perf: coalesce batches before sending to distributor channels in RepartitionExec by gabotechs · Pull Request #22010 · apache/datafusion

gabotechs · 2026-05-04T16:43:43Z

Which issue does this PR close?

Closes #.

Rationale for this change

Today RepartitionExec emits one small batch per (input batch × non-empty output partition), then coalesces those small batches back into target-sized ones on the consumer side. That means the channel layer (memory accounting, gate, await suspensions) does work proportional to num_partitions per input batch, even though each small batch only carries batch_size / num_partitions rows. Moving the coalescing producer-side, before the gate, collapses the channel traffic and reduces sensitivity to high output fanout.

This gets much worst in https://github.com/datafusion-contrib/datafusion-distributed, which uses RepartitionExec as the foundational backbone for network shuffles: it scales up the RepartitionExec output partitions to match P (partitions) * W (workers), where P is typically in the range of 12-24, and W can be from 1 to thousands. The fanout overhead that the current RepartitionExec introduces is pretty big in that setup.

Ideally, the overhead of fanning many small RecordBatches should be as small as possible.

What changes are included in this PR?

A producer-side coalescer per output partition, shared across input tasks, that runs before the channel's gate.
The redundant consumer-side coalescer in PerPartitionStream is removed.
BatchPartitioner is unchanged — only the layer between the partitioner and the channel.

Are these changes tested?

Yes, by existing tests

Are there any user-facing changes?

No

…butor channel, instead of after.

gabotechs · 2026-05-04T16:51:07Z

run benchmarks

adriangbot · 2026-05-04T16:54:25Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4372819392-2010-r97fl 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: tpch
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-05-04T16:54:28Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4372819392-2009-rm68r 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: tpcds
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-05-04T16:54:36Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4372819392-2008-hw4tt 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-05-04T17:08:35Z

🤖 Benchmark completed (GKE) | trigger

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

CPU Details (lscpu)

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

Details

Comparing HEAD and gabotechs_repartition-coalesce-batches-berfore-channels
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                           HEAD ┃ gabotechs_repartition-coalesce-batches-berfore-channels ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │ 40.46 / 41.52 ±1.18 / 43.63 ms │                          40.41 / 41.15 ±1.24 / 43.64 ms │     no change │
│ QQuery 2  │ 21.08 / 21.60 ±0.36 / 22.12 ms │                          20.79 / 21.44 ±0.86 / 23.13 ms │     no change │
│ QQuery 3  │ 36.64 / 39.44 ±2.36 / 43.12 ms │                          35.29 / 36.96 ±0.99 / 37.74 ms │ +1.07x faster │
│ QQuery 4  │ 18.66 / 19.26 ±0.56 / 19.94 ms │                          18.11 / 18.69 ±0.45 / 19.49 ms │     no change │
│ QQuery 5  │ 44.54 / 46.28 ±1.33 / 47.63 ms │                          42.16 / 43.31 ±1.07 / 44.58 ms │ +1.07x faster │
│ QQuery 6  │ 17.29 / 17.54 ±0.14 / 17.72 ms │                          17.32 / 17.49 ±0.16 / 17.70 ms │     no change │
│ QQuery 7  │ 51.30 / 52.24 ±0.97 / 53.66 ms │                          48.05 / 50.55 ±1.84 / 53.21 ms │     no change │
│ QQuery 8  │ 47.50 / 47.78 ±0.19 / 48.07 ms │                          46.76 / 47.04 ±0.15 / 47.21 ms │     no change │
│ QQuery 9  │ 52.44 / 54.23 ±2.35 / 58.86 ms │                          52.35 / 53.57 ±1.05 / 55.19 ms │     no change │
│ QQuery 10 │ 66.06 / 66.58 ±0.70 / 67.96 ms │                          65.60 / 65.80 ±0.20 / 66.17 ms │     no change │
│ QQuery 11 │ 14.13 / 14.28 ±0.13 / 14.50 ms │                          13.69 / 14.22 ±0.67 / 15.50 ms │     no change │
│ QQuery 12 │ 26.84 / 27.95 ±1.07 / 29.90 ms │                          26.02 / 26.83 ±0.72 / 27.96 ms │     no change │
│ QQuery 13 │ 35.47 / 36.35 ±0.64 / 37.38 ms │                          36.08 / 38.17 ±1.93 / 40.58 ms │     no change │
│ QQuery 14 │ 27.21 / 27.83 ±0.52 / 28.66 ms │                          26.81 / 27.00 ±0.11 / 27.10 ms │     no change │
│ QQuery 15 │ 33.66 / 34.42 ±0.67 / 35.47 ms │                          33.03 / 33.75 ±1.10 / 35.95 ms │     no change │
│ QQuery 16 │ 15.32 / 15.50 ±0.11 / 15.65 ms │                          15.40 / 15.48 ±0.09 / 15.66 ms │     no change │
│ QQuery 17 │ 77.93 / 78.94 ±0.83 / 80.05 ms │                          76.46 / 78.81 ±2.00 / 82.42 ms │     no change │
│ QQuery 18 │ 69.44 / 70.82 ±0.73 / 71.39 ms │                          64.71 / 65.86 ±1.70 / 69.22 ms │ +1.08x faster │
│ QQuery 19 │ 37.96 / 38.43 ±0.47 / 39.27 ms │                          37.63 / 38.14 ±0.40 / 38.79 ms │     no change │
│ QQuery 20 │ 39.46 / 39.69 ±0.16 / 39.96 ms │                          38.92 / 39.06 ±0.09 / 39.21 ms │     no change │
│ QQuery 21 │ 60.35 / 62.58 ±1.20 / 63.60 ms │                          58.74 / 61.39 ±2.76 / 66.68 ms │     no change │
│ QQuery 22 │ 24.23 / 24.50 ±0.18 / 24.77 ms │                          23.61 / 24.10 ±0.37 / 24.71 ms │     no change │
└───────────┴────────────────────────────────┴─────────────────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                                                      ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                                                      │ 877.77ms │
│ Total Time (gabotechs_repartition-coalesce-batches-berfore-channels)   │ 858.79ms │
│ Average Time (HEAD)                                                    │  39.90ms │
│ Average Time (gabotechs_repartition-coalesce-batches-berfore-channels) │  39.04ms │
│ Queries Faster                                                         │        3 │
│ Queries Slower                                                         │        0 │
│ Queries with No Change                                                 │       19 │
│ Queries with Failure                                                   │        0 │
└────────────────────────────────────────────────────────────────────────┴──────────┘

