perf: route final aggregate buckets by column by Rachelint · Pull Request #25959 · apache/datafusion

Rachelint · 2026-10-02T05:40:24Z

Which issue does this PR close?

Part of #20773. This experimental follow-up is based on #25724; it does not close the issue.

Rationale for this change

This POC combines column-wise bucket coalescing with a smaller Partial aggregation working set. Earlier ECS measurements found that a 2 MiB allocated-byte threshold reduced Partial lookup costs, while repeated string-buffer growth consumed part of that benefit. Reserving the next payload buffer from the previous flush reduced that overhead.

What changes are included in this PR?

Preserve the existing gather and gathered-range StringView scatter kernel from Arrow fork commit cd9284f884d05e329acb293e1e95353ce555315e.
Restore 32 Final buckets after the separate 128/256-bucket experiments. Intermediate compaction remains disabled.
When experimental bucketed aggregation is enabled, flush eligible Partial tables after an input batch reaches either the existing group-count threshold or 2 MiB of allocated storage. This includes the hashmap, group keys, scratch storage and accumulator states; it is not a hard memory limit.
Keep flushing recurring groups and remove recurrence sampling for this POC policy. Continue counting emitted groups for the existing skip-partial probe. Nested aggregate states and soft-limit cases retain their existing exclusions.
On threshold flushes, reuse the single-StringView hashmap allocation and reserve one new payload buffer from the previous payload length. The u32 view-offset limit retains a rollover path. Emitted arrays own their buffers; memory-pressure output still releases allocations.
Keep profiling probes and ECS environment-variable switches out of the submitted implementation.

The intended tradeoff is more downstream rows for cheaper Partial aggregation. Earlier ECS measurements for Q33/Q34 showed approximately 41%/42% less cumulative Partial time, 30% more Partial output rows, and 7.8%/6.9% lower query latency. These are cross-round experimental measurements, not benchmark-bot results or a guarantee for other queries.

What is the testing strategy for this PR?

Validation for efd384f9f:

cargo fmt --all and cargo fmt --all --check passed.
cargo clippy --all-targets --all-features -- -D warnings passed on ECS.
cargo check -p datafusion-physical-plan passed.
cargo test -p datafusion-physical-expr-common --lib binary_view_map: 14 passed.
cargo test -p datafusion-physical-plan --lib aggregates::: 282 passed.
The ECS source hashes match the submitted files. Tests cover repeated allocation reuse, long/inline/null string output ownership, byte-threshold flushing, recurring-group results, and aggregate output/spill paths.
Validation is scoped to the POC's affected paths as requested; the full workspace extended test suite was not rerun.

The requested clickbench_partitioned comparison uses baseline cf6819d7ef43de4da3eda8ff75118fa51ddc6c01. Both sides use 32 Final buckets, the same gather/scatter kernel, no intermediate compaction, and DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD=262144. This isolates the new Partial policy from the previous Final fanout experiments.

Are there any user-facing changes?

Default behavior remains unchanged because experimental bucketed aggregation is disabled by default. When enabled, eligible Partial tables now flush by allocated bytes even if keys recur. This may increase downstream work; the full partitioned ClickBench run is intended to measure that tradeoff. This remains a POC with an Arrow fork dependency.

…, off by default)

…threshold while its groups do not recur

…ting buckets of unique groups

…f the threshold

…e its input holds them

Rachelint · 2026-10-02T05:43:51Z

run benchmark clickbench_partitioned

baseline:
ref: f9df89e
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

adriangbot · 2026-10-02T05:47:27Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5946287393-3018-vf2zl 6.12.94+ #1 SMP Wed Aug 19 07:47:20 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 codex/in-progress-column-buckets (b964422) to f9df89e diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "f9df89e2b1077deaae9be106682c9f1e4d0d61d2"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-10-02T06:02:36Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing codex/in-progress-column-buckets (b964422) to f9df89e diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "f9df89e2b1077deaae9be106682c9f1e4d0d61d2"

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 codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │    1.38 ms │                          1.36 ms │     no change │
│ QQuery 1  │   12.29 ms │                         12.30 ms │     no change │
│ QQuery 2  │   38.01 ms │                         36.56 ms │     no change │
│ QQuery 3  │   33.08 ms │                         31.66 ms │     no change │
│ QQuery 4  │  194.67 ms │                        193.91 ms │     no change │
│ QQuery 5  │  298.01 ms │                        298.51 ms │     no change │
│ QQuery 6  │    1.31 ms │                          1.35 ms │     no change │
│ QQuery 7  │   13.30 ms │                         13.50 ms │     no change │
│ QQuery 8  │  295.90 ms │                        299.70 ms │     no change │
│ QQuery 9  │  516.57 ms │                        517.07 ms │     no change │
│ QQuery 10 │   69.82 ms │                         69.32 ms │     no change │
│ QQuery 11 │   81.46 ms │                         80.23 ms │     no change │
│ QQuery 12 │  281.54 ms │                        305.99 ms │  1.09x slower │
│ QQuery 13 │  368.27 ms │                        385.62 ms │     no change │
│ QQuery 14 │  306.21 ms │                        310.13 ms │     no change │
│ QQuery 15 │  222.74 ms │                        225.81 ms │     no change │
│ QQuery 16 │  660.43 ms │                        682.68 ms │     no change │
│ QQuery 17 │  593.47 ms │                        624.23 ms │  1.05x slower │
│ QQuery 18 │ 1233.76 ms │                       1208.48 ms │     no change │
│ QQuery 19 │   28.77 ms │                         28.43 ms │     no change │
│ QQuery 20 │  531.75 ms │                        525.29 ms │     no change │
│ QQuery 21 │  531.71 ms │                        539.23 ms │     no change │
│ QQuery 22 │ 1012.15 ms │                       1058.72 ms │     no change │
│ QQuery 23 │ 3250.78 ms │                       3103.50 ms │     no change │
│ QQuery 24 │   41.26 ms │                         41.00 ms │     no change │
│ QQuery 25 │  107.96 ms │                        105.95 ms │     no change │
│ QQuery 26 │   41.99 ms │                         42.66 ms │     no change │
│ QQuery 27 │  524.42 ms │                        515.76 ms │     no change │
│ QQuery 28 │ 3090.43 ms │                       2999.42 ms │     no change │
│ QQuery 29 │   43.42 ms │                         41.92 ms │     no change │
│ QQuery 30 │  298.27 ms │                        296.34 ms │     no change │
│ QQuery 31 │  263.19 ms │                        256.64 ms │     no change │
│ QQuery 32 │  781.12 ms │                        729.83 ms │ +1.07x faster │
│ QQuery 33 │ 1599.50 ms │                       1526.51 ms │     no change │
│ QQuery 34 │ 1587.52 ms │                       1512.00 ms │     no change │
│ QQuery 35 │  253.66 ms │                        234.93 ms │ +1.08x faster │
│ QQuery 36 │   73.04 ms │                         65.56 ms │ +1.11x faster │
│ QQuery 37 │   37.74 ms │                         35.75 ms │ +1.06x faster │
│ QQuery 38 │   41.68 ms │                         42.40 ms │     no change │
│ QQuery 39 │  157.71 ms │                        152.13 ms │     no change │
│ QQuery 40 │   16.31 ms │                         15.65 ms │     no change │
│ QQuery 41 │   15.05 ms │                         14.53 ms │     no change │
│ QQuery 42 │   14.22 ms │                         14.31 ms │     no change │
└───────────┴────────────┴──────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 19565.87ms │
│ Total Time (codex_in-progress-column-buckets)   │ 19196.85ms │
│ Average Time (HEAD)                             │   455.02ms │
│ Average Time (codex_in-progress-column-buckets) │   446.44ms │
│ Queries Faster                                  │          4 │
│ Queries Slower                                  │          2 │
│ Queries with No Change                          │         37 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Distribution per query (min / mean ±stddev / max):

Comparing HEAD and codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃       codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │           1.38 / 4.42 ±5.93 / 16.27 ms │           1.36 / 4.43 ±5.86 / 16.15 ms │     no change │
│ QQuery 1  │         12.29 / 12.73 ±0.45 / 13.55 ms │         12.30 / 12.72 ±0.29 / 13.12 ms │     no change │
│ QQuery 2  │         38.01 / 38.41 ±0.29 / 38.79 ms │         36.56 / 36.75 ±0.17 / 36.99 ms │     no change │
│ QQuery 3  │         33.08 / 34.22 ±0.72 / 35.09 ms │         31.66 / 32.18 ±0.33 / 32.48 ms │ +1.06x faster │
│ QQuery 4  │      194.67 / 198.62 ±2.38 / 201.86 ms │      193.91 / 195.55 ±2.02 / 199.33 ms │     no change │
│ QQuery 5  │      298.01 / 307.02 ±7.18 / 317.64 ms │      298.51 / 307.29 ±4.93 / 311.73 ms │     no change │
│ QQuery 6  │            1.31 / 1.53 ±0.23 / 1.93 ms │            1.35 / 1.49 ±0.22 / 1.92 ms │     no change │
│ QQuery 7  │         13.30 / 13.41 ±0.10 / 13.59 ms │         13.50 / 13.82 ±0.17 / 13.97 ms │     no change │
│ QQuery 8  │      295.90 / 301.89 ±3.92 / 305.49 ms │      299.70 / 306.42 ±4.99 / 314.65 ms │     no change │
│ QQuery 9  │     516.57 / 530.53 ±16.68 / 560.56 ms │     517.07 / 525.67 ±11.10 / 547.09 ms │     no change │
│ QQuery 10 │         69.82 / 72.46 ±1.73 / 74.28 ms │         69.32 / 70.48 ±1.62 / 73.69 ms │     no change │
│ QQuery 11 │         81.46 / 82.95 ±0.80 / 83.82 ms │        80.23 / 85.90 ±7.56 / 100.84 ms │     no change │
│ QQuery 12 │      281.54 / 289.40 ±7.70 / 302.37 ms │      305.99 / 317.98 ±7.97 / 327.76 ms │  1.10x slower │
│ QQuery 13 │      368.27 / 379.99 ±8.17 / 392.51 ms │     385.62 / 396.45 ±13.25 / 422.47 ms │     no change │
│ QQuery 14 │     306.21 / 317.88 ±11.70 / 339.27 ms │     310.13 / 331.54 ±18.30 / 353.82 ms │     no change │
│ QQuery 15 │     222.74 / 238.89 ±18.62 / 275.05 ms │     225.81 / 266.08 ±33.40 / 319.69 ms │  1.11x slower │
│ QQuery 16 │     660.43 / 695.60 ±25.18 / 729.92 ms │      682.68 / 690.25 ±4.61 / 696.07 ms │     no change │
│ QQuery 17 │     593.47 / 610.87 ±16.09 / 636.42 ms │     624.23 / 643.89 ±19.38 / 674.65 ms │  1.05x slower │
│ QQuery 18 │  1233.76 / 1286.18 ±53.81 / 1386.70 ms │  1208.48 / 1285.91 ±52.32 / 1364.57 ms │     no change │
│ QQuery 19 │         28.77 / 31.32 ±3.72 / 38.67 ms │        28.43 / 35.43 ±11.75 / 58.80 ms │  1.13x slower │
│ QQuery 20 │     531.75 / 546.77 ±17.45 / 573.52 ms │     525.29 / 537.63 ±10.74 / 557.18 ms │     no change │
│ QQuery 21 │     531.71 / 546.69 ±22.43 / 590.69 ms │      539.23 / 549.41 ±9.02 / 565.75 ms │     no change │
│ QQuery 22 │  1012.15 / 1036.68 ±16.38 / 1060.92 ms │  1058.72 / 1079.40 ±17.77 / 1109.89 ms │     no change │
│ QQuery 23 │ 3250.78 / 3538.98 ±243.67 / 3934.40 ms │ 3103.50 / 3555.05 ±374.43 / 4187.90 ms │     no change │
│ QQuery 24 │         41.26 / 43.99 ±2.78 / 47.54 ms │       41.00 / 60.00 ±30.03 / 119.68 ms │  1.36x slower │
│ QQuery 25 │     107.96 / 169.66 ±91.38 / 345.71 ms │      105.95 / 106.52 ±0.34 / 106.90 ms │ +1.59x faster │
│ QQuery 26 │         41.99 / 43.37 ±0.92 / 44.38 ms │        42.66 / 58.85 ±16.50 / 79.02 ms │  1.36x slower │
│ QQuery 27 │      524.42 / 530.22 ±5.78 / 540.97 ms │      515.76 / 523.38 ±6.10 / 534.48 ms │     no change │
│ QQuery 28 │  3090.43 / 3123.11 ±33.19 / 3180.03 ms │  2999.42 / 3026.05 ±30.55 / 3079.91 ms │     no change │
│ QQuery 29 │        43.42 / 52.86 ±11.55 / 73.61 ms │       41.92 / 85.42 ±33.44 / 141.98 ms │  1.62x slower │
│ QQuery 30 │     298.27 / 319.70 ±25.68 / 367.63 ms │      296.34 / 302.25 ±6.54 / 314.52 ms │ +1.06x faster │
│ QQuery 31 │      263.19 / 270.64 ±6.95 / 280.52 ms │      256.64 / 262.76 ±3.96 / 267.86 ms │     no change │
│ QQuery 32 │    781.12 / 857.94 ±87.22 / 1028.41 ms │     729.83 / 814.04 ±76.45 / 929.81 ms │ +1.05x faster │
│ QQuery 33 │  1599.50 / 1702.38 ±76.11 / 1826.81 ms │  1526.51 / 1618.77 ±48.75 / 1662.71 ms │     no change │
│ QQuery 34 │  1587.52 / 1690.60 ±59.36 / 1740.36 ms │ 1512.00 / 1631.58 ±103.59 / 1819.34 ms │     no change │
│ QQuery 35 │     253.66 / 292.37 ±48.57 / 382.80 ms │     234.93 / 272.98 ±41.42 / 345.61 ms │ +1.07x faster │
│ QQuery 36 │         73.04 / 77.99 ±2.57 / 80.51 ms │         65.56 / 72.73 ±5.67 / 79.57 ms │ +1.07x faster │
│ QQuery 37 │         37.74 / 41.02 ±3.89 / 47.88 ms │         35.75 / 38.03 ±2.72 / 43.18 ms │ +1.08x faster │
│ QQuery 38 │         41.68 / 47.10 ±4.59 / 53.84 ms │         42.40 / 44.50 ±1.48 / 46.91 ms │ +1.06x faster │
│ QQuery 39 │     157.71 / 170.12 ±12.08 / 191.62 ms │     152.13 / 171.07 ±13.00 / 186.01 ms │     no change │
│ QQuery 40 │         16.31 / 19.93 ±5.93 / 31.70 ms │         15.65 / 17.51 ±2.57 / 22.56 ms │ +1.14x faster │
│ QQuery 41 │         15.05 / 17.82 ±5.25 / 28.32 ms │         14.53 / 15.12 ±0.49 / 15.99 ms │ +1.18x faster │
│ QQuery 42 │         14.22 / 15.92 ±2.55 / 20.94 ms │         14.31 / 16.98 ±4.88 / 26.73 ms │  1.07x slower │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 20604.17ms │
│ Total Time (codex_in-progress-column-buckets)   │ 20420.26ms │
│ Average Time (HEAD)                             │   479.17ms │
│ Average Time (codex_in-progress-column-buckets) │   474.89ms │
│ Queries Faster                                  │         10 │
│ Queries Slower                                  │          8 │
│ Queries with No Change                          │         25 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	105.0s
Peak memory	17.7 GiB
Avg memory	5.5 GiB
CPU user	986.9s
CPU sys	97.8s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	105.0s
Peak memory	18.2 GiB
Avg memory	5.5 GiB
CPU user	975.6s
CPU sys	97.9s
Peak spill	0 B

