slower β β QQuery 27 β 513.93 / 522.18 Β±7.94 / 537.29 ms β 931.37 / 938.34 Β±7.57 / 948.85 ms β 1.80x slower β β QQuery 28 β 2910.74 / 2932.87 Β±22.86 / 2973.08 ms β 3384.85 / 3464.43 Β±86.74 / 3620.68 ms β 1.18x slower β β QQuery 29 β 40.51 / 43.70 Β±6.09 / 55.86 ms β 40.24 / 42.70 Β±4.19 / 51.06 ms β no change β β QQuery 30 β 318.41 / 327.78 Β±10.93 / 348.93 ms β 423.06 / 433.05 Β±7.58 / 444.36 ms β 1.32x slower β β QQuery 31 β 290.12 / 298.61 Β±8.59 / 314.70 ms β 403.19 / 418.14 Β±12.77 / 434.19 ms β 1.40x slower β β QQuery 32 β 1144.66 / 1251.26 Β±97.11 / 1419.66 ms β 1093.44 / 1194.66 Β±77.30 / 1326.71 ms β no change β β QQuery 33 β 1639.75 / 1650.85 Β±13.77 / 1677.68 ms β 2599.95 / 2719.55 Β±93.07 / 2843.18 ms β 1.65x slower β β QQuery 34 β 1667.80 / 1697.92 Β±30.83 / 1738.21 ms β 2653.98 / 2691.33 Β±22.00 / 2721.96 ms β 1.59x slower β β QQuery 35 β 302.27 / 382.98 Β±92.11 / 514.72 ms β 309.25 / 335.55 Β±30.47 / 387.43 ms β +1.14x faster β β QQuery 36 β 72.38 / 84.13 Β±17.22 / 118.31 ms β 113.03 / 123.54 Β±8.40 / 135.33 ms β 1.47x slower β β QQuery 37 β 35.70 / 40.57 Β±4.93 / 48.45 ms β 58.48 / 66.93 Β±7.54 / 79.18 ms β 1.65x slower β β QQuery 38 β 42.90 / 49.59 Β±5.61 / 59.39 ms β 71.52 / 72.97 Β±1.09 / 74.32 ms β 1.47x slower β β QQuery 39 β 144.82 / 154.32 Β±5.48 / 161.73 ms β 233.77 / 256.14 Β±18.37 / 276.69 ms β 1.66x slower β β QQuery 40 β 14.77 / 16.51 Β±2.99 / 22.46 ms β 14.54 / 14.89 Β±0.21 / 15.18 ms β +1.11x faster β β QQuery 41 β 13.96 / 14.74 Β±1.11 / 16.91 ms β 14.00 / 14.27 Β±0.20 / 14.53 ms β no change β β QQuery 42 β 13.71 / 15.32 Β±3.03 / 21.38 ms β 13.64 / 17.14 Β±5.44 / 27.74 ms β 1.12x slower β βββββββββββββ΄βββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ ββββββββββββββββββββββββββββββββββββββ³βββββββββββββ β Benchmark Summary β β β‘ββββββββββββββββββββββββββββββββββββββββββββββββββ© β Total Time (HEAD) β 20750.10ms β β Total Time (dict-encoding-bench) β 31688.34ms β β Average Time (HEAD) β 482.56ms β β Average Time (dict-encoding-bench) β 736.94ms β β Queries Faster β 4 β β Queries Slower β 27 β β Queries with No Change β 12 β β Queries with Failure β 0 β ββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ ``` </p> </details> <details><summary>Memory Pool Peaks</summary> Peak `MemoryPool` reservation per query β what DataFusion's accounting believes it reserved. Recorded only for benchmarks that write a results JSON (the `dfbench` suites, and the suites `bench.sh` runs through the Criterion SQL harness since apache/datafusion#25644), and only under a memory limit: `DATAFUSION_RUNTIME_MEMORY_LIMIT`, or one the suite sets itself. Base: `cb09c73e3` | Changed: `cb09c73e3` **`clickbench_partitioned`** | Query | Base | Changed | Change | | --- | --- | --- | --- | | Query 0 | 0 B | 0 B | 0.0% | | Query 1 | 104 B | 104 B | +0.0% | | Query 2 | 936 B | 936 B | +0.0% | | Query 3 | 312 B | 312 B | +0.0% | | Query 