Resource Usage

tpch — base (merge-base)

Metric	Value
Wall time	5.0s
Peak memory	5.7 GiB
Avg memory	5.1 GiB
CPU user	32.8s
CPU sys	2.2s
Peak spill	0 B

tpch — branch

Metric	Value
Wall time	5.0s
Peak memory	5.5 GiB
Avg memory	5.0 GiB
CPU user	31.3s
CPU sys	2.4s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-05-04T17:13:54Z

🤖 Benchmark completed (GKE) | trigger

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

CPU Details (lscpu)

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

Details

Comparing HEAD and gabotechs_repartition-coalesce-batches-berfore-channels
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃ gabotechs_repartition-coalesce-batches-berfore-channels ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │              7.46 / 8.00 ±0.78 / 9.55 ms │                             7.18 / 7.67 ±0.83 / 9.33 ms │     no change │
│ QQuery 2  │        148.00 / 148.61 ±0.41 / 149.21 ms │                       145.30 / 145.80 ±0.38 / 146.29 ms │     no change │
│ QQuery 3  │        117.05 / 118.12 ±0.68 / 119.20 ms │                       113.67 / 115.85 ±1.43 / 118.10 ms │     no change │
│ QQuery 4  │    1483.20 / 1499.50 ±18.91 / 1530.66 ms │                    1363.56 / 1373.26 ±8.02 / 1384.92 ms │ +1.09x faster │
│ QQuery 5  │        180.36 / 182.84 ±1.45 / 184.77 ms │                       175.21 / 176.51 ±2.07 / 180.64 ms │     no change │
│ QQuery 6  │        171.41 / 174.93 ±2.03 / 177.28 ms │                       153.52 / 156.09 ±2.44 / 159.83 ms │ +1.12x faster │
│ QQuery 7  │        332.78 / 333.83 ±1.45 / 336.70 ms │                       338.12 / 341.26 ±2.82 / 345.90 ms │     no change │
│ QQuery 8  │        116.31 / 117.40 ±0.89 / 118.95 ms │                       114.87 / 115.99 ±0.93 / 117.49 ms │     no change │
│ QQuery 9  │         99.63 / 104.48 ±4.85 / 113.12 ms │                          95.65 / 97.16 ±1.11 / 98.54 ms │ +1.08x faster │
│ QQuery 10 │        103.89 / 104.87 ±0.81 / 105.81 ms │                       102.77 / 103.59 ±0.62 / 104.69 ms │     no change │
│ QQuery 11 │    1037.80 / 1068.55 ±22.60 / 1107.06 ms │                       961.56 / 972.97 ±8.85 / 988.62 ms │ +1.10x faster │
│ QQuery 12 │           46.39 / 47.20 ±0.62 / 48.09 ms │                          44.33 / 45.17 ±1.47 / 48.11 ms │     no change │
│ QQuery 13 │        407.61 / 411.17 ±2.61 / 415.66 ms │                       395.62 / 397.89 ±1.57 / 399.47 ms │     no change │
│ QQuery 14 │     1043.99 / 1049.68 ±5.59 / 1057.68 ms │                    1044.78 / 1049.72 ±2.50 / 1051.46 ms │     no change │
│ QQuery 15 │           14.38 / 15.02 ±0.38 / 15.44 ms │                          15.21 / 15.46 ±0.20 / 15.69 ms │     no change │
│ QQuery 16 │              7.47 / 7.71 ±0.26 / 8.19 ms │                             7.69 / 7.80 ±0.14 / 8.05 ms │     no change │
│ QQuery 17 │        203.05 / 203.99 ±0.85 / 205.33 ms │                       216.93 / 220.05 ±2.95 / 224.29 ms │  1.08x slower │
│ QQuery 18 │        122.79 / 125.20 ±1.88 / 128.60 ms │                       127.88 / 129.63 ±1.33 / 131.22 ms │     no change │
│ QQuery 19 │        160.88 / 168.30 ±4.32 / 173.88 ms │                       158.58 / 160.70 ±2.06 / 164.56 ms │     no change │
│ QQuery 20 │           14.50 / 14.78 ±0.20 / 15.04 ms │                          13.72 / 14.03 ±0.44 / 14.90 ms │ +1.05x faster │
│ QQuery 21 │           21.05 / 21.24 ±0.27 / 21.76 ms │                          19.71 / 20.00 ±0.17 / 20.21 ms │ +1.06x faster │
│ QQuery 22 │       515.48 / 570.44 ±29.34 / 597.68 ms │                       483.80 / 486.49 ±2.50 / 491.00 ms │ +1.17x faster │
│ QQuery 23 │    1101.83 / 1132.44 ±20.98 / 1155.58 ms │                    1083.93 / 1089.76 ±4.69 / 1096.69 ms │     no change │
