adriangbot commented on PR #24086:
URL: https://github.com/apache/datafusion/pull/24086#issuecomment-5180800619

   πŸ€– Benchmark completed (GKE) | 
[trigger](https://github.com/apache/datafusion/pull/24086#issuecomment-5180420975)
   
   **Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
   
   Comparing claude/datafusion-rowgroup-buffering-ec9abc 
(5ff7af62acc00fbfe33496798fd22245958be41b) to 39d5064 (merge-base) 
[diff](https://github.com/apache/datafusion/compare/39d50641424dfbc9fadad9cad17619c64d06076b..5ff7af62acc00fbfe33496798fd22245958be41b)
   
   <details><summary>Run configuration</summary>
   
   ```yaml
   run benchmark tpch
   env:
     DATAFUSION_RUNTIME_MEMORY_LIMIT: "4G"
     DF_FETCH_BUDGET: "104857600"
     DF_FETCH_POLICY: "streaming"
     SIMULATE_LATENCY: "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 claude_datafusion-rowgroup-buffering-ec9abc
   --------------------
   Benchmark tpch_sf1.json
   --------------------
   
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
   ┃ Query     ┃       HEAD ┃ claude_datafusion-rowgroup-buffering-ec9abc ┃     
   Change ┃
   
┑━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
   β”‚ QQuery 1  β”‚  534.12 ms β”‚                                   298.61 ms β”‚ 
+1.79x faster β”‚
   β”‚ QQuery 2  β”‚  813.48 ms β”‚                                   951.53 ms β”‚  
1.17x slower β”‚
   β”‚ QQuery 3  β”‚  985.52 ms β”‚                                   855.26 ms β”‚ 
+1.15x faster β”‚
   β”‚ QQuery 4  β”‚  707.90 ms β”‚                                   579.89 ms β”‚ 
+1.22x faster β”‚
   β”‚ QQuery 5  β”‚ 1148.40 ms β”‚                                   957.40 ms β”‚ 
+1.20x faster β”‚
   β”‚ QQuery 6  β”‚  533.40 ms β”‚                                   246.16 ms β”‚ 
+2.17x faster β”‚
   β”‚ QQuery 7  β”‚ 1000.29 ms β”‚                                  1024.75 ms β”‚     
no change β”‚
   β”‚ QQuery 8  β”‚ 1117.15 ms β”‚                                  1099.31 ms β”‚     
no change β”‚
   β”‚ QQuery 9  β”‚ 1200.50 ms β”‚                                   934.69 ms β”‚ 
+1.28x faster β”‚
   β”‚ QQuery 10 β”‚  932.86 ms β”‚                                   757.91 ms β”‚ 
+1.23x faster β”‚
   β”‚ QQuery 11 β”‚  742.60 ms β”‚                                   813.57 ms β”‚  
1.10x slower β”‚
   β”‚ QQuery 12 β”‚  727.03 ms β”‚                                   554.46 ms β”‚ 
+1.31x faster β”‚
   β”‚ QQuery 13 β”‚  401.31 ms β”‚                                   502.65 ms β”‚  
1.25x slower β”‚
   β”‚ QQuery 14 β”‚  624.70 ms β”‚                                   453.51 ms β”‚ 
+1.38x faster β”‚
   β”‚ QQuery 15 β”‚ 1188.19 ms β”‚                                   650.69 ms β”‚ 
+1.83x faster β”‚
   β”‚ QQuery 16 β”‚  325.14 ms β”‚                                   440.47 ms β”‚  
1.35x slower β”‚
   β”‚ QQuery 17 β”‚ 1131.37 ms β”‚                                   740.73 ms β”‚ 
+1.53x faster β”‚
   β”‚ QQuery 18 β”‚ 1372.30 ms β”‚                                  1098.58 ms β”‚ 
+1.25x faster β”‚
   β”‚ QQuery 19 β”‚  543.64 ms β”‚                                   402.56 ms β”‚ 
+1.35x faster β”‚
   β”‚ QQuery 20 β”‚ 1050.12 ms β”‚                                   744.68 ms β”‚ 
+1.41x faster β”‚
   β”‚ QQuery 21 β”‚ 2061.98 ms β”‚                                  1285.63 ms β”‚ 
+1.60x faster β”‚
   β”‚ QQuery 22 β”‚  497.37 ms β”‚                                   755.17 ms β”‚  
1.52x slower β”‚
   
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
   ┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
   ┃ Benchmark Summary                                          ┃            ┃
   ┑━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
   β”‚ Total Time (HEAD)                                          β”‚ 19639.36ms β”‚
   β”‚ Total Time (claude_datafusion-rowgroup-buffering-ec9abc)   β”‚ 16148.21ms β”‚
   β”‚ Average Time (HEAD)                                        β”‚   892.70ms β”‚
   β”‚ Average Time (claude_datafusion-rowgroup-buffering-ec9abc) β”‚   734.01ms β”‚
   β”‚ Queries Faster                                             β”‚         15 β”‚
   β”‚ Queries Slower                                             β”‚          5 β”‚
   β”‚ Queries with No Change                                     β”‚          2 β”‚
   β”‚ Queries with Failure                                       β”‚          0 β”‚
   β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
   