File an issue against this benchmark runner

Rachelint · 2026-10-02T06:50:48Z

run benchmark clickbench_partitioned

baseline:
ref: f9df89e
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

adriangbot · 2026-10-02T06:55:15Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highcpu-48 (12 vCPU / 65 GiB) | Linux bench-c5946957588-3021-2czs2 6.12.94+ #1 SMP Wed Aug 19 07:47:20 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  48
On-line CPU(s) list:                     0-47
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     48
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:                               3 MiB (48 instances)
L1i cache:                               3 MiB (48 instances)
L2 cache:                                96 MiB (48 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-47
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 codex/in-progress-column-buckets (b964422) to f9df89e diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "f9df89e2b1077deaae9be106682c9f1e4d0d61d2"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-10-02T07:03:48Z

🤖 Benchmark completed (GKE) | trigger

Instance: c4a-highcpu-48 (12 vCPU / 65 GiB)

Comparing codex/in-progress-column-buckets (b964422) to f9df89e diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "f9df89e2b1077deaae9be106682c9f1e4d0d61d2"

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  48
On-line CPU(s) list:                     0-47
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     48
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:                               3 MiB (48 instances)
L1i cache:                               3 MiB (48 instances)
L2 cache:                                96 MiB (48 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-47
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 codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │    1.24 ms │                          1.25 ms │     no change │
│ QQuery 1  │   11.92 ms │                         11.73 ms │     no change │
│ QQuery 2  │   36.70 ms │                         35.28 ms │     no change │
│ QQuery 3  │   30.95 ms │                         29.88 ms │     no change │
│ QQuery 4  │  187.28 ms │                        183.18 ms │     no change │
│ QQuery 5  │  278.31 ms │                        276.88 ms │     no change │
│ QQuery 6  │    1.30 ms │                          1.29 ms │     no change │
│ QQuery 7  │   12.82 ms │                         12.84 ms │     no change │
│ QQuery 8  │  293.76 ms │                        291.95 ms │     no change │
│ QQuery 9  │  461.30 ms │                        466.04 ms │     no change │
│ QQuery 10 │   66.91 ms │                         66.54 ms │     no change │
│ QQuery 11 │   78.48 ms │                         78.10 ms │     no change │
│ QQuery 12 │  269.22 ms │                        267.24 ms │     no change │
│ QQuery 13 │  358.31 ms │                        353.87 ms │     no change │
│ QQuery 14 │  282.92 ms │                        279.27 ms │     no change │
│ QQuery 15 │  219.66 ms │                        214.20 ms │     no change │
│ QQuery 16 │  628.01 ms │                        652.37 ms │     no change │
│ QQuery 17 │  583.34 ms │                        576.94 ms │     no change │
│ QQuery 18 │ 1147.07 ms │                       1143.14 ms │     no change │
│ QQuery 19 │   27.47 ms │                         27.20 ms │     no change │
│ QQuery 20 │  527.85 ms │                        520.44 ms │     no change │
│ QQuery 21 │  522.99 ms │                        524.11 ms │     no change │
│ QQuery 22 │ 1024.18 ms │                       1015.46 ms │     no change │
│ QQuery 23 │ 3300.93 ms │                       3169.74 ms │     no change │
│ QQuery 24 │   40.21 ms │                         41.32 ms │     no change │
│ QQuery 25 │  105.06 ms │                        104.74 ms │     no change │
│ QQuery 26 │   41.01 ms │                         40.20 ms │     no change │
│ QQuery 27 │  514.18 ms │                        510.59 ms │     no change │
│ QQuery 28 │ 2983.64 ms │                       2934.84 ms │     no change │
│ QQuery 29 │   41.41 ms │                         40.47 ms │     no change │
│ QQuery 30 │  293.22 ms │                        288.76 ms │     no change │
│ QQuery 31 │  248.93 ms │                        246.45 ms │     no change │
│ QQuery 32 │  699.21 ms │                        688.17 ms │     no change │
│ QQuery 33 │ 1475.16 ms │                       1425.23 ms │     no change │
│ QQuery 34 │ 1444.86 ms │                       1502.72 ms │     no change │
│ QQuery 35 │  233.97 ms │                        237.69 ms │     no change │
│ QQuery 36 │   66.42 ms │                         69.11 ms │     no change │
│ QQuery 37 │   34.97 ms │                         35.43 ms │     no change │
│ QQuery 38 │   40.83 ms │                         42.85 ms │     no change │
│ QQuery 39 │  151.02 ms │                        143.27 ms │ +1.05x faster │
│ QQuery 40 │   14.30 ms │                         14.04 ms │     no change │
│ QQuery 41 │   13.64 ms │                         13.21 ms │     no change │
│ QQuery 42 │   13.02 ms │                         13.14 ms │     no change │
└───────────┴────────────┴──────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 18807.97ms │
│ Total Time (codex_in-progress-column-buckets)   │ 18591.18ms │
│ Average Time (HEAD)                             │   437.39ms │
│ Average Time (codex_in-progress-column-buckets) │   432.35ms │
│ Queries Faster                                  │          1 │
│ Queries Slower                                  │          0 │
│ Queries with No Change                          │         42 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Distribution per query (min / mean ±stddev / max):

Comparing HEAD and codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃      codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.24 / 3.93 ±5.33 / 14.59 ms │          1.25 / 4.01 ±5.46 / 14.93 ms │     no change │
│ QQuery 1  │        11.92 / 12.01 ±0.08 / 12.15 ms │        11.73 / 11.95 ±0.16 / 12.14 ms │     no change │
│ QQuery 2  │        36.70 / 36.81 ±0.12 / 37.03 ms │        35.28 / 35.74 ±0.23 / 35.96 ms │     no change │
│ QQuery 3  │        30.95 / 31.38 ±0.53 / 32.38 ms │        29.88 / 30.30 ±0.31 / 30.67 ms │     no change │
│ QQuery 4  │     187.28 / 189.75 ±1.69 / 191.81 ms │     183.18 / 186.22 ±1.80 / 188.07 ms │     no change │
│ QQuery 5  │     278.31 / 284.44 ±4.58 / 290.98 ms │     276.88 / 285.70 ±6.54 / 295.86 ms │     no change │
│ QQuery 6  │           1.30 / 1.44 ±0.22 / 1.87 ms │           1.29 / 1.42 ±0.22 / 1.85 ms │     no change │
│ QQuery 7  │        12.82 / 12.96 ±0.09 / 13.05 ms │        12.84 / 13.01 ±0.12 / 13.20 ms │     no change │
│ QQuery 8  │     293.76 / 296.25 ±2.64 / 300.83 ms │     291.95 / 297.15 ±6.18 / 308.73 ms │     no change │
│ QQuery 9  │     461.30 / 470.81 ±6.76 / 478.73 ms │     466.04 / 477.04 ±8.67 / 486.47 ms │     no change │
│ QQuery 10 │        66.91 / 67.63 ±0.47 / 68.28 ms │        66.54 / 67.33 ±1.24 / 69.79 ms │     no change │
│ QQuery 11 │        78.48 / 78.89 ±0.38 / 79.60 ms │        78.10 / 78.89 ±0.62 / 79.80 ms │     no change │
│ QQuery 12 │     269.22 / 274.83 ±5.27 / 284.12 ms │     267.24 / 275.27 ±8.63 / 290.92 ms │     no change │
│ QQuery 13 │     358.31 / 359.64 ±1.10 / 360.81 ms │    353.87 / 367.91 ±10.69 / 386.67 ms │     no change │
│ QQuery 14 │     282.92 / 288.34 ±6.12 / 299.42 ms │     279.27 / 285.83 ±7.41 / 299.67 ms │     no change │
│ QQuery 15 │    219.66 / 229.45 ±11.02 / 250.14 ms │     214.20 / 222.57 ±8.60 / 238.17 ms │     no change │
│ QQuery 16 │    628.01 / 646.21 ±12.22 / 661.53 ms │    652.37 / 668.35 ±12.96 / 683.07 ms │     no change │
│ QQuery 17 │    583.34 / 602.12 ±18.88 / 624.99 ms │    576.94 / 591.40 ±12.37 / 610.60 ms │     no change │
│ QQuery 18 │ 1147.07 / 1199.79 ±60.66 / 1317.87 ms │ 1143.14 / 1171.18 ±22.96 / 1201.39 ms │     no change │
│ QQuery 19 │      27.47 / 45.46 ±35.43 / 116.31 ms │       27.20 / 38.01 ±20.64 / 79.28 ms │ +1.20x faster │
│ QQuery 20 │    527.85 / 547.19 ±26.44 / 597.69 ms │     520.44 / 529.98 ±7.97 / 544.03 ms │     no change │
│ QQuery 21 │    522.99 / 537.87 ±16.13 / 568.63 ms │    524.11 / 535.58 ±11.22 / 553.76 ms │     no change │
│ QQuery 22 │ 1024.18 / 1071.65 ±49.50 / 1155.99 ms │ 1015.46 / 1056.69 ±34.67 / 1114.01 ms │     no change │
│ QQuery 23 │ 3300.93 / 3318.84 ±18.43 / 3351.85 ms │ 3169.74 / 3250.65 ±77.08 / 3388.30 ms │     no change │
│ QQuery 24 │       40.21 / 50.63 ±19.63 / 89.89 ms │     41.32 / 162.34 ±94.25 / 323.68 ms │  3.21x slower │
│ QQuery 25 │   105.06 / 210.23 ±207.00 / 624.21 ms │     104.74 / 105.76 ±1.19 / 107.41 ms │ +1.99x faster │
│ QQuery 26 │        41.01 / 42.15 ±1.46 / 44.97 ms │        40.20 / 41.27 ±0.87 / 42.54 ms │     no change │
│ QQuery 27 │     514.18 / 522.81 ±5.54 / 530.46 ms │     510.59 / 516.20 ±4.88 / 523.01 ms │     no change │
│ QQuery 28 │ 2983.64 / 3011.54 ±21.52 / 3042.01 ms │ 2934.84 / 2960.39 ±28.33 / 3011.38 ms │     no change │
│ QQuery 29 │        41.41 / 46.77 ±6.81 / 58.56 ms │       40.47 / 54.97 ±13.16 / 72.32 ms │  1.18x slower │
│ QQuery 30 │     293.22 / 300.65 ±8.51 / 313.22 ms │    288.76 / 309.30 ±21.02 / 336.90 ms │     no change │
│ QQuery 31 │    248.93 / 269.82 ±19.69 / 306.78 ms │     246.45 / 250.31 ±3.13 / 254.23 ms │ +1.08x faster │
│ QQuery 32 │   699.21 / 789.87 ±105.58 / 990.86 ms │    688.17 / 748.73 ±53.57 / 842.10 ms │ +1.05x faster │
│ QQuery 33 │ 1475.16 / 1560.20 ±62.51 / 1648.28 ms │ 1425.23 / 1520.41 ±62.30 / 1616.05 ms │     no change │
│ QQuery 34 │ 1444.86 / 1531.25 ±43.70 / 1564.72 ms │ 1502.72 / 1584.07 ±72.16 / 1691.89 ms │     no change │
│ QQuery 35 │    233.97 / 294.72 ±89.63 / 468.84 ms │    237.69 / 296.14 ±85.41 / 462.12 ms │     no change │
│ QQuery 36 │        66.42 / 72.21 ±3.38 / 75.91 ms │        69.11 / 74.81 ±5.32 / 84.31 ms │     no change │
│ QQuery 37 │        34.97 / 35.87 ±0.74 / 36.78 ms │        35.43 / 40.95 ±4.56 / 48.11 ms │  1.14x slower │
│ QQuery 38 │        40.83 / 46.27 ±4.46 / 53.48 ms │        42.85 / 43.98 ±1.20 / 45.80 ms │     no change │
│ QQuery 39 │     151.02 / 153.55 ±1.55 / 155.25 ms │     143.27 / 152.73 ±6.32 / 161.48 ms │     no change │
│ QQuery 40 │        14.30 / 14.60 ±0.29 / 15.00 ms │        14.04 / 17.48 ±6.25 / 29.96 ms │  1.20x slower │
│ QQuery 41 │        13.64 / 16.40 ±2.27 / 18.92 ms │        13.21 / 14.67 ±2.53 / 19.73 ms │ +1.12x faster │
│ QQuery 42 │        13.02 / 13.35 ±0.23 / 13.73 ms │        13.14 / 15.87 ±4.90 / 25.64 ms │  1.19x slower │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 19590.59ms │
│ Total Time (codex_in-progress-column-buckets)   │ 19392.56ms │
│ Average Time (HEAD)                             │   455.60ms │
│ Average Time (codex_in-progress-column-buckets) │   450.99ms │
│ Queries Faster                                  │          5 │
│ Queries Slower                                  │          5 │
│ Queries with No Change                          │         33 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	100.0s
Peak memory	18.4 GiB
Avg memory	5.6 GiB
CPU user	942.2s
CPU sys	90.6s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	100.0s
Peak memory	16.7 GiB
Avg memory	5.4 GiB
CPU user	936.3s
CPU sys	89.0s
Peak spill	0 B

File an issue against this benchmark runner

Rachelint · 2026-10-04T07:35:45Z

run benchmark clickbench_partitioned

baseline:
ref: f9df89e
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

adriangbot · 2026-10-04T07:38:28Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5977753017-3052-fr6m7 6.12.94+ #1 SMP Wed Aug 19 07:47:20 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 codex/in-progress-column-buckets (313a3a9) to f9df89e diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "f9df89e2b1077deaae9be106682c9f1e4d0d61d2"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-10-04T07:49:28Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing codex/in-progress-column-buckets (313a3a9) to f9df89e diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "f9df89e2b1077deaae9be106682c9f1e4d0d61d2"

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 codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │    1.27 ms │                          1.27 ms │     no change │
│ QQuery 1  │   11.56 ms │                         11.67 ms │     no change │
│ QQuery 2  │   36.86 ms │                         35.48 ms │     no change │
│ QQuery 3  │   31.12 ms │                         30.79 ms │     no change │
│ QQuery 4  │  188.35 ms │                        184.81 ms │     no change │
│ QQuery 5  │  276.59 ms │                        278.81 ms │     no change │
│ QQuery 6  │    1.29 ms │                          1.25 ms │     no change │
│ QQuery 7  │   12.74 ms │                         12.72 ms │     no change │
│ QQuery 8  │  289.56 ms │                        293.78 ms │     no change │
│ QQuery 9  │  466.96 ms │                        461.27 ms │     no change │
│ QQuery 10 │   68.31 ms │                         65.78 ms │     no change │
│ QQuery 11 │   79.69 ms │                         76.32 ms │     no change │
│ QQuery 12 │  273.37 ms │                        267.86 ms │     no change │
│ QQuery 13 │  350.94 ms │                        352.47 ms │     no change │
│ QQuery 14 │  281.91 ms │                        279.92 ms │     no change │
│ QQuery 15 │  215.38 ms │                        217.85 ms │     no change │
│ QQuery 16 │  631.50 ms │                        557.35 ms │ +1.13x faster │
│ QQuery 17 │  576.17 ms │                        496.78 ms │ +1.16x faster │
│ QQuery 18 │ 1171.34 ms │                       1071.75 ms │ +1.09x faster │
│ QQuery 19 │   27.40 ms │                         28.06 ms │     no change │
│ QQuery 20 │  535.65 ms │                        517.11 ms │     no change │
│ QQuery 21 │  515.19 ms │                        515.25 ms │     no change │
│ QQuery 22 │ 1054.58 ms │                       1010.87 ms │     no change │
│ QQuery 23 │ 3190.04 ms │                       3065.30 ms │     no change │
│ QQuery 24 │   40.33 ms │                         40.74 ms │     no change │
│ QQuery 25 │  105.35 ms │                        104.71 ms │     no change │
│ QQuery 26 │   40.97 ms │                         41.55 ms │     no change │
│ QQuery 27 │  507.41 ms │                        513.65 ms │     no change │
│ QQuery 28 │ 2936.95 ms │                       2902.47 ms │     no change │
│ QQuery 29 │   41.66 ms │                         41.33 ms │     no change │
│ QQuery 30 │  286.07 ms │                        282.64 ms │     no change │
│ QQuery 31 │  250.54 ms │                        247.86 ms │     no change │
│ QQuery 32 │  719.52 ms │                        632.56 ms │ +1.14x faster │
│ QQuery 33 │ 1454.78 ms │                       1441.06 ms │     no change │
│ QQuery 34 │ 1472.61 ms │                       1448.27 ms │     no change │
│ QQuery 35 │  225.63 ms │                        225.72 ms │     no change │
│ QQuery 36 │   65.19 ms │                         70.61 ms │  1.08x slower │
│ QQuery 37 │   34.56 ms │                         34.93 ms │     no change │
│ QQuery 38 │   41.24 ms │                         41.36 ms │     no change │
│ QQuery 39 │  145.83 ms │                        142.07 ms │     no change │
│ QQuery 40 │   14.91 ms │                         14.19 ms │     no change │
│ QQuery 41 │   13.33 ms │                         13.37 ms │     no change │
│ QQuery 42 │   12.86 ms │                         12.86 ms │     no change │
└───────────┴────────────┴──────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 18697.53ms │
│ Total Time (codex_in-progress-column-buckets)   │ 18086.49ms │
│ Average Time (HEAD)                             │   434.83ms │
│ Average Time (codex_in-progress-column-buckets) │   420.62ms │
│ Queries Faster                                  │          4 │
│ Queries Slower                                  │          1 │
│ Queries with No Change                          │         38 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Distribution per query (min / mean ±stddev / max):