│ QQuery 24 │        679.33 / 681.55 ±2.64 / 686.67 ms │                       660.57 / 672.00 ±5.99 / 677.56 ms │     no change │
│ QQuery 25 │        320.97 / 323.39 ±1.78 / 326.15 ms │                       320.53 / 322.46 ±2.54 / 327.12 ms │     no change │
│ QQuery 26 │           78.56 / 79.20 ±0.33 / 79.51 ms │                          78.05 / 78.69 ±0.38 / 79.08 ms │     no change │
│ QQuery 27 │              7.68 / 7.82 ±0.10 / 7.94 ms │                             7.32 / 7.46 ±0.15 / 7.75 ms │     no change │
│ QQuery 28 │        154.93 / 156.03 ±0.89 / 157.58 ms │                       150.00 / 153.82 ±5.02 / 163.56 ms │     no change │
│ QQuery 29 │        258.10 / 261.09 ±2.50 / 265.42 ms │                       256.81 / 259.25 ±1.90 / 262.32 ms │     no change │
│ QQuery 30 │           43.35 / 43.99 ±0.65 / 45.14 ms │                          42.37 / 43.04 ±0.68 / 44.35 ms │     no change │
│ QQuery 31 │        172.04 / 174.63 ±3.03 / 180.25 ms │                       167.18 / 168.65 ±1.30 / 170.35 ms │     no change │
│ QQuery 32 │           14.54 / 15.05 ±0.45 / 15.74 ms │                          13.74 / 14.08 ±0.27 / 14.44 ms │ +1.07x faster │
│ QQuery 33 │        144.93 / 148.59 ±3.22 / 154.32 ms │                       140.40 / 142.04 ±1.70 / 145.01 ms │     no change │
│ QQuery 34 │              7.80 / 7.95 ±0.08 / 8.03 ms │                            7.21 / 8.77 ±2.65 / 14.06 ms │  1.10x slower │
│ QQuery 35 │        105.42 / 106.60 ±0.88 / 107.95 ms │                       102.06 / 102.31 ±0.23 / 102.59 ms │     no change │
│ QQuery 36 │              7.29 / 7.50 ±0.14 / 7.71 ms │                            6.94 / 8.74 ±3.35 / 15.45 ms │  1.17x slower │
│ QQuery 37 │              8.96 / 9.23 ±0.16 / 9.39 ms │                             8.45 / 9.03 ±0.38 / 9.59 ms │     no change │
│ QQuery 38 │           88.53 / 90.58 ±2.07 / 94.45 ms │                          85.34 / 88.51 ±2.68 / 92.41 ms │     no change │
│ QQuery 39 │        123.94 / 131.68 ±5.74 / 141.69 ms │                       118.18 / 124.02 ±5.99 / 135.53 ms │ +1.06x faster │
│ QQuery 40 │          96.65 / 99.13 ±1.53 / 100.59 ms │                         96.11 / 98.99 ±2.79 / 103.97 ms │     no change │
│ QQuery 41 │           15.44 / 15.62 ±0.17 / 15.83 ms │                          14.81 / 14.95 ±0.15 / 15.24 ms │     no change │
│ QQuery 42 │        109.57 / 111.81 ±1.83 / 114.96 ms │                       107.75 / 109.26 ±1.71 / 112.45 ms │     no change │
│ QQuery 43 │              6.37 / 6.45 ±0.06 / 6.55 ms │                             5.81 / 5.89 ±0.08 / 6.03 ms │ +1.09x faster │
│ QQuery 44 │           11.63 / 11.83 ±0.11 / 11.97 ms │                          10.73 / 11.01 ±0.27 / 11.51 ms │ +1.07x faster │
│ QQuery 45 │           47.10 / 48.25 ±0.96 / 49.70 ms │                          46.07 / 46.70 ±0.49 / 47.49 ms │     no change │
│ QQuery 46 │              9.52 / 9.68 ±0.13 / 9.85 ms │                             8.64 / 8.75 ±0.10 / 8.92 ms │ +1.11x faster │
│ QQuery 47 │        856.31 / 865.06 ±6.00 / 873.60 ms │                       745.77 / 754.16 ±5.55 / 761.80 ms │ +1.15x faster │
│ QQuery 48 │        284.85 / 292.21 ±6.44 / 302.04 ms │                       274.37 / 280.99 ±5.72 / 290.89 ms │     no change │
│ QQuery 49 │        249.44 / 253.23 ±3.32 / 259.36 ms │                       245.02 / 248.85 ±3.21 / 254.41 ms │     no change │
│ QQuery 50 │        193.57 / 197.01 ±2.52 / 200.69 ms │                       191.53 / 195.18 ±2.60 / 198.26 ms │     no change │
│ QQuery 51 │        174.08 / 175.63 ±1.67 / 178.60 ms │                       172.29 / 175.86 ±3.40 / 182.26 ms │     no change │
│ QQuery 52 │        108.99 / 110.95 ±2.03 / 114.83 ms │                       107.78 / 109.88 ±3.69 / 117.24 ms │     no change │