   Distribution per query (min / mean Β±stddev / max):
   
   Comparing HEAD and claude_datafusion-rowgroup-buffering-ec9abc
   --------------------
   Benchmark tpch_sf1.json
   --------------------
   
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
   ┃ Query     ┃                                   HEAD ┃ 
claude_datafusion-rowgroup-buffering-ec9abc ┃        Change ┃
   
┑━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
   β”‚ QQuery 1  β”‚     534.12 / 567.20 Β±28.72 / 605.57 ms β”‚          298.61 / 
354.17 Β±39.12 / 399.83 ms β”‚ +1.60x faster β”‚
   β”‚ QQuery 2  β”‚     813.48 / 900.44 Β±51.57 / 954.71 ms β”‚        951.53 / 
1059.50 Β±86.57 / 1151.62 ms β”‚  1.18x slower β”‚
   β”‚ QQuery 3  β”‚   985.52 / 1022.65 Β±28.37 / 1062.90 ms β”‚          855.26 / 
911.01 Β±46.22 / 980.68 ms β”‚ +1.12x faster β”‚
   β”‚ QQuery 4  β”‚     707.90 / 788.82 Β±80.46 / 889.80 ms β”‚          579.89 / 
636.49 Β±57.96 / 721.55 ms β”‚ +1.24x faster β”‚
   β”‚ QQuery 5  β”‚  1148.40 / 1265.64 Β±84.47 / 1388.91 ms β”‚        957.40 / 
1054.37 Β±67.04 / 1138.12 ms β”‚ +1.20x faster β”‚
   β”‚ QQuery 6  β”‚     533.40 / 622.59 Β±74.58 / 759.77 ms β”‚          246.16 / 
293.63 Β±38.70 / 345.52 ms β”‚ +2.12x faster β”‚
   β”‚ QQuery 7  β”‚ 1000.29 / 1154.81 Β±119.39 / 1297.81 ms β”‚       1024.75 / 
1096.82 Β±52.68 / 1168.88 ms β”‚ +1.05x faster β”‚
   β”‚ QQuery 8  β”‚ 1117.15 / 1296.68 Β±124.31 / 1403.26 ms β”‚       1099.31 / 
1160.93 Β±57.18 / 1239.55 ms β”‚ +1.12x faster β”‚
   β”‚ QQuery 9  β”‚  1200.50 / 1311.45 Β±90.12 / 1461.08 ms β”‚       934.69 / 
1065.95 Β±100.26 / 1179.74 ms β”‚ +1.23x faster β”‚
   β”‚ QQuery 10 β”‚   932.86 / 1020.91 Β±65.83 / 1111.16 ms β”‚        757.91 / 
844.55 Β±102.19 / 1025.35 ms β”‚ +1.21x faster β”‚
   β”‚ QQuery 11 β”‚     742.60 / 825.85 Β±84.75 / 974.20 ms β”‚          813.57 / 
870.91 Β±52.07 / 940.90 ms β”‚  1.05x slower β”‚
   β”‚ QQuery 12 β”‚     727.03 / 807.82 Β±67.01 / 887.04 ms β”‚          554.46 / 
588.54 Β±32.23 / 642.38 ms β”‚ +1.37x faster β”‚
   β”‚ QQuery 13 β”‚     401.31 / 440.43 Β±23.39 / 465.00 ms β”‚          502.65 / 
526.66 Β±27.97 / 577.12 ms β”‚  1.20x slower β”‚
   β”‚ QQuery 14 β”‚     624.70 / 705.52 Β±51.83 / 762.65 ms β”‚          453.51 / 
503.03 Β±40.46 / 564.25 ms β”‚ +1.40x faster β”‚
   β”‚ QQuery 15 β”‚  1188.19 / 1272.73 Β±51.08 / 1334.89 ms β”‚          650.69 / 
717.86 Β±71.39 / 846.01 ms β”‚ +1.77x faster β”‚
   β”‚ QQuery 16 β”‚     325.14 / 419.68 Β±72.82 / 550.04 ms β”‚          440.47 / 
517.08 Β±64.04 / 621.81 ms β”‚  1.23x slower β”‚
   β”‚ QQuery 17 β”‚ 1131.37 / 1307.73 Β±102.98 / 1439.47 ms β”‚          740.73 / 
795.30 Β±54.73 / 887.48 ms β”‚ +1.64x faster β”‚
   β”‚ QQuery 18 β”‚  1372.30 / 1496.32 Β±98.09 / 1630.95 ms β”‚        1098.58 / 
1106.98 Β±7.02 / 1119.59 ms β”‚ +1.35x faster β”‚
   β”‚ QQuery 19 β”‚     543.64 / 663.44 Β±65.52 / 738.94 ms β”‚          402.56 / 
488.91 Β±55.34 / 574.50 ms β”‚ +1.36x faster β”‚
   β”‚ QQuery 20 β”‚  1050.12 / 1081.28 Β±34.44 / 1146.65 ms β”‚          744.68 / 
814.44 Β±53.86 / 891.15 ms β”‚ +1.33x faster β”‚
   β”‚ QQuery 21 β”‚  2061.98 / 2127.53 Β±69.70 / 2259.39 ms β”‚       1285.63 / 
1354.45 Β±49.02 / 1417.05 ms β”‚ +1.57x faster β”‚
   β”‚ QQuery 22 β”‚     497.37 / 546.61 Β±31.00 / 579.40 ms β”‚           755.17 / 
755.93 Β±0.76 / 757.14 ms β”‚  1.38x slower β”‚
   
β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
   ┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
   ┃ Benchmark Summary                                          ┃            ┃
   ┑━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
   β”‚ Total Time (HEAD)                                          β”‚ 21646.11ms β”‚
   β”‚ Total Time (claude_datafusion-rowgroup-buffering-ec9abc)   β”‚ 17517.50ms β”‚
   β”‚ Average Time (HEAD)                                        β”‚   983.91ms β”‚
   β”‚ Average Time (claude_datafusion-rowgroup-buffering-ec9abc) β”‚   796.25ms β”‚
   β”‚ Queries Faster                                             β”‚         17 β”‚
   β”‚ Queries Slower                                             β”‚          5 β”‚
   β”‚ Queries with No Change                                     β”‚          0 β”‚
   β”‚ Queries with Failure                                       β”‚          0 β”‚
   β””β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”΄β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”€β”˜
   ```
   