Comparing HEAD and codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃       codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │           1.27 / 4.03 ±5.46 / 14.95 ms │           1.27 / 4.05 ±5.50 / 15.04 ms │     no change │
│ QQuery 1  │         11.56 / 11.80 ±0.22 / 12.18 ms │         11.67 / 11.85 ±0.12 / 11.97 ms │     no change │
│ QQuery 2  │         36.86 / 37.07 ±0.15 / 37.26 ms │         35.48 / 35.79 ±0.25 / 36.22 ms │     no change │
│ QQuery 3  │         31.12 / 32.41 ±1.16 / 33.99 ms │         30.79 / 31.47 ±0.89 / 33.20 ms │     no change │
│ QQuery 4  │      188.35 / 190.85 ±2.04 / 193.97 ms │      184.81 / 185.95 ±1.56 / 188.84 ms │     no change │
│ QQuery 5  │      276.59 / 282.43 ±7.34 / 296.83 ms │      278.81 / 283.47 ±3.42 / 287.10 ms │     no change │
│ QQuery 6  │            1.29 / 1.45 ±0.23 / 1.91 ms │            1.25 / 1.42 ±0.24 / 1.88 ms │     no change │
│ QQuery 7  │         12.74 / 12.91 ±0.13 / 13.05 ms │         12.72 / 12.84 ±0.08 / 12.94 ms │     no change │
│ QQuery 8  │      289.56 / 297.86 ±6.01 / 307.99 ms │      293.78 / 297.69 ±3.33 / 302.22 ms │     no change │
│ QQuery 9  │      466.96 / 476.09 ±9.34 / 490.75 ms │      461.27 / 474.75 ±8.21 / 483.64 ms │     no change │
│ QQuery 10 │         68.31 / 68.72 ±0.27 / 69.05 ms │         65.78 / 66.39 ±0.33 / 66.69 ms │     no change │
│ QQuery 11 │         79.69 / 80.61 ±1.39 / 83.34 ms │         76.32 / 77.32 ±0.81 / 78.49 ms │     no change │
│ QQuery 12 │      273.37 / 276.76 ±4.10 / 284.26 ms │      267.86 / 271.23 ±3.69 / 278.21 ms │     no change │
│ QQuery 13 │     350.94 / 361.25 ±10.56 / 378.77 ms │      352.47 / 358.47 ±5.22 / 368.01 ms │     no change │
│ QQuery 14 │      281.91 / 288.06 ±6.42 / 300.02 ms │      279.92 / 289.09 ±8.06 / 300.18 ms │     no change │
│ QQuery 15 │      215.38 / 219.99 ±2.72 / 222.24 ms │     217.85 / 233.70 ±22.51 / 277.97 ms │  1.06x slower │
│ QQuery 16 │      631.50 / 645.33 ±9.86 / 661.22 ms │     557.35 / 593.17 ±22.15 / 621.87 ms │ +1.09x faster │
│ QQuery 17 │     576.17 / 599.41 ±19.00 / 634.24 ms │      496.78 / 512.17 ±9.74 / 523.49 ms │ +1.17x faster │
│ QQuery 18 │  1171.34 / 1187.16 ±12.59 / 1209.86 ms │  1071.75 / 1120.21 ±57.22 / 1232.63 ms │ +1.06x faster │
│ QQuery 19 │        27.40 / 44.43 ±23.15 / 86.58 ms │       28.06 / 61.90 ±56.81 / 174.53 ms │  1.39x slower │
│ QQuery 20 │     535.65 / 553.59 ±20.24 / 587.04 ms │     517.11 / 530.32 ±10.44 / 546.08 ms │     no change │
│ QQuery 21 │     515.19 / 537.72 ±15.19 / 562.90 ms │     515.25 / 528.44 ±11.97 / 549.77 ms │     no change │
│ QQuery 22 │  1054.58 / 1113.49 ±46.87 / 1173.72 ms │  1010.87 / 1030.20 ±18.66 / 1064.92 ms │ +1.08x faster │
│ QQuery 23 │ 3190.04 / 3286.97 ±126.00 / 3532.17 ms │ 3065.30 / 3269.40 ±130.38 / 3443.54 ms │     no change │
│ QQuery 24 │      40.33 / 127.32 ±77.02 / 220.17 ms │       40.74 / 66.78 ±38.04 / 139.20 ms │ +1.91x faster │
│ QQuery 25 │     105.35 / 140.28 ±65.32 / 270.88 ms │    104.71 / 177.42 ±135.63 / 448.43 ms │  1.26x slower │
│ QQuery 26 │         40.97 / 41.99 ±1.09 / 43.99 ms │         41.55 / 42.21 ±0.73 / 43.25 ms │     no change │
│ QQuery 27 │     507.41 / 522.75 ±11.12 / 538.21 ms │     513.65 / 522.43 ±10.65 / 542.66 ms │     no change │
│ QQuery 28 │  2936.95 / 2970.15 ±23.49 / 2994.94 ms │  2902.47 / 2972.58 ±36.28 / 3006.83 ms │     no change │
│ QQuery 29 │        41.66 / 53.89 ±19.05 / 91.42 ms │         41.33 / 49.10 ±7.86 / 63.30 ms │ +1.10x faster │
│ QQuery 30 │     286.07 / 298.88 ±10.39 / 316.94 ms │      282.64 / 293.64 ±9.36 / 308.47 ms │     no change │
│ QQuery 31 │      250.54 / 257.42 ±3.66 / 261.20 ms │      247.86 / 256.32 ±7.13 / 266.00 ms │     no change │
│ QQuery 32 │     719.52 / 801.22 ±95.44 / 985.47 ms │     632.56 / 692.72 ±58.80 / 795.60 ms │ +1.16x faster │
│ QQuery 33 │ 1454.78 / 1602.07 ±153.21 / 1807.01 ms │  1441.06 / 1491.33 ±36.83 / 1551.64 ms │ +1.07x faster │
│ QQuery 34 │ 1472.61 / 1624.83 ±161.88 / 1846.73 ms │ 1448.27 / 1584.67 ±124.65 / 1813.34 ms │     no change │
│ QQuery 35 │    225.63 / 294.04 ±128.03 / 550.00 ms │     225.72 / 256.75 ±23.98 / 284.53 ms │ +1.15x faster │
│ QQuery 36 │         65.19 / 68.48 ±1.98 / 71.29 ms │         70.61 / 78.05 ±9.14 / 95.35 ms │  1.14x slower │
│ QQuery 37 │       34.56 / 48.41 ±27.11 / 102.63 ms │         34.93 / 37.13 ±3.69 / 44.48 ms │ +1.30x faster │
│ QQuery 38 │         41.24 / 46.47 ±3.52 / 51.89 ms │         41.36 / 46.05 ±3.62 / 52.39 ms │     no change │
│ QQuery 39 │      145.83 / 152.12 ±4.45 / 158.97 ms │      142.07 / 153.07 ±7.02 / 163.17 ms │     no change │
│ QQuery 40 │        14.91 / 29.08 ±24.69 / 78.29 ms │         14.19 / 18.96 ±7.55 / 33.81 ms │ +1.53x faster │
│ QQuery 41 │         13.33 / 14.99 ±2.85 / 20.68 ms │         13.37 / 13.52 ±0.12 / 13.69 ms │ +1.11x faster │
│ QQuery 42 │         12.86 / 14.02 ±1.84 / 17.64 ms │         12.86 / 13.10 ±0.15 / 13.31 ms │ +1.07x faster │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 19718.79ms │
│ Total Time (codex_in-progress-column-buckets)   │ 19047.10ms │
│ Average Time (HEAD)                             │   458.58ms │
│ Average Time (codex_in-progress-column-buckets) │   442.96ms │
│ Queries Faster                                  │         13 │
│ Queries Slower                                  │          4 │
│ Queries with No Change                          │         26 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	100.0s
Peak memory	17.8 GiB
Avg memory	5.5 GiB
CPU user	942.1s
CPU sys	90.2s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	100.0s
Peak memory	18.3 GiB
Avg memory	5.5 GiB
CPU user	922.4s
CPU sys	86.5s
Peak spill	0 B

File an issue against this benchmark runner

Rachelint · 2026-10-04T14:40:03Z

run benchmark clickbench_partitioned

baseline:
ref: 313a3a9
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

adriangbot · 2026-10-04T14:42:41Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5981162161-3053-bkmlp 6.12.94+ #1 SMP Wed Aug 19 07:47:20 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 codex/in-progress-column-buckets (085c221) to 313a3a9 diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "313a3a9590c30cbca9afd37c70ca4c8bdd49f10d"

Results will be posted here when complete

File an issue against this benchmark runner

Rachelint · 2026-10-04T14:54:30Z

run benchmark clickbench_partitioned

baseline:
ref: 313a3a9
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

adriangbot · 2026-10-04T14:57:54Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing codex/in-progress-column-buckets (085c221) to 313a3a9 diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "313a3a9590c30cbca9afd37c70ca4c8bdd49f10d"

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 codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ codex_in-progress-column-buckets ┃    Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ QQuery 0  │    1.25 ms │                          1.23 ms │ no change │
│ QQuery 1  │   11.92 ms │                         12.12 ms │ no change │
│ QQuery 2  │   36.08 ms │                         35.52 ms │ no change │
│ QQuery 3  │   31.09 ms │                         30.45 ms │ no change │
│ QQuery 4  │  189.65 ms │                        185.98 ms │ no change │
│ QQuery 5  │  282.12 ms │                        281.41 ms │ no change │
│ QQuery 6  │    1.33 ms │                          1.29 ms │ no change │
│ QQuery 7  │   13.01 ms │                         12.77 ms │ no change │
│ QQuery 8  │  293.19 ms │                        291.73 ms │ no change │
│ QQuery 9  │  474.87 ms │                        473.51 ms │ no change │
│ QQuery 10 │   66.79 ms │                         66.05 ms │ no change │
│ QQuery 11 │   77.91 ms │                         77.87 ms │ no change │
│ QQuery 12 │  268.28 ms │                        270.37 ms │ no change │
│ QQuery 13 │  355.94 ms │                        349.55 ms │ no change │
│ QQuery 14 │  279.74 ms │                        279.28 ms │ no change │
│ QQuery 15 │  214.91 ms │                        216.13 ms │ no change │
│ QQuery 16 │  559.53 ms │                        567.54 ms │ no change │
│ QQuery 17 │  508.07 ms │                        517.58 ms │ no change │
│ QQuery 18 │ 1088.77 ms │                       1101.65 ms │ no change │
│ QQuery 19 │   28.21 ms │                         28.12 ms │ no change │
│ QQuery 20 │  521.01 ms │                        525.25 ms │ no change │
│ QQuery 21 │  520.19 ms │                        533.74 ms │ no change │
│ QQuery 22 │ 1021.17 ms │                       1060.23 ms │ no change │
│ QQuery 23 │ 3211.29 ms │                       3168.57 ms │ no change │
│ QQuery 24 │   42.28 ms │                         40.77 ms │ no change │
│ QQuery 25 │  105.49 ms │                        106.27 ms │ no change │
│ QQuery 26 │   41.14 ms │                         41.37 ms │ no change │
│ QQuery 27 │  514.23 ms │                        513.63 ms │ no change │
│ QQuery 28 │ 2924.62 ms │                       2954.37 ms │ no change │
│ QQuery 29 │   40.80 ms │                         41.18 ms │ no change │
│ QQuery 30 │  289.15 ms │                        285.82 ms │ no change │
│ QQuery 31 │  245.56 ms │                        246.50 ms │ no change │
│ QQuery 32 │  656.37 ms │                        658.04 ms │ no change │
│ QQuery 33 │ 1484.30 ms │                       1530.46 ms │ no change │
│ QQuery 34 │ 1515.36 ms │                       1459.69 ms │ no change │
│ QQuery 35 │  233.06 ms │                        230.76 ms │ no change │
│ QQuery 36 │   66.21 ms │                         65.56 ms │ no change │
│ QQuery 37 │   35.78 ms │                         35.30 ms │ no change │
│ QQuery 38 │   42.76 ms │                         44.54 ms │ no change │
│ QQuery 39 │  149.56 ms │                        151.50 ms │ no change │
│ QQuery 40 │   14.57 ms │                         14.92 ms │ no change │
│ QQuery 41 │   14.00 ms │                         13.88 ms │ no change │
│ QQuery 42 │   13.29 ms │                         13.45 ms │ no change │
└───────────┴────────────┴──────────────────────────────────┴───────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 18484.85ms │
│ Total Time (codex_in-progress-column-buckets)   │ 18535.97ms │
│ Average Time (HEAD)                             │   429.88ms │
│ Average Time (codex_in-progress-column-buckets) │   431.07ms │
│ Queries Faster                                  │          0 │
│ Queries Slower                                  │          0 │
│ Queries with No Change                          │         43 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Distribution per query (min / mean ±stddev / max):