│ QQuery 53 │        104.21 / 105.04 ±0.71 / 106.05 ms │                       102.41 / 102.71 ±0.21 / 103.06 ms │     no change │
│ QQuery 54 │        145.44 / 147.00 ±1.20 / 149.12 ms │                       146.66 / 148.25 ±1.95 / 152.06 ms │     no change │
│ QQuery 55 │        106.50 / 107.32 ±0.51 / 108.05 ms │                       106.37 / 108.59 ±1.96 / 111.75 ms │     no change │
│ QQuery 56 │        139.42 / 140.49 ±1.16 / 142.71 ms │                       139.84 / 141.35 ±0.93 / 142.43 ms │     no change │
│ QQuery 57 │        165.64 / 168.98 ±2.76 / 173.88 ms │                       169.57 / 171.52 ±1.47 / 174.10 ms │     no change │
│ QQuery 58 │        244.43 / 247.09 ±1.37 / 248.20 ms │                       274.98 / 278.56 ±3.30 / 284.63 ms │  1.13x slower │
│ QQuery 59 │        196.17 / 197.86 ±2.17 / 201.87 ms │                       195.81 / 197.98 ±2.11 / 201.30 ms │     no change │
│ QQuery 60 │        142.06 / 146.13 ±3.09 / 150.00 ms │                       141.87 / 142.85 ±0.91 / 144.41 ms │     no change │
│ QQuery 61 │           14.64 / 14.78 ±0.09 / 14.87 ms │                          14.05 / 14.13 ±0.05 / 14.19 ms │     no change │
│ QQuery 62 │       896.06 / 937.50 ±30.47 / 979.59 ms │                       879.96 / 880.65 ±0.44 / 881.13 ms │ +1.06x faster │
│ QQuery 63 │        104.49 / 106.70 ±2.20 / 110.77 ms │                       103.06 / 103.80 ±0.79 / 105.22 ms │     no change │
│ QQuery 64 │        661.77 / 668.72 ±3.76 / 671.70 ms │                       662.28 / 672.08 ±8.22 / 686.46 ms │     no change │
│ QQuery 65 │        270.03 / 272.82 ±2.43 / 277.23 ms │                       245.66 / 251.38 ±5.82 / 258.61 ms │ +1.09x faster │
│ QQuery 66 │        226.49 / 236.66 ±7.47 / 248.85 ms │                       221.28 / 227.99 ±7.16 / 239.84 ms │     no change │
│ QQuery 67 │        311.22 / 318.26 ±7.76 / 333.36 ms │                      298.50 / 309.25 ±15.00 / 339.05 ms │     no change │
│ QQuery 68 │              9.01 / 9.16 ±0.13 / 9.39 ms │                             9.08 / 9.47 ±0.20 / 9.63 ms │     no change │
│ QQuery 69 │         99.28 / 103.46 ±4.70 / 112.62 ms │                        99.64 / 102.40 ±4.36 / 111.05 ms │     no change │
│ QQuery 70 │        328.88 / 334.52 ±4.67 / 340.85 ms │                       308.72 / 320.75 ±9.97 / 335.86 ms │     no change │
│ QQuery 71 │        136.84 / 139.08 ±1.52 / 141.30 ms │                       134.17 / 136.87 ±4.48 / 145.81 ms │     no change │
│ QQuery 72 │        576.67 / 583.46 ±6.11 / 592.79 ms │                      543.99 / 561.54 ±14.48 / 583.50 ms │     no change │
│ QQuery 73 │              7.13 / 7.34 ±0.17 / 7.63 ms │                             6.66 / 6.94 ±0.26 / 7.42 ms │ +1.06x faster │
│ QQuery 74 │       638.76 / 654.56 ±11.49 / 668.55 ms │                      583.41 / 593.74 ±10.07 / 612.43 ms │ +1.10x faster │
│ QQuery 75 │        273.56 / 276.78 ±2.25 / 279.79 ms │                       268.08 / 270.48 ±3.74 / 277.92 ms │     no change │
│ QQuery 76 │        134.07 / 136.64 ±2.50 / 141.28 ms │                       129.92 / 132.41 ±3.22 / 138.73 ms │     no change │
│ QQuery 77 │        191.94 / 193.20 ±1.45 / 195.33 ms │                       187.87 / 190.00 ±2.71 / 195.22 ms │     no change │
│ QQuery 78 │        319.07 / 325.38 ±5.07 / 332.36 ms │                       320.07 / 322.00 ±1.55 / 324.19 ms │     no change │
│ QQuery 79 │        242.26 / 244.92 ±2.68 / 250.05 ms │                       232.72 / 238.93 ±5.93 / 248.81 ms │     no change │
│ QQuery 80 │        299.09 / 302.85 ±2.00 / 304.50 ms │                       310.15 / 313.74 ±2.33 / 316.74 ms │     no change │