   </p>
   </details>
   
   <details><summary>Memory Pool Peaks</summary>
   
   Peak `MemoryPool` reservation per query β€” what DataFusion's accounting 
believes it reserved. Recorded only when the benchmark runs with 
`DATAFUSION_RUNTIME_MEMORY_LIMIT` set.
   
   Base: `39d5064 (merge-base)` | Changed: 
`claude/datafusion-rowgroup-buffering-ec9abc`
   
   **`tpch` β€” `tpch_sf1`**
   
   | Query | Base | Changed | Change |
   | --- | --- | --- | --- |
   | Query 1 | 40.0 MiB | 40.0 MiB | +0.0% |
   | Query 2 | 29.5 MiB | 29.7 MiB | +0.6% |
   | Query 3 | 10.6 MiB | 10.6 MiB | +0.0% |
   | Query 4 | 33.3 MiB | 33.3 MiB | -0.0% |
   | Query 5 | 31.1 MiB | 40.6 MiB | +30.6% |
   | Query 6 | 0 B | 0 B | 0.0% |
   | Query 7 | 87.2 MiB | 90.0 MiB | +3.3% |
   | Query 8 | 33.0 MiB | 33.4 MiB | +1.1% |
   | Query 9 | 130.7 MiB | 131.3 MiB | +0.5% |
   | Query 10 | 43.2 MiB | 41.5 MiB | -4.0% |
   | Query 11 | 31.5 MiB | 22.1 MiB | -29.9% |
   | Query 12 | 20.0 MiB | 20.0 MiB | +0.0% |
   | Query 13 | 70.1 MiB | 70.1 MiB | -0.0% |
   | Query 14 | 7.5 MiB | 7.5 MiB | +0.0% |
   | Query 15 | 13.0 MiB | 12.9 MiB | -0.6% |
   | Query 16 | 55.4 MiB | 54.8 MiB | -0.9% |
   | Query 17 | 147.6 MiB | 69.3 MiB | -53.1% |
   | Query 18 | 306.8 MiB | 306.8 MiB | +0.0% |
   | Query 19 | 721.1 KiB | 721.1 KiB | +0.0% |
   | Query 20 | 86.6 MiB | 82.5 MiB | -4.7% |
   | Query 21 | 46.8 MiB | 46.8 MiB | +0.0% |
   | Query 22 | 42.8 MiB | 32.2 MiB | -24.8% |
   
   **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 |
   | --- | --- | --- | --- | --- | --- |
   | tpch | base (`39d5064 (merge-base)`) | 306.8 MiB | 871.2 MiB | 564.4 MiB | 
2.8Γ— |
   | tpch | changed (`claude/datafusion-rowgroup-buffering-ec9abc`) | 306.8 MiB 
| 782.7 MiB | 475.9 MiB | 2.6Γ— |
   
   </details>
   
   <details><summary>Resource Usage</summary>
   
   **tpch β€” base (merge-base)**
   | Metric | Value |
   |--------|-------|
   | Wall time | 115.0s |
   | Peak memory | 871.2 MiB |
   | Avg memory | 407.9 MiB |
   | CPU user | 30.1s |
   | CPU sys | 3.9s |
   | Peak spill | 0 B |
   
   **tpch β€” branch**
   | Metric | Value |
   |--------|-------|
   | Wall time | 90.0s |
   | Peak memory | 782.7 MiB |
   | Avg memory | 395.7 MiB |
   | CPU user | 29.8s |
   | CPU sys | 3.7s |
   | Peak spill | 0 B |
   
   </details>
   
   
   ---
   [File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) 
against this benchmark runner


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

To unsubscribe, e-mail: [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]

Reply via email to