Comparing HEAD and codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃       codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │           1.25 / 4.09 ±5.57 / 15.23 ms │           1.23 / 4.03 ±5.53 / 15.09 ms │     no change │
│ QQuery 1  │         11.92 / 12.17 ±0.15 / 12.39 ms │         12.12 / 12.27 ±0.18 / 12.61 ms │     no change │
│ QQuery 2  │         36.08 / 36.46 ±0.20 / 36.64 ms │         35.52 / 35.74 ±0.23 / 36.13 ms │     no change │
│ QQuery 3  │         31.09 / 31.98 ±0.81 / 33.08 ms │         30.45 / 30.94 ±0.41 / 31.58 ms │     no change │
│ QQuery 4  │      189.65 / 190.59 ±1.18 / 192.93 ms │      185.98 / 189.50 ±4.88 / 199.10 ms │     no change │
│ QQuery 5  │      282.12 / 289.42 ±4.81 / 296.52 ms │      281.41 / 287.40 ±3.96 / 293.75 ms │     no change │
│ QQuery 6  │            1.33 / 1.48 ±0.25 / 1.97 ms │            1.29 / 1.43 ±0.23 / 1.89 ms │     no change │
│ QQuery 7  │         13.01 / 13.18 ±0.09 / 13.24 ms │         12.77 / 12.93 ±0.11 / 13.05 ms │     no change │
│ QQuery 8  │      293.19 / 298.90 ±3.68 / 302.83 ms │      291.73 / 301.51 ±7.65 / 311.69 ms │     no change │
│ QQuery 9  │      474.87 / 484.10 ±6.44 / 489.53 ms │      473.51 / 478.71 ±6.23 / 490.95 ms │     no change │
│ QQuery 10 │         66.79 / 67.84 ±0.82 / 69.30 ms │         66.05 / 66.70 ±0.41 / 67.14 ms │     no change │
│ QQuery 11 │         77.91 / 81.84 ±6.99 / 95.78 ms │         77.87 / 78.96 ±1.41 / 81.73 ms │     no change │
│ QQuery 12 │      268.28 / 276.28 ±5.96 / 285.24 ms │      270.37 / 275.13 ±7.09 / 289.25 ms │     no change │
│ QQuery 13 │      355.94 / 361.10 ±3.47 / 366.85 ms │      349.55 / 354.06 ±2.96 / 358.62 ms │     no change │
│ QQuery 14 │     279.74 / 289.81 ±10.80 / 306.63 ms │      279.28 / 284.71 ±7.34 / 298.75 ms │     no change │
│ QQuery 15 │      214.91 / 220.95 ±4.16 / 226.30 ms │      216.13 / 222.17 ±4.60 / 227.65 ms │     no change │
│ QQuery 16 │     559.53 / 579.40 ±11.17 / 592.41 ms │     567.54 / 589.30 ±17.01 / 616.71 ms │     no change │
│ QQuery 17 │      508.07 / 511.93 ±2.80 / 515.93 ms │     517.58 / 529.32 ±11.02 / 550.22 ms │     no change │
│ QQuery 18 │  1088.77 / 1120.16 ±28.83 / 1174.58 ms │ 1101.65 / 1188.85 ±128.42 / 1439.69 ms │  1.06x slower │
│ QQuery 19 │       28.21 / 51.72 ±33.21 / 113.37 ms │        28.12 / 37.14 ±17.45 / 72.04 ms │ +1.39x faster │
│ QQuery 20 │     521.01 / 538.64 ±14.43 / 556.12 ms │      525.25 / 532.22 ±7.06 / 541.59 ms │     no change │
│ QQuery 21 │      520.19 / 528.75 ±4.45 / 532.57 ms │      533.74 / 545.86 ±9.93 / 557.89 ms │     no change │
│ QQuery 22 │  1021.17 / 1041.27 ±12.07 / 1058.00 ms │  1060.23 / 1086.50 ±17.11 / 1106.06 ms │     no change │
│ QQuery 23 │ 3211.29 / 3317.22 ±112.77 / 3528.40 ms │ 3168.57 / 3313.25 ±114.36 / 3518.04 ms │     no change │
│ QQuery 24 │     42.28 / 210.13 ±240.78 / 661.78 ms │       40.77 / 74.46 ±57.29 / 188.16 ms │ +2.82x faster │
│ QQuery 25 │      105.49 / 106.93 ±1.73 / 110.21 ms │    106.27 / 199.24 ±148.79 / 491.99 ms │  1.86x slower │
│ QQuery 26 │         41.14 / 41.49 ±0.37 / 42.20 ms │         41.37 / 42.36 ±1.00 / 44.24 ms │     no change │
│ QQuery 27 │     514.23 / 532.96 ±17.86 / 565.78 ms │      513.63 / 526.61 ±8.98 / 537.02 ms │     no change │
│ QQuery 28 │  2924.62 / 2981.14 ±49.30 / 3060.91 ms │  2954.37 / 3013.03 ±31.71 / 3040.78 ms │     no change │
│ QQuery 29 │         40.80 / 47.76 ±5.97 / 55.27 ms │       41.18 / 64.45 ±34.69 / 130.78 ms │  1.35x slower │
│ QQuery 30 │     289.15 / 323.26 ±58.85 / 440.78 ms │     285.82 / 298.24 ±10.71 / 315.57 ms │ +1.08x faster │
│ QQuery 31 │     245.56 / 266.28 ±18.69 / 297.61 ms │     246.50 / 264.16 ±15.42 / 288.80 ms │     no change │
│ QQuery 32 │      656.37 / 664.30 ±8.90 / 680.00 ms │     658.04 / 703.34 ±44.18 / 782.32 ms │  1.06x slower │
│ QQuery 33 │  1484.30 / 1529.29 ±33.93 / 1577.15 ms │ 1530.46 / 1771.82 ±198.17 / 1976.03 ms │  1.16x slower │
│ QQuery 34 │  1515.36 / 1552.52 ±36.74 / 1622.90 ms │  1459.69 / 1496.79 ±45.97 / 1576.76 ms │     no change │
│ QQuery 35 │     233.06 / 287.77 ±87.72 / 461.42 ms │     230.76 / 294.46 ±74.06 / 434.83 ms │     no change │
│ QQuery 36 │         66.21 / 70.85 ±4.43 / 78.20 ms │         65.56 / 72.01 ±4.00 / 75.87 ms │     no change │
│ QQuery 37 │         35.78 / 42.57 ±9.12 / 59.20 ms │        35.30 / 41.13 ±11.22 / 63.57 ms │     no change │
│ QQuery 38 │         42.76 / 46.50 ±2.56 / 49.71 ms │         44.54 / 49.27 ±4.44 / 57.02 ms │  1.06x slower │
│ QQuery 39 │      149.56 / 153.91 ±3.78 / 159.09 ms │     151.50 / 161.15 ±11.11 / 181.94 ms │     no change │
│ QQuery 40 │         14.57 / 15.41 ±0.60 / 16.25 ms │         14.92 / 15.99 ±1.65 / 19.25 ms │     no change │
│ QQuery 41 │         14.00 / 14.12 ±0.11 / 14.27 ms │         13.88 / 16.06 ±3.86 / 23.75 ms │  1.14x slower │
│ QQuery 42 │         13.29 / 14.29 ±1.73 / 17.74 ms │         13.45 / 16.74 ±6.15 / 29.04 ms │  1.17x slower │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 19250.77ms │
│ Total Time (codex_in-progress-column-buckets)   │ 19579.96ms │
│ Average Time (HEAD)                             │   447.69ms │
│ Average Time (codex_in-progress-column-buckets) │   455.35ms │
│ Queries Faster                                  │          3 │
│ Queries Slower                                  │          8 │
│ Queries with No Change                          │         32 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	100.0s
Peak memory	18.3 GiB
Avg memory	5.7 GiB
CPU user	928.0s
CPU sys	89.1s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	100.0s
Peak memory	18.3 GiB
Avg memory	5.7 GiB
CPU user	930.9s
CPU sys	93.6s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-10-04T14:57:55Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5981287627-3054-rh9gl 6.12.94+ #1 SMP Wed Aug 19 07:47:20 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 codex/in-progress-column-buckets (cf6819d) to 313a3a9 diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "313a3a9590c30cbca9afd37c70ca4c8bdd49f10d"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-10-04T15:12:21Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing codex/in-progress-column-buckets (cf6819d) to 313a3a9 diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "313a3a9590c30cbca9afd37c70ca4c8bdd49f10d"

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 codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │    1.25 ms │                          1.25 ms │     no change │
│ QQuery 1  │   11.87 ms │                         11.71 ms │     no change │
│ QQuery 2  │   35.90 ms │                         35.65 ms │     no change │
│ QQuery 3  │   30.88 ms │                         30.36 ms │     no change │
│ QQuery 4  │  186.99 ms │                        184.78 ms │     no change │
│ QQuery 5  │  276.92 ms │                        279.25 ms │     no change │
│ QQuery 6  │    1.29 ms │                          1.29 ms │     no change │
│ QQuery 7  │   12.81 ms │                         12.90 ms │     no change │
│ QQuery 8  │  295.52 ms │                        292.65 ms │     no change │
│ QQuery 9  │  465.39 ms │                        460.59 ms │     no change │
│ QQuery 10 │   66.24 ms │                         66.05 ms │     no change │
│ QQuery 11 │   76.96 ms │                         76.27 ms │     no change │
│ QQuery 12 │  265.29 ms │                        269.20 ms │     no change │
│ QQuery 13 │  354.06 ms │                        346.71 ms │     no change │
│ QQuery 14 │  275.73 ms │                        274.42 ms │     no change │
│ QQuery 15 │  214.06 ms │                        219.02 ms │     no change │
│ QQuery 16 │  569.33 ms │                        561.70 ms │     no change │
│ QQuery 17 │  504.09 ms │                        511.00 ms │     no change │
│ QQuery 18 │ 1091.80 ms │                       1099.00 ms │     no change │
│ QQuery 19 │   27.90 ms │                         27.92 ms │     no change │
│ QQuery 20 │  521.90 ms │                        529.66 ms │     no change │
│ QQuery 21 │  517.75 ms │                        534.36 ms │     no change │
│ QQuery 22 │ 1024.57 ms │                       1074.75 ms │     no change │
│ QQuery 23 │ 3063.11 ms │                       3124.96 ms │     no change │
│ QQuery 24 │   41.32 ms │                         40.08 ms │     no change │
│ QQuery 25 │  105.52 ms │                        105.11 ms │     no change │
│ QQuery 26 │   41.00 ms │                         40.29 ms │     no change │
│ QQuery 27 │  504.72 ms │                        510.93 ms │     no change │
│ QQuery 28 │ 2945.04 ms │                       2924.49 ms │     no change │
│ QQuery 29 │   41.07 ms │                         40.61 ms │     no change │
│ QQuery 30 │  282.75 ms │                        280.85 ms │     no change │
│ QQuery 31 │  247.22 ms │                        247.70 ms │     no change │
│ QQuery 32 │  666.85 ms │                        629.12 ms │ +1.06x faster │
│ QQuery 33 │ 1430.79 ms │                       1481.03 ms │     no change │
│ QQuery 34 │ 1481.65 ms │                       1474.93 ms │     no change │
│ QQuery 35 │  223.60 ms │                        233.00 ms │     no change │
│ QQuery 36 │   66.11 ms │                         66.71 ms │     no change │
│ QQuery 37 │   35.15 ms │                         35.29 ms │     no change │
│ QQuery 38 │   43.21 ms │                         40.83 ms │ +1.06x faster │
│ QQuery 39 │  152.69 ms │                        139.99 ms │ +1.09x faster │
│ QQuery 40 │   14.19 ms │                         14.60 ms │     no change │
│ QQuery 41 │   13.57 ms │                         13.51 ms │     no change │
│ QQuery 42 │   13.04 ms │                         13.20 ms │     no change │
└───────────┴────────────┴──────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 18241.07ms │
│ Total Time (codex_in-progress-column-buckets)   │ 18357.71ms │
│ Average Time (HEAD)                             │   424.21ms │
│ Average Time (codex_in-progress-column-buckets) │   426.92ms │
│ Queries Faster                                  │          3 │
│ Queries Slower                                  │          0 │
│ Queries with No Change                          │         40 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Distribution per query (min / mean ±stddev / max):

Comparing HEAD and codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃       codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │           1.25 / 3.97 ±5.38 / 14.74 ms │           1.25 / 3.96 ±5.38 / 14.71 ms │     no change │
│ QQuery 1  │         11.87 / 11.99 ±0.11 / 12.15 ms │         11.71 / 11.90 ±0.12 / 12.07 ms │     no change │
│ QQuery 2  │         35.90 / 36.06 ±0.17 / 36.37 ms │         35.65 / 36.05 ±0.24 / 36.38 ms │     no change │
│ QQuery 3  │         30.88 / 31.73 ±0.69 / 32.69 ms │         30.36 / 30.91 ±0.42 / 31.65 ms │     no change │
│ QQuery 4  │      186.99 / 190.09 ±2.25 / 193.04 ms │      184.78 / 187.72 ±1.69 / 189.39 ms │     no change │
│ QQuery 5  │      276.92 / 282.99 ±3.54 / 286.97 ms │      279.25 / 282.96 ±2.80 / 286.14 ms │     no change │
│ QQuery 6  │            1.29 / 1.47 ±0.23 / 1.91 ms │            1.29 / 1.45 ±0.23 / 1.90 ms │     no change │
│ QQuery 7  │         12.81 / 12.94 ±0.13 / 13.18 ms │         12.90 / 13.01 ±0.11 / 13.20 ms │     no change │
│ QQuery 8  │      295.52 / 297.19 ±1.64 / 299.96 ms │      292.65 / 297.12 ±4.60 / 305.92 ms │     no change │
│ QQuery 9  │      465.39 / 470.17 ±5.23 / 480.13 ms │      460.59 / 472.36 ±7.19 / 482.60 ms │     no change │
│ QQuery 10 │         66.24 / 66.51 ±0.22 / 66.86 ms │         66.05 / 68.89 ±4.18 / 77.02 ms │     no change │
│ QQuery 11 │         76.96 / 77.44 ±0.47 / 78.04 ms │         76.27 / 77.49 ±0.89 / 78.50 ms │     no change │
│ QQuery 12 │      265.29 / 274.27 ±7.59 / 284.64 ms │      269.20 / 275.18 ±6.15 / 284.82 ms │     no change │
│ QQuery 13 │      354.06 / 358.49 ±4.20 / 363.66 ms │     346.71 / 359.15 ±13.20 / 383.18 ms │     no change │
│ QQuery 14 │      275.73 / 283.21 ±6.05 / 290.99 ms │      274.42 / 283.74 ±8.67 / 299.93 ms │     no change │
│ QQuery 15 │      214.06 / 218.68 ±3.72 / 224.58 ms │     219.02 / 235.08 ±14.67 / 256.77 ms │  1.08x slower │
│ QQuery 16 │     569.33 / 581.17 ±17.98 / 616.58 ms │     561.70 / 585.16 ±14.78 / 605.78 ms │     no change │
│ QQuery 17 │      504.09 / 507.74 ±3.53 / 514.33 ms │      511.00 / 517.36 ±6.60 / 529.85 ms │     no change │
│ QQuery 18 │  1091.80 / 1114.69 ±19.66 / 1151.18 ms │  1099.00 / 1173.76 ±92.19 / 1346.02 ms │  1.05x slower │
│ QQuery 19 │        27.90 / 36.51 ±12.65 / 60.75 ms │        27.92 / 48.50 ±25.07 / 86.03 ms │  1.33x slower │
│ QQuery 20 │     521.90 / 534.08 ±11.38 / 549.59 ms │      529.66 / 538.67 ±9.37 / 551.98 ms │     no change │
│ QQuery 21 │      517.75 / 530.32 ±9.38 / 544.39 ms │     534.36 / 544.72 ±14.22 / 572.85 ms │     no change │
│ QQuery 22 │  1024.57 / 1047.05 ±24.15 / 1090.00 ms │  1074.75 / 1124.79 ±46.90 / 1182.78 ms │  1.07x slower │
│ QQuery 23 │ 3063.11 / 3273.09 ±141.00 / 3482.60 ms │ 3124.96 / 3346.28 ±254.08 / 3790.26 ms │     no change │
│ QQuery 24 │       41.32 / 86.86 ±57.64 / 180.73 ms │         40.08 / 40.90 ±0.57 / 41.83 ms │ +2.12x faster │
│ QQuery 25 │    105.52 / 208.99 ±200.69 / 610.29 ms │    105.11 / 195.94 ±178.43 / 552.78 ms │ +1.07x faster │
│ QQuery 26 │         41.00 / 41.30 ±0.18 / 41.54 ms │         40.29 / 41.99 ±0.98 / 43.25 ms │     no change │
│ QQuery 27 │      504.72 / 515.46 ±9.79 / 532.07 ms │     510.93 / 533.86 ±21.16 / 573.03 ms │     no change │
│ QQuery 28 │  2945.04 / 2983.41 ±32.78 / 3042.68 ms │  2924.49 / 2943.31 ±16.97 / 2971.56 ms │     no change │
│ QQuery 29 │       41.07 / 73.90 ±34.61 / 120.02 ms │         40.61 / 41.14 ±0.45 / 41.65 ms │ +1.80x faster │
│ QQuery 30 │      282.75 / 293.25 ±6.91 / 301.48 ms │     280.85 / 293.78 ±12.01 / 315.45 ms │     no change │
│ QQuery 31 │      247.22 / 255.89 ±5.26 / 261.64 ms │     247.70 / 258.30 ±13.98 / 285.16 ms │     no change │
│ QQuery 32 │     666.85 / 676.44 ±11.06 / 697.56 ms │   629.12 / 738.98 ±148.78 / 1032.26 ms │  1.09x slower │
│ QQuery 33 │  1430.79 / 1549.10 ±84.96 / 1655.62 ms │   1481.03 / 1492.88 ±9.15 / 1505.46 ms │     no change │
│ QQuery 34 │  1481.65 / 1510.28 ±29.23 / 1561.71 ms │ 1474.93 / 1686.36 ±228.70 / 1983.31 ms │  1.12x slower │
│ QQuery 35 │     223.60 / 247.52 ±33.33 / 312.63 ms │      233.00 / 234.24 ±0.84 / 235.46 ms │ +1.06x faster │
│ QQuery 36 │         66.11 / 71.23 ±4.35 / 77.18 ms │         66.71 / 69.08 ±1.42 / 70.49 ms │     no change │
│ QQuery 37 │         35.15 / 38.14 ±4.74 / 47.53 ms │         35.29 / 37.80 ±4.09 / 45.91 ms │     no change │
│ QQuery 38 │         43.21 / 46.16 ±2.33 / 49.95 ms │         40.83 / 44.01 ±2.97 / 49.01 ms │     no change │
│ QQuery 39 │      152.69 / 155.96 ±3.08 / 160.99 ms │     139.99 / 158.87 ±11.57 / 176.49 ms │     no change │
│ QQuery 40 │         14.19 / 16.02 ±1.88 / 19.22 ms │         14.60 / 14.78 ±0.20 / 15.06 ms │ +1.08x faster │
│ QQuery 41 │         13.57 / 15.40 ±3.40 / 22.21 ms │         13.51 / 13.80 ±0.19 / 14.03 ms │ +1.12x faster │
│ QQuery 42 │         13.04 / 13.23 ±0.11 / 13.38 ms │         13.20 / 14.70 ±2.63 / 19.94 ms │  1.11x slower │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 19040.40ms │
│ Total Time (codex_in-progress-column-buckets)   │ 19370.91ms │
│ Average Time (HEAD)                             │   442.80ms │
│ Average Time (codex_in-progress-column-buckets) │   450.49ms │
│ Queries Faster                                  │          6 │
│ Queries Slower                                  │          7 │
│ Queries with No Change                          │         30 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	100.0s
Peak memory	18.2 GiB
Avg memory	5.6 GiB
CPU user	924.3s
CPU sys	85.0s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	100.0s
Peak memory	18.5 GiB
Avg memory	5.6 GiB
CPU user	922.7s
CPU sys	90.9s
Peak spill	0 B