│ QQuery 81 │           26.02 / 26.33 ±0.35 / 26.96 ms │                          26.13 / 26.32 ±0.18 / 26.61 ms │     no change │
│ QQuery 82 │           39.51 / 40.17 ±0.70 / 41.50 ms │                          39.76 / 41.10 ±2.38 / 45.86 ms │     no change │
│ QQuery 83 │           35.49 / 35.84 ±0.22 / 36.14 ms │                          37.94 / 38.08 ±0.20 / 38.45 ms │  1.06x slower │
│ QQuery 84 │           46.50 / 46.68 ±0.14 / 46.84 ms │                          46.92 / 47.52 ±0.31 / 47.81 ms │     no change │
│ QQuery 85 │        139.89 / 141.31 ±0.76 / 142.07 ms │                       142.33 / 145.26 ±2.84 / 150.48 ms │     no change │
│ QQuery 86 │           37.24 / 38.56 ±1.53 / 41.54 ms │                          37.49 / 38.56 ±0.71 / 39.54 ms │     no change │
│ QQuery 87 │              3.53 / 3.65 ±0.18 / 4.01 ms │                             3.52 / 3.80 ±0.49 / 4.78 ms │     no change │
│ QQuery 88 │           99.16 / 99.61 ±0.31 / 99.97 ms │                       100.01 / 100.91 ±0.55 / 101.60 ms │     no change │
│ QQuery 89 │        117.80 / 121.40 ±6.05 / 133.45 ms │                       118.07 / 119.97 ±2.95 / 125.81 ms │     no change │
│ QQuery 90 │           22.77 / 23.26 ±0.52 / 24.15 ms │                          23.08 / 23.49 ±0.31 / 23.94 ms │     no change │
│ QQuery 91 │           58.98 / 59.69 ±0.68 / 60.84 ms │                          59.32 / 60.04 ±0.51 / 60.74 ms │     no change │
│ QQuery 92 │           56.00 / 56.58 ±0.36 / 57.09 ms │                          57.44 / 58.56 ±1.06 / 60.47 ms │     no change │
│ QQuery 93 │        160.13 / 160.86 ±0.87 / 162.53 ms │                       167.44 / 169.18 ±1.78 / 172.11 ms │  1.05x slower │
│ QQuery 94 │           60.62 / 61.10 ±0.40 / 61.56 ms │                          61.65 / 62.33 ±0.47 / 63.01 ms │     no change │
│ QQuery 95 │        111.15 / 111.70 ±0.30 / 112.03 ms │                       114.74 / 115.42 ±0.64 / 116.24 ms │     no change │
│ QQuery 96 │           67.29 / 69.00 ±1.23 / 71.06 ms │                          68.57 / 70.33 ±1.85 / 73.65 ms │     no change │
│ QQuery 97 │        112.63 / 114.49 ±1.40 / 116.71 ms │                       112.05 / 113.37 ±0.72 / 114.16 ms │     no change │
│ QQuery 98 │        148.33 / 151.47 ±2.44 / 155.88 ms │                       150.17 / 152.30 ±1.70 / 154.61 ms │     no change │
│ QQuery 99 │ 10670.11 / 10756.94 ±83.25 / 10864.99 ms │                10809.45 / 10913.68 ±78.62 / 10995.20 ms │     no change │
└───────────┴──────────────────────────────────────────┴─────────────────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                                                      ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                                                      │ 31315.37ms │
│ Total Time (gabotechs_repartition-coalesce-batches-berfore-channels)   │ 30756.79ms │
│ Average Time (HEAD)                                                    │   316.32ms │
│ Average Time (gabotechs_repartition-coalesce-batches-berfore-channels) │   310.67ms │
│ Queries Faster                                                         │         17 │
│ Queries Slower                                                         │          6 │
│ Queries with No Change                                                 │         76 │
│ Queries with Failure                                                   │          0 │
└────────────────────────────────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	160.0s
Peak memory	6.3 GiB
Avg memory	5.6 GiB
CPU user	249.8s
CPU sys	8.3s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	155.0s
Peak memory	6.4 GiB
Avg memory	5.6 GiB
CPU user	236.4s
CPU sys	7.2s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-05-04T17:14:25Z