4 | 829.6 MiB | 844.0 MiB | +1.7% | | Query 5 | 1.2 GiB | 1.6 GiB | +32.8% | | Query 6 | 0 B | 0 B | 0.0% | | Query 7 | 50.2 MiB | 50.2 MiB | -0.0% | | Query 8 | 942.3 MiB | 947.2 MiB | +0.5% | | Query 9 | 593.9 MiB | 594.0 MiB | +0.0% | | Query 10 | 119.9 MiB | 91.0 MiB | -24.2% | | Query 11 | 125.6 MiB | 111.8 MiB | -11.0% | | Query 12 | 1.3 GiB | 1.7 GiB | +31.3% | | Query 13 | 1.6 GiB | 1.9 GiB | +20.1% | | Query 14 | 1.3 GiB | 2.0 GiB | +55.5% | | Query 15 | 1.2 GiB | 1.2 GiB | +0.6% | | Query 16 | 3.4 GiB | 4.3 GiB | +25.5% | | Query 17 | 3.4 GiB | 4.2 GiB | +23.2% | | Query 18 | 8.1 GiB | 8.5 GiB | +4.8% | | Query 19 | 0 B | 0 B | 0.0% | | Query 20 | 104 B | 104 B | +0.0% | | Query 21 | 3.7 MiB | 1.1 MiB | -70.9% | | Query 22 | 3.4 MiB | 2.3 MiB | -33.3% | | Query 23 | 3.0 GiB | 9.8 MiB | -99.7% | | Query 24 | 60.1 MiB | 7.2 MiB | -88.0% | | Query 25 | 168.3 MiB | 12.4 MiB | -92.6% | | Query 26 | 59.3 MiB | 8.3 MiB | -86.1% | | Query 27 | 2.3 MiB | 2.6 MiB | +12.7% | | Query 28 | 1.7 GiB | 1.7 GiB | +0.4% | | Query 29 | 624 B | 624 B | +0.0% | | Query 30 | 706.9 MiB | 656.4 MiB | -7.1% | | Query 31 | 1.5 GiB | 1.5 GiB | -4.5% | | Query 32 | 7.1 GiB | 7.2 GiB | +0.7% | | Query 33 | 7.6 GiB | 10.5 GiB | +38.4% | | Query 34 | 7.7 GiB | 10.3 GiB | +34.1% | | Query 35 | 659.8 MiB | 669.4 MiB | +1.5% | | Query 36 | 118.8 MiB | 102.5 MiB | -13.7% | | Query 37 | 7.3 MiB | 7.8 MiB | +6.9% | | Query 38 | 5.2 MiB | 5.2 MiB | +0.2% | | Query 39 | 297.8 MiB | 232.3 MiB | -22.0% | | Query 40 | 2.5 MiB | 2.5 MiB | -0.8% | | Query 41 | 3.1 MiB | 3.1 MiB | +0.0% | | Query 42 | 1.6 MiB | 1.6 MiB | +0.0% | **Pool accounting vs. process RSS** Max pool peak is the largest reservation any single query in the run reached; peak RSS covers the whole invocation, including data loading and allocator retention, and the two high-water marks need not coincide in time. The gap is therefore an upper bound on what the pool did not account for, not a measurement of it. | Benchmark | Side | Max pool peak | Peak RSS | Gap | RSS / pool | | --- | --- | --- | --- | --- | --- | | clickbench_partitioned | base (`cb09c73e3`) | 8.1 GiB | 17.3 GiB | 9.2 GiB | 2.1Γ | | clickbench_partitioned | changed (`cb09c73e3`) | 10.5 GiB | 14.4 GiB | 3.8 GiB | 1.4Γ | </details> <details><summary>Resource Usage</summary> **clickbench_partitioned β base (merge-base)** | Metric | Value | |--------|-------| | Wall time | 105.0s | | Peak memory | 17.3 GiB | | Avg memory | 5.9 GiB | | CPU user | 1016.9s | | CPU sys | 106.7s | | Peak spill | 0 B | **clickbench_partitioned β branch** | Metric | Value | |--------|-------| | Wall time | 160.0s | | Peak memory | 14.4 GiB | | Avg memory | 6.1 GiB | | CPU user | 1522.1s | | CPU sys | 138.0s | | Peak spill | 0 B | </details> --- [File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [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]