File an issue against this benchmark runner

Rachelint · 2026-10-06T03:59:18Z

run benchmark clickbench_partitioned

baseline:
ref: cf6819d
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

adriangbot · 2026-10-06T04:04:45Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highcpu-48 (12 vCPU / 65 GiB) | Linux bench-c6009043623-3086-tr25s 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  48
On-line CPU(s) list:                     0-47
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     48
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:                               3 MiB (48 instances)
L1i cache:                               3 MiB (48 instances)
L2 cache:                                96 MiB (48 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-47
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 codex/in-progress-column-buckets (33aad7a) to cf6819d diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "cf6819d7ef43de4da3eda8ff75118fa51ddc6c01"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-10-06T04:19:42Z

🤖 Benchmark completed (GKE) | trigger

Instance: c4a-highcpu-48 (12 vCPU / 65 GiB)

Comparing codex/in-progress-column-buckets (33aad7a) to cf6819d diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "cf6819d7ef43de4da3eda8ff75118fa51ddc6c01"

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  48
On-line CPU(s) list:                     0-47
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     48
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:                               3 MiB (48 instances)
L1i cache:                               3 MiB (48 instances)
L2 cache:                                96 MiB (48 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-47
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 codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │    1.24 ms │                          1.22 ms │     no change │
│ QQuery 1  │   11.61 ms │                         11.70 ms │     no change │
│ QQuery 2  │   35.20 ms │                         35.22 ms │     no change │
│ QQuery 3  │   30.06 ms │                         29.79 ms │     no change │
│ QQuery 4  │  188.44 ms │                        185.80 ms │     no change │
│ QQuery 5  │  278.12 ms │                        278.73 ms │     no change │
│ QQuery 6  │    1.27 ms │                          1.28 ms │     no change │
│ QQuery 7  │   12.73 ms │                         12.68 ms │     no change │
│ QQuery 8  │  292.53 ms │                        293.54 ms │     no change │
│ QQuery 9  │  459.25 ms │                        461.85 ms │     no change │
│ QQuery 10 │   65.74 ms │                         65.17 ms │     no change │
│ QQuery 11 │   76.19 ms │                         76.54 ms │     no change │
│ QQuery 12 │  265.77 ms │                        266.88 ms │     no change │
│ QQuery 13 │  354.95 ms │                        348.24 ms │     no change │
│ QQuery 14 │  279.42 ms │                        273.14 ms │     no change │
│ QQuery 15 │  218.08 ms │                        217.58 ms │     no change │
│ QQuery 16 │  564.88 ms │                        569.30 ms │     no change │
│ QQuery 17 │  514.62 ms │                        510.69 ms │     no change │
│ QQuery 18 │ 1093.51 ms │                       1074.87 ms │     no change │
│ QQuery 19 │   27.51 ms │                         27.58 ms │     no change │
│ QQuery 20 │  525.61 ms │                        518.95 ms │     no change │
│ QQuery 21 │  517.62 ms │                        512.52 ms │     no change │
│ QQuery 22 │ 1028.74 ms │                       1006.66 ms │     no change │
│ QQuery 23 │ 3120.30 ms │                       3011.57 ms │     no change │
│ QQuery 24 │   40.32 ms │                         40.00 ms │     no change │
│ QQuery 25 │  104.56 ms │                        103.24 ms │     no change │
│ QQuery 26 │   40.91 ms │                         40.43 ms │     no change │
│ QQuery 27 │  519.12 ms │                        507.39 ms │     no change │
│ QQuery 28 │ 2920.54 ms │                       2905.70 ms │     no change │
│ QQuery 29 │   40.53 ms │                         40.73 ms │     no change │
│ QQuery 30 │  283.71 ms │                        294.04 ms │     no change │
│ QQuery 31 │  241.13 ms │                        257.47 ms │  1.07x slower │
│ QQuery 32 │  619.79 ms │                        680.68 ms │  1.10x slower │
│ QQuery 33 │ 1450.27 ms │                       1484.10 ms │     no change │
│ QQuery 34 │ 1445.99 ms │                       1544.05 ms │  1.07x slower │
│ QQuery 35 │  224.09 ms │                        225.96 ms │     no change │
│ QQuery 36 │   65.53 ms │                         69.60 ms │  1.06x slower │
│ QQuery 37 │   35.08 ms │                         34.45 ms │     no change │
│ QQuery 38 │   45.66 ms │                         41.03 ms │ +1.11x faster │
│ QQuery 39 │  145.70 ms │                        145.81 ms │     no change │
│ QQuery 40 │   14.27 ms │                         14.32 ms │     no change │
│ QQuery 41 │   13.24 ms │                         13.37 ms │     no change │
│ QQuery 42 │   13.08 ms │                         13.06 ms │     no change │
└───────────┴────────────┴──────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 18226.91ms │
│ Total Time (codex_in-progress-column-buckets)   │ 18246.97ms │
│ Average Time (HEAD)                             │   423.88ms │
│ Average Time (codex_in-progress-column-buckets) │   424.35ms │
│ Queries Faster                                  │          1 │
│ Queries Slower                                  │          4 │
│ Queries with No Change                          │         38 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Distribution per query (min / mean ±stddev / max):

Comparing HEAD and codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃       codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │           1.24 / 3.95 ±5.37 / 14.70 ms │           1.22 / 3.91 ±5.30 / 14.51 ms │     no change │
│ QQuery 1  │         11.61 / 11.81 ±0.15 / 11.98 ms │         11.70 / 11.88 ±0.11 / 12.06 ms │     no change │
│ QQuery 2  │         35.20 / 35.45 ±0.18 / 35.66 ms │         35.22 / 35.62 ±0.25 / 35.95 ms │     no change │
│ QQuery 3  │         30.06 / 30.58 ±0.58 / 31.67 ms │         29.79 / 30.07 ±0.16 / 30.30 ms │     no change │
│ QQuery 4  │      188.44 / 189.94 ±1.63 / 192.16 ms │      185.80 / 188.69 ±2.09 / 192.11 ms │     no change │
│ QQuery 5  │      278.12 / 282.60 ±5.64 / 293.34 ms │      278.73 / 283.63 ±5.18 / 293.69 ms │     no change │
│ QQuery 6  │            1.27 / 1.43 ±0.23 / 1.88 ms │            1.28 / 1.40 ±0.19 / 1.79 ms │     no change │
│ QQuery 7  │         12.73 / 12.91 ±0.17 / 13.23 ms │         12.68 / 12.79 ±0.08 / 12.93 ms │     no change │
│ QQuery 8  │      292.53 / 294.88 ±2.32 / 299.06 ms │      293.54 / 302.35 ±6.33 / 312.95 ms │     no change │
│ QQuery 9  │     459.25 / 477.24 ±16.97 / 509.19 ms │      461.85 / 474.77 ±7.77 / 485.17 ms │     no change │
│ QQuery 10 │         65.74 / 66.28 ±0.42 / 66.76 ms │         65.17 / 65.96 ±0.49 / 66.58 ms │     no change │
│ QQuery 11 │         76.19 / 77.06 ±0.95 / 78.89 ms │       76.54 / 82.81 ±10.10 / 102.95 ms │  1.07x slower │
│ QQuery 12 │      265.77 / 279.50 ±9.37 / 291.37 ms │     266.88 / 277.32 ±10.99 / 298.43 ms │     no change │
│ QQuery 13 │      354.95 / 364.68 ±8.55 / 380.10 ms │     348.24 / 364.41 ±12.62 / 385.97 ms │     no change │
│ QQuery 14 │      279.42 / 288.66 ±7.08 / 294.87 ms │      273.14 / 279.07 ±6.79 / 291.89 ms │     no change │
│ QQuery 15 │     218.08 / 227.69 ±10.70 / 244.20 ms │      217.58 / 221.08 ±4.94 / 230.84 ms │     no change │
│ QQuery 16 │     564.88 / 595.72 ±19.48 / 624.23 ms │      569.30 / 578.69 ±5.54 / 584.49 ms │     no change │
│ QQuery 17 │      514.62 / 519.21 ±3.26 / 524.64 ms │      510.69 / 520.34 ±9.86 / 536.59 ms │     no change │
│ QQuery 18 │  1093.51 / 1155.09 ±66.46 / 1274.71 ms │   1074.87 / 1087.75 ±8.00 / 1097.99 ms │ +1.06x faster │
│ QQuery 19 │        27.51 / 48.90 ±26.85 / 91.63 ms │         27.58 / 32.72 ±4.19 / 36.74 ms │ +1.49x faster │
│ QQuery 20 │     525.61 / 536.33 ±12.45 / 555.08 ms │      518.95 / 526.38 ±6.45 / 535.50 ms │     no change │
│ QQuery 21 │     517.62 / 533.01 ±10.26 / 547.10 ms │     512.52 / 527.69 ±13.08 / 543.08 ms │     no change │
│ QQuery 22 │  1028.74 / 1160.07 ±89.12 / 1269.70 ms │  1006.66 / 1025.46 ±13.88 / 1049.05 ms │ +1.13x faster │
│ QQuery 23 │ 3120.30 / 3273.52 ±117.09 / 3471.57 ms │ 3011.57 / 3191.38 ±175.95 / 3528.01 ms │     no change │
│ QQuery 24 │       40.32 / 98.50 ±57.56 / 167.94 ms │     40.00 / 126.55 ±131.26 / 379.55 ms │  1.28x slower │
│ QQuery 25 │    104.56 / 168.06 ±109.42 / 385.92 ms │      103.24 / 107.11 ±4.33 / 114.47 ms │ +1.57x faster │
│ QQuery 26 │         40.91 / 42.78 ±1.83 / 46.20 ms │         40.43 / 40.95 ±0.75 / 42.43 ms │     no change │
│ QQuery 27 │      519.12 / 524.28 ±4.51 / 532.33 ms │      507.39 / 520.81 ±8.87 / 532.52 ms │     no change │
│ QQuery 28 │  2920.54 / 2950.65 ±36.41 / 3020.60 ms │  2905.70 / 2929.40 ±19.41 / 2957.41 ms │     no change │
│ QQuery 29 │         40.53 / 40.81 ±0.25 / 41.17 ms │         40.73 / 45.65 ±9.29 / 64.23 ms │  1.12x slower │
│ QQuery 30 │      283.71 / 290.17 ±7.51 / 304.30 ms │     294.04 / 308.99 ±19.57 / 347.23 ms │  1.06x slower │
│ QQuery 31 │      241.13 / 249.50 ±8.53 / 262.54 ms │      257.47 / 265.56 ±6.34 / 273.67 ms │  1.06x slower │
│ QQuery 32 │     619.79 / 663.16 ±51.08 / 756.35 ms │     680.68 / 737.31 ±74.14 / 873.72 ms │  1.11x slower │
│ QQuery 33 │  1450.27 / 1517.97 ±51.77 / 1563.70 ms │  1484.10 / 1600.61 ±74.13 / 1703.00 ms │  1.05x slower │
│ QQuery 34 │  1445.99 / 1568.67 ±94.21 / 1715.08 ms │  1544.05 / 1563.42 ±17.29 / 1595.70 ms │     no change │
│ QQuery 35 │     224.09 / 261.14 ±61.93 / 384.80 ms │     225.96 / 289.65 ±70.65 / 400.69 ms │  1.11x slower │
│ QQuery 36 │         65.53 / 70.30 ±7.36 / 84.91 ms │         69.60 / 79.06 ±9.85 / 96.02 ms │  1.12x slower │
│ QQuery 37 │         35.08 / 37.94 ±3.17 / 42.52 ms │         34.45 / 34.99 ±0.47 / 35.84 ms │ +1.08x faster │
│ QQuery 38 │         45.66 / 51.40 ±6.30 / 62.67 ms │         41.03 / 44.68 ±3.91 / 51.71 ms │ +1.15x faster │
│ QQuery 39 │      145.70 / 153.38 ±6.11 / 163.51 ms │     145.81 / 161.73 ±11.26 / 178.32 ms │  1.05x slower │
│ QQuery 40 │         14.27 / 17.30 ±4.11 / 25.45 ms │         14.32 / 14.55 ±0.15 / 14.78 ms │ +1.19x faster │
│ QQuery 41 │         13.24 / 17.49 ±7.98 / 33.44 ms │         13.37 / 13.51 ±0.11 / 13.66 ms │ +1.29x faster │
│ QQuery 42 │         13.08 / 14.67 ±2.46 / 19.58 ms │         13.06 / 13.29 ±0.35 / 13.97 ms │ +1.10x faster │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 19204.70ms │
│ Total Time (codex_in-progress-column-buckets)   │ 19023.99ms │
│ Average Time (HEAD)                             │   446.62ms │
│ Average Time (codex_in-progress-column-buckets) │   442.42ms │
│ Queries Faster                                  │          9 │
│ Queries Slower                                  │         10 │
│ Queries with No Change                          │         24 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	100.0s
Peak memory	17.4 GiB
Avg memory	5.2 GiB
CPU user	922.9s
CPU sys	92.5s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	100.0s
Peak memory	18.0 GiB
Avg memory	5.3 GiB
CPU user	925.6s
CPU sys	84.6s
Peak spill	0 B

File an issue against this benchmark runner

Rachelint · 2026-10-06T04:42:52Z

run benchmark clickbench_partitioned

baseline:
ref: cf6819d
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

adriangbot · 2026-10-06T04:48:03Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highcpu-48 (12 vCPU / 65 GiB) | Linux bench-c6009488158-3087-9tzfx 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  48
On-line CPU(s) list:                     0-47
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     48
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:                               3 MiB (48 instances)
L1i cache:                               3 MiB (48 instances)
L2 cache:                                96 MiB (48 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-47
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 codex/in-progress-column-buckets (df55545) to cf6819d diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "cf6819d7ef43de4da3eda8ff75118fa51ddc6c01"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-10-06T04:59:04Z

🤖 Benchmark completed (GKE) | trigger

Instance: c4a-highcpu-48 (12 vCPU / 65 GiB)