🤖 Benchmark completed (GKE) | trigger

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

CPU Details (lscpu)

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

Details

Comparing HEAD and gabotechs_repartition-coalesce-batches-berfore-channels
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃ gabotechs_repartition-coalesce-batches-berfore-channels ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.26 / 4.79 ±6.93 / 18.65 ms │                            1.23 / 4.75 ±6.89 / 18.53 ms │     no change │
│ QQuery 1  │        12.46 / 12.83 ±0.22 / 13.06 ms │                          12.50 / 13.00 ±0.34 / 13.57 ms │     no change │
│ QQuery 2  │        37.19 / 37.54 ±0.28 / 37.81 ms │                          37.05 / 37.40 ±0.24 / 37.66 ms │     no change │
│ QQuery 3  │        32.32 / 33.01 ±0.59 / 34.07 ms │                          32.24 / 33.06 ±0.66 / 34.04 ms │     no change │
│ QQuery 4  │     245.78 / 250.61 ±2.69 / 253.44 ms │                       239.58 / 240.32 ±0.74 / 241.66 ms │     no change │
│ QQuery 5  │     293.51 / 296.46 ±2.29 / 299.96 ms │                       287.59 / 289.87 ±1.58 / 291.90 ms │     no change │
│ QQuery 6  │           7.11 / 7.42 ±0.23 / 7.76 ms │                             7.36 / 7.67 ±0.24 / 8.02 ms │     no change │
│ QQuery 7  │        13.66 / 13.90 ±0.15 / 14.14 ms │                          13.91 / 15.03 ±1.98 / 18.97 ms │  1.08x slower │
│ QQuery 8  │     336.96 / 342.66 ±4.52 / 350.05 ms │                       326.56 / 329.77 ±3.90 / 337.35 ms │     no change │
│ QQuery 9  │     474.87 / 481.30 ±4.35 / 487.45 ms │                      470.73 / 483.65 ±10.51 / 501.23 ms │     no change │
│ QQuery 10 │        73.89 / 74.59 ±0.41 / 75.07 ms │                          75.11 / 76.44 ±1.09 / 78.05 ms │     no change │
│ QQuery 11 │        84.52 / 85.16 ±0.83 / 86.76 ms │                          85.85 / 86.54 ±0.70 / 87.77 ms │     no change │
│ QQuery 12 │     282.94 / 287.85 ±4.16 / 294.09 ms │                       279.78 / 283.96 ±3.46 / 288.68 ms │     no change │
│ QQuery 13 │     403.29 / 409.91 ±6.25 / 420.51 ms │                      377.45 / 397.22 ±12.33 / 414.73 ms │     no change │
│ QQuery 14 │     295.55 / 299.43 ±3.18 / 304.92 ms │                       291.12 / 295.59 ±2.50 / 298.38 ms │     no change │
│ QQuery 15 │     289.38 / 297.26 ±5.39 / 304.98 ms │                       287.81 / 291.27 ±2.95 / 295.20 ms │     no change │
│ QQuery 16 │     628.03 / 637.32 ±5.79 / 644.33 ms │                       632.63 / 633.57 ±0.93 / 635.35 ms │     no change │
│ QQuery 17 │     633.30 / 647.04 ±8.05 / 655.82 ms │                      628.58 / 640.93 ±12.32 / 660.90 ms │     no change │
│ QQuery 18 │  1268.93 / 1280.96 ±6.70 / 1287.49 ms │                   1295.20 / 1320.00 ±19.53 / 1353.94 ms │     no change │
│ QQuery 19 │        29.08 / 33.78 ±8.83 / 51.43 ms │                          29.92 / 33.09 ±6.02 / 45.12 ms │     no change │
│ QQuery 20 │    524.94 / 535.15 ±10.42 / 554.48 ms │                       529.02 / 535.77 ±8.04 / 550.64 ms │     no change │
│ QQuery 21 │    603.08 / 612.27 ±12.63 / 637.19 ms │                       602.95 / 612.81 ±8.84 / 626.54 ms │     no change │
│ QQuery 22 │ 1089.34 / 1099.75 ±11.03 / 1119.41 ms │                    1084.01 / 1092.61 ±6.64 / 1100.61 ms │     no change │
│ QQuery 23 │ 3418.32 / 3452.59 ±19.61 / 3478.32 ms │                   3397.70 / 3433.90 ±19.15 / 3450.70 ms │     no change │
│ QQuery 24 │        42.45 / 45.32 ±4.61 / 54.51 ms │                          42.49 / 46.05 ±5.71 / 57.39 ms │     no change │
│ QQuery 25 │     114.23 / 117.26 ±4.38 / 125.91 ms │                       114.84 / 116.60 ±1.67 / 119.32 ms │     no change │
│ QQuery 26 │        43.30 / 45.53 ±2.89 / 51.22 ms │                          44.21 / 46.53 ±1.53 / 48.50 ms │     no change │
│ QQuery 27 │     681.82 / 686.99 ±2.86 / 689.68 ms │                       679.79 / 688.54 ±5.65 / 694.22 ms │     no change │
│ QQuery 28 │ 3047.37 / 3066.65 ±14.36 / 3088.91 ms │                   3069.70 / 3099.23 ±30.33 / 3155.95 ms │     no change │
│ QQuery 29 │        43.02 / 48.45 ±6.21 / 57.96 ms │                         43.11 / 53.09 ±13.16 / 76.63 ms │  1.10x slower │
│ QQuery 30 │     313.18 / 317.67 ±3.56 / 323.03 ms │                       314.32 / 317.42 ±2.60 / 321.47 ms │     no change │
│ QQuery 31 │    302.76 / 320.13 ±14.42 / 346.73 ms │                       304.78 / 308.60 ±4.02 / 315.85 ms │     no change │
│ QQuery 32 │  1015.23 / 1019.34 ±3.73 / 1024.70 ms │                    990.85 / 1008.80 ±14.74 / 1030.51 ms │     no change │
│ QQuery 33 │ 1491.39 / 1509.54 ±10.99 / 1524.95 ms │                   1528.62 / 1540.00 ±15.32 / 1569.48 ms │     no change │
│ QQuery 34 │ 1534.76 / 1549.94 ±12.70 / 1565.19 ms │                   1521.86 / 1535.59 ±12.24 / 1551.64 ms │     no change │
│ QQuery 35 │    308.10 / 322.94 ±11.17 / 335.25 ms │                      291.36 / 315.11 ±15.33 / 331.84 ms │     no change │
│ QQuery 36 │        65.81 / 70.39 ±4.44 / 78.17 ms │                          67.99 / 75.29 ±8.52 / 91.30 ms │  1.07x slower │
│ QQuery 37 │        36.66 / 37.68 ±1.12 / 39.69 ms │                          36.80 / 39.83 ±3.97 / 47.70 ms │  1.06x slower │
│ QQuery 38 │        40.93 / 45.87 ±4.96 / 54.54 ms │                          41.15 / 44.66 ±3.22 / 49.61 ms │     no change │
│ QQuery 39 │     127.61 / 138.86 ±6.38 / 146.83 ms │                       146.78 / 157.88 ±7.34 / 164.77 ms │  1.14x slower │
│ QQuery 40 │        14.87 / 17.03 ±3.75 / 24.52 ms │                          15.41 / 19.34 ±4.33 / 24.82 ms │  1.14x slower │
│ QQuery 41 │        14.71 / 17.24 ±4.30 / 25.82 ms │                          14.69 / 14.85 ±0.15 / 15.11 ms │ +1.16x faster │
│ QQuery 42 │        14.17 / 16.26 ±3.79 / 23.84 ms │                          14.09 / 15.96 ±3.16 / 22.26 ms │     no change │
└───────────┴───────────────────────────────────────┴─────────────────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                                                      ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                                                      │ 20626.67ms │
│ Total Time (gabotechs_repartition-coalesce-batches-berfore-channels)   │ 20631.60ms │
│ Average Time (HEAD)                                                    │   479.69ms │
│ Average Time (gabotechs_repartition-coalesce-batches-berfore-channels) │   479.80ms │
│ Queries Faster                                                         │          1 │
│ Queries Slower                                                         │          6 │
│ Queries with No Change                                                 │         36 │
│ Queries with Failure                                                   │          0 │
└────────────────────────────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	105.0s
Peak memory	30.7 GiB
Avg memory	23.2 GiB
CPU user	1088.8s
CPU sys	68.7s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	105.0s
Peak memory	28.9 GiB
Avg memory	23.0 GiB
CPU user	1068.9s
CPU sys	72.0s
Peak spill	0 B