Comparing codex/in-progress-column-buckets (df55545) to cf6819d diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "cf6819d7ef43de4da3eda8ff75118fa51ddc6c01"

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  48
On-line CPU(s) list:                     0-47
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     48
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:                               3 MiB (48 instances)
L1i cache:                               3 MiB (48 instances)
L2 cache:                                96 MiB (48 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-47
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 codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ codex_in-progress-column-buckets ┃       Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━┩
│ QQuery 0  │    1.26 ms │                          1.25 ms │    no change │
│ QQuery 1  │   11.55 ms │                         11.73 ms │    no change │
│ QQuery 2  │   35.51 ms │                         35.28 ms │    no change │
│ QQuery 3  │   30.24 ms │                         30.13 ms │    no change │
│ QQuery 4  │  185.49 ms │                        192.79 ms │    no change │
│ QQuery 5  │  283.29 ms │                        279.58 ms │    no change │
│ QQuery 6  │    1.29 ms │                          1.29 ms │    no change │
│ QQuery 7  │   12.87 ms │                         12.87 ms │    no change │
│ QQuery 8  │  290.33 ms │                        301.81 ms │    no change │
│ QQuery 9  │  472.87 ms │                        465.51 ms │    no change │
│ QQuery 10 │   66.16 ms │                         67.69 ms │    no change │
│ QQuery 11 │   76.60 ms │                         79.11 ms │    no change │
│ QQuery 12 │  267.73 ms │                        276.79 ms │    no change │
│ QQuery 13 │  352.57 ms │                        370.73 ms │ 1.05x slower │
│ QQuery 14 │  280.06 ms │                        278.47 ms │    no change │
│ QQuery 15 │  215.69 ms │                        227.51 ms │ 1.05x slower │
│ QQuery 16 │  573.84 ms │                        585.11 ms │    no change │
│ QQuery 17 │  497.87 ms │                        530.40 ms │ 1.07x slower │
│ QQuery 18 │ 1107.97 ms │                       1155.30 ms │    no change │
│ QQuery 19 │   27.17 ms │                         27.34 ms │    no change │
│ QQuery 20 │  516.62 ms │                        516.66 ms │    no change │
│ QQuery 21 │  511.65 ms │                        516.57 ms │    no change │
│ QQuery 22 │ 1031.51 ms │                        996.65 ms │    no change │
│ QQuery 23 │ 3251.76 ms │                       3133.30 ms │    no change │
│ QQuery 24 │   40.04 ms │                         40.89 ms │    no change │
│ QQuery 25 │  103.24 ms │                        105.04 ms │    no change │
│ QQuery 26 │   40.19 ms │                         41.01 ms │    no change │
│ QQuery 27 │  501.88 ms │                        514.85 ms │    no change │
│ QQuery 28 │ 2906.90 ms │                       2936.27 ms │    no change │
│ QQuery 29 │   40.62 ms │                         40.96 ms │    no change │
│ QQuery 30 │  281.24 ms │                        300.52 ms │ 1.07x slower │
│ QQuery 31 │  246.94 ms │                        273.25 ms │ 1.11x slower │
│ QQuery 32 │  663.91 ms │                        742.83 ms │ 1.12x slower │
│ QQuery 33 │ 1512.19 ms │                       1475.77 ms │    no change │
│ QQuery 34 │ 1503.46 ms │                       1468.30 ms │    no change │
│ QQuery 35 │  226.29 ms │                        243.02 ms │ 1.07x slower │
│ QQuery 36 │   67.50 ms │                         73.57 ms │ 1.09x slower │
│ QQuery 37 │   34.29 ms │                         34.83 ms │    no change │
│ QQuery 38 │   45.19 ms │                         43.00 ms │    no change │
│ QQuery 39 │  135.87 ms │                        149.19 ms │ 1.10x slower │
│ QQuery 40 │   14.16 ms │                         14.43 ms │    no change │
│ QQuery 41 │   13.25 ms │                         13.27 ms │    no change │
│ QQuery 42 │   12.88 ms │                         13.08 ms │    no change │
└───────────┴────────────┴──────────────────────────────────┴──────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 18491.97ms │
│ Total Time (codex_in-progress-column-buckets)   │ 18617.95ms │
│ Average Time (HEAD)                             │   430.05ms │
│ Average Time (codex_in-progress-column-buckets) │   432.98ms │
│ Queries Faster                                  │          0 │
│ Queries Slower                                  │          9 │
│ Queries with No Change                          │         34 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Distribution per query (min / mean ±stddev / max):

Comparing HEAD and codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃       codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │           1.26 / 3.95 ±5.34 / 14.62 ms │           1.25 / 3.95 ±5.35 / 14.65 ms │     no change │
│ QQuery 1  │         11.55 / 11.70 ±0.10 / 11.80 ms │         11.73 / 11.92 ±0.11 / 12.01 ms │     no change │
│ QQuery 2  │         35.51 / 35.67 ±0.09 / 35.75 ms │         35.28 / 35.49 ±0.14 / 35.66 ms │     no change │
│ QQuery 3  │         30.24 / 30.74 ±0.57 / 31.80 ms │         30.13 / 30.40 ±0.18 / 30.67 ms │     no change │
│ QQuery 4  │      185.49 / 188.07 ±2.18 / 191.37 ms │      192.79 / 194.95 ±2.44 / 199.15 ms │     no change │
│ QQuery 5  │      283.29 / 286.58 ±4.65 / 295.76 ms │      279.58 / 286.19 ±4.44 / 292.08 ms │     no change │
│ QQuery 6  │            1.29 / 1.44 ±0.22 / 1.88 ms │            1.29 / 1.43 ±0.22 / 1.86 ms │     no change │
│ QQuery 7  │         12.87 / 12.98 ±0.08 / 13.09 ms │         12.87 / 12.97 ±0.05 / 13.02 ms │     no change │
│ QQuery 8  │      290.33 / 300.78 ±7.11 / 309.78 ms │      301.81 / 306.15 ±3.00 / 310.07 ms │     no change │
│ QQuery 9  │     472.87 / 489.12 ±12.06 / 502.42 ms │      465.51 / 476.25 ±6.78 / 484.74 ms │     no change │
│ QQuery 10 │         66.16 / 72.45 ±8.55 / 88.63 ms │        67.69 / 73.21 ±10.74 / 94.69 ms │     no change │
│ QQuery 11 │         76.60 / 76.92 ±0.25 / 77.31 ms │         79.11 / 82.38 ±3.71 / 89.51 ms │  1.07x slower │
│ QQuery 12 │      267.73 / 277.39 ±7.24 / 290.04 ms │     276.79 / 285.10 ±10.81 / 305.40 ms │     no change │
│ QQuery 13 │      352.57 / 358.61 ±6.25 / 370.66 ms │     370.73 / 384.69 ±11.87 / 398.85 ms │  1.07x slower │
│ QQuery 14 │     280.06 / 290.91 ±10.15 / 310.09 ms │      278.47 / 289.83 ±6.38 / 295.96 ms │     no change │
│ QQuery 15 │      215.69 / 221.51 ±4.76 / 229.68 ms │      227.51 / 232.07 ±2.39 / 234.17 ms │     no change │
│ QQuery 16 │     573.84 / 589.23 ±20.56 / 629.74 ms │     585.11 / 608.17 ±24.93 / 656.59 ms │     no change │
│ QQuery 17 │     497.87 / 525.78 ±26.72 / 572.73 ms │      530.40 / 541.67 ±6.80 / 551.00 ms │     no change │
│ QQuery 18 │  1107.97 / 1139.02 ±25.02 / 1170.44 ms │  1155.30 / 1194.82 ±33.29 / 1247.63 ms │     no change │
│ QQuery 19 │        27.17 / 34.14 ±13.55 / 61.24 ms │         27.34 / 27.86 ±0.57 / 28.92 ms │ +1.23x faster │
│ QQuery 20 │     516.62 / 547.07 ±35.23 / 615.41 ms │      516.66 / 525.67 ±8.77 / 539.79 ms │     no change │
│ QQuery 21 │     511.65 / 526.32 ±12.29 / 543.13 ms │     516.57 / 537.33 ±14.48 / 556.81 ms │     no change │
│ QQuery 22 │  1031.51 / 1051.03 ±20.33 / 1089.44 ms │   996.65 / 1017.22 ±19.85 / 1048.63 ms │     no change │
│ QQuery 23 │ 3251.76 / 3434.99 ±275.43 / 3975.33 ms │  3133.30 / 3239.56 ±86.26 / 3334.93 ms │ +1.06x faster │
│ QQuery 24 │         40.04 / 42.56 ±1.56 / 44.71 ms │     40.89 / 171.81 ±201.96 / 564.22 ms │  4.04x slower │
│ QQuery 25 │      103.24 / 104.28 ±1.55 / 107.30 ms │      105.04 / 106.82 ±1.30 / 108.95 ms │     no change │
│ QQuery 26 │         40.19 / 40.53 ±0.33 / 41.11 ms │         41.01 / 41.40 ±0.38 / 42.00 ms │     no change │
│ QQuery 27 │     501.88 / 531.37 ±17.25 / 551.73 ms │      514.85 / 524.19 ±6.54 / 533.42 ms │     no change │
│ QQuery 28 │  2906.90 / 2965.13 ±44.99 / 3043.79 ms │  2936.27 / 2992.01 ±31.61 / 3032.64 ms │     no change │
│ QQuery 29 │         40.62 / 40.89 ±0.30 / 41.43 ms │         40.96 / 45.78 ±8.62 / 62.98 ms │  1.12x slower │
│ QQuery 30 │      281.24 / 288.16 ±8.20 / 302.67 ms │     300.52 / 331.36 ±32.12 / 381.99 ms │  1.15x slower │
│ QQuery 31 │      246.94 / 254.00 ±5.38 / 261.03 ms │     273.25 / 297.44 ±36.50 / 369.90 ms │  1.17x slower │
│ QQuery 32 │     663.91 / 699.57 ±24.08 / 738.65 ms │     742.83 / 813.84 ±95.29 / 999.30 ms │  1.16x slower │
│ QQuery 33 │ 1512.19 / 1643.24 ±119.66 / 1869.24 ms │ 1475.77 / 1610.63 ±171.29 / 1941.96 ms │     no change │
│ QQuery 34 │  1503.46 / 1520.01 ±20.51 / 1558.97 ms │  1468.30 / 1577.83 ±59.65 / 1638.70 ms │     no change │
│ QQuery 35 │     226.29 / 288.41 ±72.14 / 378.74 ms │     243.02 / 258.29 ±16.84 / 289.95 ms │ +1.12x faster │
│ QQuery 36 │         67.50 / 73.70 ±4.72 / 78.29 ms │         73.57 / 78.78 ±4.33 / 84.37 ms │  1.07x slower │
│ QQuery 37 │         34.29 / 36.93 ±3.41 / 43.62 ms │         34.83 / 37.82 ±2.60 / 41.72 ms │     no change │
│ QQuery 38 │         45.19 / 48.63 ±2.58 / 52.36 ms │         43.00 / 46.27 ±3.72 / 53.31 ms │     no change │
│ QQuery 39 │      135.87 / 147.98 ±7.09 / 155.14 ms │      149.19 / 159.04 ±5.32 / 164.08 ms │  1.07x slower │
│ QQuery 40 │         14.16 / 19.96 ±7.07 / 32.52 ms │         14.43 / 16.99 ±4.16 / 25.28 ms │ +1.18x faster │
│ QQuery 41 │         13.25 / 14.49 ±1.90 / 18.26 ms │         13.27 / 13.76 ±0.42 / 14.44 ms │ +1.05x faster │
│ QQuery 42 │         12.88 / 14.12 ±2.07 / 18.26 ms │         13.08 / 17.84 ±5.44 / 25.72 ms │  1.26x slower │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 19280.35ms │
│ Total Time (codex_in-progress-column-buckets)   │ 19541.16ms │
│ Average Time (HEAD)                             │   448.38ms │
│ Average Time (codex_in-progress-column-buckets) │   454.45ms │
│ Queries Faster                                  │          5 │
│ Queries Slower                                  │         10 │
│ Queries with No Change                          │         28 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	100.0s
Peak memory	17.6 GiB
Avg memory	5.4 GiB
CPU user	928.3s
CPU sys	91.5s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	100.0s
Peak memory	18.5 GiB
Avg memory	5.4 GiB
CPU user	949.7s
CPU sys	84.2s
Peak spill	0 B

File an issue against this benchmark runner

Rachelint · 2026-10-06T09:04:00Z

run benchmark clickbench_partitioned

baseline:
ref: cf6819d
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

adriangbot · 2026-10-06T09:06:42Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c6012972965-3088-585rb 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Comparing codex/in-progress-column-buckets (efd384f) to cf6819d diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "cf6819d7ef43de4da3eda8ff75118fa51ddc6c01"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-10-06T09:19:21Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing codex/in-progress-column-buckets (efd384f) to cf6819d diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "cf6819d7ef43de4da3eda8ff75118fa51ddc6c01"

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 codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │    1.23 ms │                          1.22 ms │     no change │
│ QQuery 1  │   11.57 ms │                         11.74 ms │     no change │
│ QQuery 2  │   35.78 ms │                         35.48 ms │     no change │
│ QQuery 3  │   30.46 ms │                         30.50 ms │     no change │
│ QQuery 4  │  184.60 ms │                        160.08 ms │ +1.15x faster │
│ QQuery 5  │  279.40 ms │                        245.80 ms │ +1.14x faster │
│ QQuery 6  │    1.28 ms │                          1.29 ms │     no change │
│ QQuery 7  │   12.80 ms │                         12.72 ms │     no change │
│ QQuery 8  │  291.60 ms │                        263.19 ms │ +1.11x faster │
│ QQuery 9  │  460.82 ms │                        467.84 ms │     no change │
│ QQuery 10 │   65.57 ms │                         67.80 ms │     no change │
│ QQuery 11 │   77.34 ms │                         78.29 ms │     no change │
│ QQuery 12 │  264.81 ms │                        229.91 ms │ +1.15x faster │
│ QQuery 13 │  356.45 ms │                        379.86 ms │  1.07x slower │
│ QQuery 14 │  279.77 ms │                        262.00 ms │ +1.07x faster │
│ QQuery 15 │  216.55 ms │                        179.50 ms │ +1.21x faster │
│ QQuery 16 │  556.35 ms │                        502.69 ms │ +1.11x faster │
│ QQuery 17 │  514.97 ms │                        434.21 ms │ +1.19x faster │
│ QQuery 18 │ 1088.75 ms │                        980.74 ms │ +1.11x faster │
│ QQuery 19 │   27.81 ms │                         27.53 ms │     no change │
│ QQuery 20 │  528.66 ms │                        528.86 ms │     no change │
│ QQuery 21 │  523.41 ms │                        532.95 ms │     no change │
│ QQuery 22 │ 1023.64 ms │                       1015.07 ms │     no change │
│ QQuery 23 │ 3072.71 ms │                       3294.05 ms │  1.07x slower │
│ QQuery 24 │   40.36 ms │                         40.48 ms │     no change │
│ QQuery 25 │  105.14 ms │                        105.87 ms │     no change │
│ QQuery 26 │   40.98 ms │                         40.96 ms │     no change │
│ QQuery 27 │  516.11 ms │                        510.56 ms │     no change │
│ QQuery 28 │ 2874.77 ms │                       2869.46 ms │     no change │
│ QQuery 29 │   40.57 ms │                         40.39 ms │     no change │
│ QQuery 30 │  282.67 ms │                        256.11 ms │ +1.10x faster │
│ QQuery 31 │  246.41 ms │                        242.78 ms │     no change │
│ QQuery 32 │  635.04 ms │                        662.94 ms │     no change │
│ QQuery 33 │ 1453.40 ms │                       1302.74 ms │ +1.12x faster │
│ QQuery 34 │ 1474.83 ms │                       1299.09 ms │ +1.14x faster │
│ QQuery 35 │  228.29 ms │                        202.61 ms │ +1.13x faster │
│ QQuery 36 │   66.56 ms │                         60.17 ms │ +1.11x faster │
│ QQuery 37 │   35.46 ms │                         34.59 ms │     no change │
│ QQuery 38 │   43.01 ms │                         42.61 ms │     no change │
│ QQuery 39 │  137.41 ms │                        132.85 ms │     no change │
│ QQuery 40 │   14.36 ms │                         14.42 ms │     no change │
│ QQuery 41 │   13.65 ms │                         13.48 ms │     no change │
│ QQuery 42 │   13.10 ms │                         12.98 ms │     no change │
└───────────┴────────────┴──────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 18168.45ms │
│ Total Time (codex_in-progress-column-buckets)   │ 17628.42ms │
│ Average Time (HEAD)                             │   422.52ms │
│ Average Time (codex_in-progress-column-buckets) │   409.96ms │
│ Queries Faster                                  │         14 │
│ Queries Slower                                  │          2 │
│ Queries with No Change                          │         27 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Distribution per query (min / mean ±stddev / max):

Comparing HEAD and codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃       codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │           1.23 / 4.01 ±5.46 / 14.93 ms │           1.22 / 4.15 ±5.77 / 15.70 ms │     no change │
│ QQuery 1  │         11.57 / 11.83 ±0.18 / 12.11 ms │         11.74 / 11.88 ±0.15 / 12.15 ms │     no change │
│ QQuery 2  │         35.78 / 36.16 ±0.41 / 36.92 ms │         35.48 / 35.83 ±0.20 / 36.04 ms │     no change │
│ QQuery 3  │         30.46 / 31.33 ±0.65 / 32.43 ms │         30.50 / 30.88 ±0.29 / 31.23 ms │     no change │
│ QQuery 4  │      184.60 / 188.33 ±3.33 / 192.88 ms │      160.08 / 164.58 ±2.53 / 166.69 ms │ +1.14x faster │
│ QQuery 5  │      279.40 / 283.91 ±3.89 / 290.15 ms │      245.80 / 254.33 ±5.08 / 260.01 ms │ +1.12x faster │
│ QQuery 6  │            1.28 / 1.43 ±0.22 / 1.87 ms │            1.29 / 1.44 ±0.23 / 1.89 ms │     no change │
│ QQuery 7  │         12.80 / 13.03 ±0.20 / 13.36 ms │         12.72 / 12.91 ±0.12 / 13.07 ms │     no change │
│ QQuery 8  │      291.60 / 296.81 ±4.09 / 303.64 ms │      263.19 / 267.07 ±3.03 / 271.50 ms │ +1.11x faster │
│ QQuery 9  │     460.82 / 478.05 ±11.79 / 497.37 ms │     467.84 / 480.23 ±11.73 / 500.02 ms │     no change │
│ QQuery 10 │         65.57 / 65.93 ±0.31 / 66.49 ms │         67.80 / 69.65 ±1.97 / 73.43 ms │  1.06x slower │
│ QQuery 11 │         77.34 / 78.00 ±0.62 / 79.00 ms │         78.29 / 79.28 ±0.82 / 80.52 ms │     no change │
│ QQuery 12 │      264.81 / 275.74 ±8.15 / 288.95 ms │      229.91 / 235.02 ±6.22 / 246.20 ms │ +1.17x faster │
│ QQuery 13 │      356.45 / 365.04 ±7.34 / 377.75 ms │      379.86 / 388.23 ±5.00 / 395.27 ms │  1.06x slower │
│ QQuery 14 │      279.77 / 283.73 ±3.83 / 290.89 ms │      262.00 / 267.80 ±5.02 / 276.40 ms │ +1.06x faster │
│ QQuery 15 │     216.55 / 225.85 ±10.10 / 242.54 ms │      179.50 / 183.47 ±5.12 / 192.80 ms │ +1.23x faster │
│ QQuery 16 │     556.35 / 586.54 ±18.66 / 611.18 ms │     502.69 / 522.98 ±22.77 / 566.88 ms │ +1.12x faster │
│ QQuery 17 │      514.97 / 521.76 ±5.24 / 530.23 ms │      434.21 / 445.21 ±6.74 / 454.78 ms │ +1.17x faster │
│ QQuery 18 │  1088.75 / 1116.54 ±44.42 / 1204.93 ms │     980.74 / 993.70 ±8.36 / 1002.19 ms │ +1.12x faster │
│ QQuery 19 │       27.81 / 50.30 ±32.34 / 111.14 ms │         27.53 / 27.87 ±0.23 / 28.24 ms │ +1.80x faster │
│ QQuery 20 │     528.66 / 544.43 ±12.33 / 564.04 ms │     528.86 / 540.97 ±10.61 / 557.85 ms │     no change │
│ QQuery 21 │      523.41 / 529.75 ±5.89 / 539.59 ms │      532.95 / 543.20 ±9.17 / 560.41 ms │     no change │
│ QQuery 22 │  1023.64 / 1036.64 ±12.21 / 1059.26 ms │  1015.07 / 1051.85 ±28.21 / 1095.53 ms │     no change │
│ QQuery 23 │ 3072.71 / 3287.02 ±132.72 / 3489.28 ms │  3294.05 / 3352.51 ±76.62 / 3500.71 ms │     no change │
│ QQuery 24 │       40.36 / 77.05 ±54.54 / 181.32 ms │       40.48 / 55.29 ±28.84 / 112.97 ms │ +1.39x faster │
│ QQuery 25 │    105.14 / 203.51 ±192.20 / 587.91 ms │    105.87 / 238.24 ±249.14 / 736.11 ms │  1.17x slower │
│ QQuery 26 │         40.98 / 41.84 ±1.01 / 43.65 ms │         40.96 / 41.54 ±0.78 / 43.09 ms │     no change │
│ QQuery 27 │      516.11 / 522.81 ±6.61 / 534.30 ms │     510.56 / 525.38 ±12.42 / 545.41 ms │     no change │
│ QQuery 28 │  2874.77 / 2917.77 ±32.06 / 2953.85 ms │  2869.46 / 2886.48 ±17.47 / 2917.64 ms │     no change │
│ QQuery 29 │         40.57 / 40.74 ±0.15 / 40.97 ms │         40.39 / 40.76 ±0.36 / 41.42 ms │     no change │
│ QQuery 30 │     282.67 / 300.42 ±16.79 / 325.94 ms │     256.11 / 269.46 ±16.91 / 302.23 ms │ +1.11x faster │
│ QQuery 31 │      246.41 / 252.37 ±7.06 / 265.47 ms │      242.78 / 250.11 ±7.27 / 263.73 ms │     no change │
│ QQuery 32 │   635.04 / 820.60 ±185.06 / 1074.28 ms │     662.94 / 772.58 ±83.44 / 871.08 ms │ +1.06x faster │
│ QQuery 33 │ 1453.40 / 1545.66 ±115.61 / 1772.90 ms │ 1302.74 / 1450.91 ±207.78 / 1863.20 ms │ +1.07x faster │
│ QQuery 34 │  1474.83 / 1511.89 ±36.73 / 1569.04 ms │ 1299.09 / 1493.82 ±152.35 / 1713.61 ms │     no change │
│ QQuery 35 │     228.29 / 286.71 ±68.35 / 412.37 ms │     202.61 / 268.80 ±78.63 / 393.22 ms │ +1.07x faster │
│ QQuery 36 │         66.56 / 68.96 ±2.15 / 72.48 ms │         60.17 / 65.81 ±5.30 / 74.23 ms │     no change │
│ QQuery 37 │         35.46 / 39.93 ±4.56 / 48.17 ms │         34.59 / 38.77 ±5.18 / 48.21 ms │     no change │
│ QQuery 38 │         43.01 / 46.99 ±2.68 / 49.72 ms │         42.61 / 45.18 ±1.73 / 47.66 ms │     no change │
│ QQuery 39 │     137.41 / 154.64 ±10.40 / 168.67 ms │      132.85 / 138.52 ±4.64 / 146.37 ms │ +1.12x faster │
│ QQuery 40 │         14.36 / 14.60 ±0.27 / 15.10 ms │         14.42 / 14.90 ±0.53 / 15.92 ms │     no change │
│ QQuery 41 │         13.65 / 16.48 ±5.29 / 27.05 ms │         13.48 / 14.98 ±2.45 / 19.87 ms │ +1.10x faster │
│ QQuery 42 │         13.10 / 13.38 ±0.27 / 13.84 ms │         12.98 / 13.71 ±0.98 / 15.63 ms │     no change │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 19188.51ms │
│ Total Time (codex_in-progress-column-buckets)   │ 18590.27ms │
│ Average Time (HEAD)                             │   446.24ms │
│ Average Time (codex_in-progress-column-buckets) │   432.33ms │
│ Queries Faster                                  │         17 │
│ Queries Slower                                  │          3 │
│ Queries with No Change                          │         23 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	100.0s
Peak memory	16.7 GiB
Avg memory	5.5 GiB
CPU user	922.5s
CPU sys	88.0s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	95.0s
Peak memory	17.6 GiB
Avg memory	5.7 GiB
CPU user	881.0s
CPU sys	88.1s
Peak spill	0 B

File an issue against this benchmark runner

Rachelint · 2026-10-06T09:30:43Z

run benchmark clickbench_partitioned

baseline:
ref: cf6819d
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

adriangbot · 2026-10-06T09:34:35Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c6013401450-3089-llhsk 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

Architecture:                            aarch64
CPU op-mode(s):                          64-bit
Byte Order:                              Little Endian
CPU(s):                                  16
On-line CPU(s) list:                     0-15
Vendor ID:                               ARM
Model name:                              Neoverse-V2
Model:                                   1
Thread(s) per core:                      1
Core(s) per cluster:                     16
Socket(s):                               -
Cluster(s):                              1
Stepping:                                r0p1
BogoMIPS:                                2000.00
Flags:                                   fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache:                               1 MiB (16 instances)
L1i cache:                               1 MiB (16 instances)
L2 cache:                                32 MiB (16 instances)
L3 cache:                                80 MiB (1 instance)
NUMA node(s):                            1
NUMA node0 CPU(s):                       0-15
Vulnerability Gather data sampling:      Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit:             Not affected
Vulnerability L1tf:                      Not affected
Vulnerability Mds:                       Not affected
Vulnerability Meltdown:                  Not affected
Vulnerability Mmio stale data:           Not affected
Vulnerability Reg file data sampling:    Not affected
Vulnerability Retbleed:                  Not affected
Vulnerability Spec rstack overflow:      Not affected
Vulnerability Spec store bypass:         Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:                Mitigation; __user pointer sanitization
Vulnerability Spectre v2:                Mitigation; CSV2, BHB
Vulnerability Srbds:                     Not affected
Vulnerability Tsa:                       Not affected
Vulnerability Tsx async abort:           Not affected
Vulnerability Vmscape:                   Not affected

Comparing codex/in-progress-column-buckets (efd384f) to cf6819d diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "cf6819d7ef43de4da3eda8ff75118fa51ddc6c01"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-10-06T09:43:04Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing codex/in-progress-column-buckets (efd384f) to cf6819d diff

Run configuration

run benchmark clickbench_partitioned
env:
  DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"
baseline:
  ref: "cf6819d7ef43de4da3eda8ff75118fa51ddc6c01"

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 codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │    1.25 ms │                          1.26 ms │     no change │
│ QQuery 1  │   11.64 ms │                         11.64 ms │     no change │
│ QQuery 2  │   35.45 ms │                         35.27 ms │     no change │
│ QQuery 3  │   30.67 ms │                         30.86 ms │     no change │
│ QQuery 4  │  185.80 ms │                        162.64 ms │ +1.14x faster │
│ QQuery 5  │  280.34 ms │                        246.16 ms │ +1.14x faster │
│ QQuery 6  │    1.27 ms │                          1.29 ms │     no change │
│ QQuery 7  │   12.62 ms │                         12.82 ms │     no change │
│ QQuery 8  │  293.55 ms │                        260.41 ms │ +1.13x faster │
│ QQuery 9  │  453.63 ms │                        466.61 ms │     no change │
│ QQuery 10 │   65.91 ms │                         68.32 ms │     no change │
│ QQuery 11 │   76.34 ms │                         79.12 ms │     no change │
│ QQuery 12 │  264.53 ms │                        228.70 ms │ +1.16x faster │
│ QQuery 13 │  358.44 ms │                        379.47 ms │  1.06x slower │
│ QQuery 14 │  276.96 ms │                        256.66 ms │ +1.08x faster │
│ QQuery 15 │  218.18 ms │                        178.27 ms │ +1.22x faster │
│ QQuery 16 │  563.72 ms │                        499.05 ms │ +1.13x faster │
│ QQuery 17 │  504.99 ms │                        438.56 ms │ +1.15x faster │
│ QQuery 18 │ 1076.71 ms │                        967.54 ms │ +1.11x faster │
│ QQuery 19 │   27.78 ms │                         27.71 ms │     no change │
│ QQuery 20 │  533.04 ms │                        527.20 ms │     no change │
│ QQuery 21 │  518.60 ms │                        524.95 ms │     no change │
│ QQuery 22 │ 1054.16 ms │                       1052.12 ms │     no change │
│ QQuery 23 │ 3225.77 ms │                       3231.36 ms │     no change │
│ QQuery 24 │   40.03 ms │                         40.27 ms │     no change │
│ QQuery 25 │  104.87 ms │                        105.26 ms │     no change │
│ QQuery 26 │   40.35 ms │                         41.55 ms │     no change │
│ QQuery 27 │  523.07 ms │                        513.72 ms │     no change │
│ QQuery 28 │ 2926.85 ms │                       2849.19 ms │     no change │
│ QQuery 29 │   40.57 ms │                         40.38 ms │     no change │
│ QQuery 30 │  283.52 ms │                        260.66 ms │ +1.09x faster │
│ QQuery 31 │  244.48 ms │                        245.88 ms │     no change │
│ QQuery 32 │  633.45 ms │                        638.70 ms │     no change │
│ QQuery 33 │ 1448.44 ms │                       1306.18 ms │ +1.11x faster │
│ QQuery 34 │ 1565.27 ms │                       1315.71 ms │ +1.19x faster │
│ QQuery 35 │  222.63 ms │                        203.72 ms │ +1.09x faster │
│ QQuery 36 │   63.70 ms │                         61.27 ms │     no change │
│ QQuery 37 │   34.76 ms │                         35.83 ms │     no change │
│ QQuery 38 │   40.17 ms │                         41.75 ms │     no change │
│ QQuery 39 │  144.18 ms │                        137.32 ms │     no change │
│ QQuery 40 │   14.08 ms │                         14.14 ms │     no change │
│ QQuery 41 │   13.38 ms │                         13.50 ms │     no change │
│ QQuery 42 │   12.77 ms │                         12.90 ms │     no change │
└───────────┴────────────┴──────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 18467.89ms │
│ Total Time (codex_in-progress-column-buckets)   │ 17565.92ms │
│ Average Time (HEAD)                             │   429.49ms │
│ Average Time (codex_in-progress-column-buckets) │   408.51ms │
│ Queries Faster                                  │         13 │
│ Queries Slower                                  │          1 │
│ Queries with No Change                          │         29 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Distribution per query (min / mean ±stddev / max):

Comparing HEAD and codex_in-progress-column-buckets
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃       codex_in-progress-column-buckets ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │           1.25 / 3.97 ±5.41 / 14.78 ms │           1.26 / 3.98 ±5.41 / 14.80 ms │     no change │
│ QQuery 1  │         11.64 / 11.77 ±0.09 / 11.91 ms │         11.64 / 11.78 ±0.08 / 11.87 ms │     no change │
│ QQuery 2  │         35.45 / 35.81 ±0.22 / 36.13 ms │         35.27 / 35.90 ±0.71 / 37.26 ms │     no change │
│ QQuery 3  │         30.67 / 31.29 ±0.63 / 32.50 ms │         30.86 / 30.91 ±0.04 / 30.96 ms │     no change │
│ QQuery 4  │      185.80 / 187.67 ±1.38 / 189.55 ms │      162.64 / 164.23 ±1.13 / 166.02 ms │ +1.14x faster │
│ QQuery 5  │      280.34 / 286.18 ±6.37 / 297.72 ms │      246.16 / 249.62 ±1.77 / 250.97 ms │ +1.15x faster │
│ QQuery 6  │            1.27 / 1.42 ±0.23 / 1.87 ms │            1.29 / 1.44 ±0.24 / 1.91 ms │     no change │
│ QQuery 7  │         12.62 / 12.84 ±0.16 / 13.03 ms │         12.82 / 12.94 ±0.11 / 13.12 ms │     no change │
│ QQuery 8  │      293.55 / 296.18 ±3.71 / 303.32 ms │      260.41 / 264.54 ±2.87 / 269.00 ms │ +1.12x faster │
│ QQuery 9  │      453.63 / 467.58 ±8.52 / 477.46 ms │      466.61 / 471.21 ±4.25 / 479.27 ms │     no change │
│ QQuery 10 │         65.91 / 66.35 ±0.30 / 66.78 ms │         68.32 / 74.04 ±9.90 / 93.79 ms │  1.12x slower │
│ QQuery 11 │         76.34 / 77.17 ±0.70 / 78.23 ms │         79.12 / 79.97 ±0.93 / 81.68 ms │     no change │
│ QQuery 12 │      264.53 / 271.87 ±5.94 / 281.85 ms │      228.70 / 232.95 ±6.15 / 245.17 ms │ +1.17x faster │
│ QQuery 13 │      358.44 / 367.61 ±6.69 / 374.03 ms │      379.47 / 391.33 ±7.39 / 401.30 ms │  1.06x slower │
│ QQuery 14 │     276.96 / 286.27 ±10.92 / 306.15 ms │     256.66 / 271.12 ±13.92 / 295.22 ms │ +1.06x faster │
│ QQuery 15 │     218.18 / 228.37 ±16.90 / 262.06 ms │      178.27 / 186.30 ±7.72 / 200.76 ms │ +1.23x faster │
│ QQuery 16 │     563.72 / 587.95 ±18.06 / 618.40 ms │     499.05 / 519.04 ±27.40 / 573.20 ms │ +1.13x faster │
│ QQuery 17 │      504.99 / 512.26 ±9.28 / 529.64 ms │      438.56 / 444.98 ±9.58 / 463.72 ms │ +1.15x faster │
│ QQuery 18 │  1076.71 / 1121.89 ±35.77 / 1170.98 ms │    967.54 / 991.96 ±19.52 / 1024.44 ms │ +1.13x faster │
│ QQuery 19 │         27.78 / 32.20 ±8.59 / 49.38 ms │         27.71 / 28.29 ±0.78 / 29.80 ms │ +1.14x faster │
│ QQuery 20 │     533.04 / 545.77 ±14.88 / 571.40 ms │      527.20 / 534.89 ±9.96 / 553.83 ms │     no change │
│ QQuery 21 │      518.60 / 529.11 ±8.36 / 540.93 ms │      524.95 / 532.06 ±7.17 / 544.67 ms │     no change │
│ QQuery 22 │  1054.16 / 1099.78 ±51.43 / 1195.45 ms │  1052.12 / 1071.85 ±25.46 / 1121.40 ms │     no change │
│ QQuery 23 │ 3225.77 / 3450.85 ±223.71 / 3811.84 ms │ 3231.36 / 3431.07 ±211.96 / 3821.60 ms │     no change │
│ QQuery 24 │         40.03 / 40.63 ±0.59 / 41.72 ms │       40.27 / 85.88 ±85.34 / 256.47 ms │  2.11x slower │
│ QQuery 25 │    104.87 / 214.86 ±216.90 / 648.64 ms │      105.26 / 106.37 ±1.07 / 108.33 ms │ +2.02x faster │
│ QQuery 26 │         40.35 / 41.28 ±0.81 / 42.73 ms │       41.55 / 64.23 ±43.38 / 150.98 ms │  1.56x slower │
│ QQuery 27 │      523.07 / 532.27 ±9.80 / 550.84 ms │      513.72 / 521.13 ±7.03 / 533.61 ms │     no change │
│ QQuery 28 │  2926.85 / 2971.51 ±53.98 / 3075.59 ms │  2849.19 / 2872.27 ±12.27 / 2882.95 ms │     no change │
│ QQuery 29 │       40.57 / 74.30 ±43.48 / 149.45 ms │         40.38 / 45.30 ±8.82 / 62.94 ms │ +1.64x faster │
│ QQuery 30 │     283.52 / 298.18 ±12.10 / 315.35 ms │      260.66 / 269.47 ±7.17 / 282.05 ms │ +1.11x faster │
│ QQuery 31 │     244.48 / 251.84 ±10.11 / 271.75 ms │      245.88 / 254.08 ±5.04 / 260.27 ms │     no change │
│ QQuery 32 │     633.45 / 686.91 ±38.61 / 752.43 ms │     638.70 / 727.36 ±94.25 / 866.84 ms │  1.06x slower │
│ QQuery 33 │ 1448.44 / 1571.22 ±140.43 / 1841.49 ms │  1306.18 / 1373.72 ±55.21 / 1470.09 ms │ +1.14x faster │
│ QQuery 34 │  1565.27 / 1637.93 ±54.03 / 1722.65 ms │ 1315.71 / 1432.53 ±120.23 / 1622.62 ms │ +1.14x faster │
│ QQuery 35 │     222.63 / 279.43 ±66.09 / 404.68 ms │     203.72 / 224.97 ±28.96 / 281.80 ms │ +1.24x faster │
│ QQuery 36 │         63.70 / 72.17 ±9.48 / 90.17 ms │         61.27 / 63.74 ±2.31 / 66.72 ms │ +1.13x faster │
│ QQuery 37 │         34.76 / 36.99 ±2.75 / 42.24 ms │         35.83 / 37.11 ±1.07 / 38.75 ms │     no change │
│ QQuery 38 │         40.17 / 44.80 ±3.20 / 50.22 ms │         41.75 / 47.32 ±3.68 / 52.50 ms │  1.06x slower │
│ QQuery 39 │      144.18 / 152.82 ±7.27 / 161.19 ms │      137.32 / 143.26 ±6.31 / 155.15 ms │ +1.07x faster │
│ QQuery 40 │         14.08 / 15.68 ±2.80 / 21.26 ms │         14.14 / 14.39 ±0.20 / 14.68 ms │ +1.09x faster │
│ QQuery 41 │         13.38 / 16.26 ±3.69 / 22.93 ms │         13.50 / 16.51 ±3.68 / 21.83 ms │     no change │
│ QQuery 42 │         12.77 / 13.05 ±0.31 / 13.65 ms │         12.90 / 13.86 ±1.41 / 16.67 ms │  1.06x slower │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                               ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                               │ 19464.28ms │
│ Total Time (codex_in-progress-column-buckets)   │ 18349.90ms │
│ Average Time (HEAD)                             │   452.66ms │
│ Average Time (codex_in-progress-column-buckets) │   426.74ms │
│ Queries Faster                                  │         19 │
│ Queries Slower                                  │          7 │
│ Queries with No Change                          │         17 │
│ Queries with Failure                            │          0 │
└─────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	100.0s
Peak memory	18.3 GiB
Avg memory	5.6 GiB
CPU user	920.1s
CPU sys	94.5s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	95.0s
Peak memory	17.3 GiB
Avg memory	5.7 GiB
CPU user	877.6s
CPU sys	88.0s
Peak spill	0 B

File an issue against this benchmark runner

Rachelint · 2026-10-06T11:13:03Z

run benchmark clickbench_partitioned

baseline:
ref: 126109f
changed:
ref: 70ac2ee
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

Rachelint · 2026-10-06T11:13:19Z

run benchmark clickbench_partitioned

baseline:
ref: efd384f
changed:
ref: 70ac2ee
env:
DATAFUSION_EXECUTION_HASH_AGGREGATE_BUCKET_THRESHOLD: "262144"

adriangbot · 2026-10-06T11:15:24Z

Benchmark for this request failed before finishing (Kubernetes reason: BackoffLimitExceeded).

Benchmarks requested: clickbench_partitioned

Runner log (last 40 lines)

2026-10-06T11:14:42.725450Z  INFO runner starting benchmark runner bench_type=Datafusion, pr_url=https://github.com/apache/datafusion/pull/25959, benchmarks=clickbench_partitioned
2026-10-06T11:14:42.787264Z  INFO benchmark_controller::runner::bench_datafusion === Cloning PR branch ===
2026-10-06T11:14:42.787285Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["clone", "--depth=200", "https://github.com/apache/datafusion.git", "/workspace/datafusion-branch"], cwd="/"
2026-10-06T11:14:47.797873Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["fetch", "origin", "refs/pull/25959/head:codex/in-progress-column-buckets", "main"], cwd="/workspace/datafusion-branch"
2026-10-06T11:14:52.800430Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["checkout", "codex/in-progress-column-buckets"], cwd="/workspace/datafusion-branch"
2026-10-06T11:14:57.801739Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["merge-base", "HEAD", "origin/main"], cwd="/workspace/datafusion-branch"
2026-10-06T11:15:02.804028Z  INFO benchmark_controller::runner::bench_datafusion === Checking out custom changed ref === changed_ref=70ac2eed7ad82bd0915ad526cd735bdcacbf7458
2026-10-06T11:15:02.804044Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["fetch", "origin", "refs/pull/25959/head"], cwd="/workspace/datafusion-branch"
2026-10-06T11:15:07.805759Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["fetch", "origin"], cwd="/workspace/datafusion-branch"
2026-10-06T11:15:12.808243Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["-c", "advice.detachedHead=false", "checkout", "70ac2eed7ad82bd0915ad526cd735bdcacbf7458"], cwd="/workspace/datafusion-branch"
2026-10-06T11:15:17.810436Z ERROR runner benchmark failed error=git checkout: git -c advice.detachedHead=false checkout 70ac2eed7ad82bd0915ad526cd735bdcacbf7458 exited with code 128
stdout:

stderr:
fatal: reference is not a tree: 70ac2eed7ad82bd0915ad526cd735bdcacbf7458

Kubernetes message

Job has reached the specified backoff limit

File an issue against this benchmark runner

adriangbot · 2026-10-06T11:15:58Z

Benchmark for this request failed before finishing (Kubernetes reason: BackoffLimitExceeded).

Benchmarks requested: clickbench_partitioned

Runner log (last 40 lines)

2026-10-06T11:15:17.957421Z  INFO runner starting benchmark runner bench_type=Datafusion, pr_url=https://github.com/apache/datafusion/pull/25959, benchmarks=clickbench_partitioned
2026-10-06T11:15:18.017761Z  INFO benchmark_controller::runner::bench_datafusion === Cloning PR branch ===
2026-10-06T11:15:18.017781Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["clone", "--depth=200", "https://github.com/apache/datafusion.git", "/workspace/datafusion-branch"], cwd="/"
2026-10-06T11:15:23.028430Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["fetch", "origin", "refs/pull/25959/head:codex/in-progress-column-buckets", "main"], cwd="/workspace/datafusion-branch"
2026-10-06T11:15:28.030819Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["checkout", "codex/in-progress-column-buckets"], cwd="/workspace/datafusion-branch"
2026-10-06T11:15:33.032880Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["merge-base", "HEAD", "origin/main"], cwd="/workspace/datafusion-branch"
2026-10-06T11:15:38.035088Z  INFO benchmark_controller::runner::bench_datafusion === Checking out custom changed ref === changed_ref=70ac2eed7ad82bd0915ad526cd735bdcacbf7458
2026-10-06T11:15:38.035103Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["fetch", "origin", "refs/pull/25959/head"], cwd="/workspace/datafusion-branch"
2026-10-06T11:15:43.036584Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["fetch", "origin"], cwd="/workspace/datafusion-branch"
2026-10-06T11:15:48.038258Z  INFO benchmark_controller::runner::shell running command cmd=git, args=["-c", "advice.detachedHead=false", "checkout", "70ac2eed7ad82bd0915ad526cd735bdcacbf7458"], cwd="/workspace/datafusion-branch"
2026-10-06T11:15:53.040462Z ERROR runner benchmark failed error=git checkout: git -c advice.detachedHead=false checkout 70ac2eed7ad82bd0915ad526cd735bdcacbf7458 exited with code 128
stdout:

stderr:
fatal: reference is not a tree: 70ac2eed7ad82bd0915ad526cd735bdcacbf7458

Kubernetes message

Job has reached the specified backoff limit

File an issue against this benchmark runner

jayzhan211 and others added 7 commits September 25, 2026 09:39

feat: aggregate large final hash tables as hash buckets (experimental…

4daca18

…, off by default)

feat: flush a partial hash aggregation table that reaches the bucket …

0ba8308

…threshold while its groups do not recur

perf: reuse one hash table across aggregation buckets and skip compac…

efa8dde

…ting buckets of unique groups

refactor: move the aggregation bucket driver into its own module

f91b6ba

perf: move a final table of unique groups into buckets at a quarter o…

2cf0fe9

…f the threshold

perf: let the table of an aggregation bucket keep its group keys wher…

f9df89e

…e its input holds them

perf: route final aggregate buckets by column

b964422

github-actions Bot added documentation Improvements or additions to documentation sqllogictest SQL Logic Tests (.slt) common Related to common crate physical-plan Changes to the physical-plan crate labels Oct 2, 2026

perf: benchmark 32 aggregate buckets without compaction

313a3a9

perf: scatter StringView columns into aggregate buckets

085c221

perf: retain gather when scattering StringView bucket ranges

cf6819d

perf: increase final aggregation fanout to 128 buckets

33aad7a

perf: increase final aggregation fanout to 256 buckets

df55545

perf: bound partial aggregation working sets

efd384f

github-actions Bot added the physical-expr Changes to the physical-expr crates label Oct 6, 2026

This was referenced Oct 6, 2026

bench: isolate Partial aggregation gains without Final buckets (do not merge) #26081

Draft

bench: isolate continuous partial flushing against main #26098

Draft

perf: reuse StringView storage across partial flushes #26109

Closed

Conversation

Rachelint commented Oct 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Which issue does this PR close?

Rationale for this change

What changes are included in this PR?

What is the testing strategy for this PR?

Are there any user-facing changes?

Uh oh!

Rachelint commented Oct 2, 2026

Uh oh!

adriangbot commented Oct 2, 2026

Uh oh!

adriangbot commented Oct 2, 2026

Uh oh!

Rachelint commented Oct 2, 2026

Uh oh!

adriangbot commented Oct 2, 2026

Uh oh!

adriangbot commented Oct 2, 2026

Uh oh!

Rachelint commented Oct 4, 2026

Uh oh!

adriangbot commented Oct 4, 2026

Uh oh!

adriangbot commented Oct 4, 2026

Uh oh!

Rachelint commented Oct 4, 2026

Uh oh!

adriangbot commented Oct 4, 2026

Uh oh!

Rachelint commented Oct 4, 2026

Uh oh!

adriangbot commented Oct 4, 2026

Uh oh!

adriangbot commented Oct 4, 2026

Uh oh!

adriangbot commented Oct 4, 2026

Uh oh!

Rachelint commented Oct 6, 2026

Uh oh!

adriangbot commented Oct 6, 2026

Uh oh!

adriangbot commented Oct 6, 2026

Uh oh!

Rachelint commented Oct 6, 2026

Uh oh!

adriangbot commented Oct 6, 2026

Uh oh!

adriangbot commented Oct 6, 2026

Uh oh!

Rachelint commented Oct 6, 2026

Uh oh!

adriangbot commented Oct 6, 2026

Uh oh!

adriangbot commented Oct 6, 2026

Uh oh!

Rachelint commented Oct 6, 2026

Uh oh!

adriangbot commented Oct 6, 2026

Uh oh!

adriangbot commented Oct 6, 2026

Uh oh!

Rachelint commented Oct 6, 2026

Uh oh!

Rachelint commented Oct 6, 2026

Uh oh!

adriangbot commented Oct 6, 2026

Uh oh!

adriangbot commented Oct 6, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Rachelint commented Oct 2, 2026 •

edited

Loading