File an issue against this benchmark runner

gabotechs · 2026-05-04T19:15:43Z

run benchmarks
env:
DATAFUSION_EXECUTION_TARGET_PARTITIONS: 64

adriangbot · 2026-05-04T19:17:37Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4373801306-2016-np8lv 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-05-04T19:19:34Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4373801306-2018-6nwl5 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: tpch
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-05-04T19:19:35Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4373801306-2017-csbcx 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: tpcds
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-05-04T19:29:41Z

🤖 Benchmark completed (GKE) | trigger

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

CPU Details (lscpu)

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

Details

Comparing HEAD and gabotechs_repartition-coalesce-batches-berfore-channels
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                              HEAD ┃ gabotechs_repartition-coalesce-batches-berfore-channels ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │    44.31 / 44.96 ±1.14 / 47.23 ms │                          43.30 / 44.10 ±0.89 / 45.83 ms │     no change │
│ QQuery 2  │    30.10 / 31.02 ±0.74 / 32.10 ms │                          29.04 / 29.80 ±0.82 / 31.25 ms │     no change │
│ QQuery 3  │    65.02 / 66.26 ±1.35 / 68.62 ms │                          54.32 / 55.33 ±1.08 / 56.86 ms │ +1.20x faster │
│ QQuery 4  │    24.51 / 25.20 ±0.57 / 25.79 ms │                          24.82 / 25.35 ±0.55 / 26.42 ms │     no change │
│ QQuery 5  │    93.43 / 94.24 ±0.75 / 95.54 ms │                          77.51 / 80.48 ±1.97 / 82.68 ms │ +1.17x faster │
│ QQuery 6  │    19.19 / 19.47 ±0.36 / 20.16 ms │                          19.33 / 19.47 ±0.12 / 19.65 ms │     no change │
│ QQuery 7  │   95.89 / 99.24 ±3.09 / 105.01 ms │                          86.87 / 89.10 ±2.33 / 93.02 ms │ +1.11x faster │
│ QQuery 8  │    84.54 / 88.58 ±2.71 / 91.66 ms │                          81.80 / 82.98 ±1.51 / 85.86 ms │ +1.07x faster │
│ QQuery 9  │    93.27 / 94.47 ±0.67 / 95.29 ms │                          88.08 / 90.75 ±1.70 / 92.83 ms │     no change │
│ QQuery 10 │    70.58 / 72.76 ±1.17 / 73.79 ms │                          72.04 / 73.23 ±0.98 / 74.52 ms │     no change │
│ QQuery 11 │    23.00 / 23.82 ±0.57 / 24.31 ms │                          21.85 / 22.92 ±0.74 / 23.59 ms │     no change │
│ QQuery 12 │    46.62 / 47.78 ±1.06 / 49.69 ms │                          43.49 / 46.29 ±1.52 / 47.94 ms │     no change │
│ QQuery 13 │    46.61 / 49.39 ±2.21 / 52.92 ms │                          46.38 / 48.64 ±2.19 / 52.68 ms │     no change │
│ QQuery 14 │    33.37 / 33.97 ±0.57 / 34.77 ms │                          31.97 / 32.45 ±0.34 / 32.97 ms │     no change │
│ QQuery 15 │    41.25 / 41.81 ±0.45 / 42.34 ms │                          40.93 / 41.46 ±0.35 / 41.84 ms │     no change │
│ QQuery 16 │    23.00 / 23.60 ±0.43 / 24.22 ms │                          21.79 / 22.54 ±0.64 / 23.73 ms │     no change │
│ QQuery 17 │ 116.19 / 118.09 ±1.22 / 119.50 ms │                       109.22 / 111.50 ±1.23 / 112.66 ms │ +1.06x faster │
│ QQuery 18 │ 139.69 / 140.42 ±0.42 / 140.92 ms │                          89.65 / 91.11 ±1.09 / 92.50 ms │ +1.54x faster │
│ QQuery 19 │    40.97 / 42.01 ±0.75 / 42.90 ms │                          39.70 / 40.35 ±0.50 / 41.12 ms │     no change │
│ QQuery 20 │    64.96 / 66.08 ±1.04 / 67.62 ms │                          62.53 / 64.43 ±1.31 / 65.63 ms │     no change │
│ QQuery 21 │ 100.95 / 102.27 ±1.49 / 105.19 ms │                          90.71 / 91.86 ±0.84 / 93.24 ms │ +1.11x faster │
│ QQuery 22 │    27.61 / 28.48 ±0.80 / 29.90 ms │                          26.48 / 27.39 ±0.59 / 28.08 ms │     no change │
└───────────┴───────────────────────────────────┴─────────────────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                                                      ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                                                      │ 1353.91ms │
│ Total Time (gabotechs_repartition-coalesce-batches-berfore-channels)   │ 1231.53ms │
│ Average Time (HEAD)                                                    │   61.54ms │
│ Average Time (gabotechs_repartition-coalesce-batches-berfore-channels) │   55.98ms │
│ Queries Faster                                                         │         7 │
│ Queries Slower                                                         │         0 │
│ Queries with No Change                                                 │        15 │
│ Queries with Failure                                                   │         0 │
└────────────────────────────────────────────────────────────────────────┴───────────┘

Resource Usage

tpch — base (merge-base)

Metric	Value
Wall time	10.0s
Peak memory	6.9 GiB
Avg memory	5.4 GiB
CPU user	54.4s
CPU sys	2.6s
Peak spill	0 B

tpch — branch

Metric	Value
Wall time	10.0s
Peak memory	6.8 GiB
Avg memory	5.3 GiB
CPU user	48.9s
CPU sys	2.5s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-05-04T19:33:07Z

🤖 Benchmark completed (GKE) | trigger

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

CPU Details (lscpu)

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

Details

Comparing HEAD and gabotechs_repartition-coalesce-batches-berfore-channels
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃ gabotechs_repartition-coalesce-batches-berfore-channels ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.37 / 4.83 ±6.78 / 18.39 ms │                            1.34 / 4.78 ±6.75 / 18.29 ms │     no change │
│ QQuery 1  │        13.61 / 15.05 ±0.73 / 15.55 ms │                          14.65 / 15.02 ±0.21 / 15.28 ms │     no change │
│ QQuery 2  │        38.44 / 39.46 ±0.61 / 40.23 ms │                          37.74 / 38.10 ±0.29 / 38.59 ms │     no change │
│ QQuery 3  │        32.70 / 33.62 ±1.26 / 36.00 ms │                          30.78 / 31.49 ±0.37 / 31.80 ms │ +1.07x faster │
│ QQuery 4  │     255.92 / 258.26 ±1.96 / 260.63 ms │                       198.45 / 203.50 ±2.56 / 205.31 ms │ +1.27x faster │
│ QQuery 5  │    295.95 / 311.89 ±10.76 / 322.26 ms │                       267.93 / 270.18 ±1.66 / 272.76 ms │ +1.15x faster │
│ QQuery 6  │         9.37 / 10.34 ±0.49 / 10.73 ms │                            9.16 / 9.81 ±0.75 / 11.26 ms │ +1.05x faster │
│ QQuery 7  │        17.58 / 19.48 ±3.58 / 26.64 ms │                          17.10 / 17.68 ±0.30 / 17.93 ms │ +1.10x faster │
│ QQuery 8  │     328.65 / 332.01 ±3.22 / 338.16 ms │                       294.62 / 308.71 ±7.42 / 315.70 ms │ +1.08x faster │
│ QQuery 9  │     371.97 / 380.18 ±4.92 / 385.27 ms │                       389.57 / 398.98 ±6.82 / 409.30 ms │     no change │
│ QQuery 10 │        78.65 / 80.73 ±3.26 / 87.21 ms │                          79.33 / 81.84 ±1.90 / 84.87 ms │     no change │
│ QQuery 11 │        91.45 / 93.02 ±1.05 / 94.30 ms │                          91.50 / 93.55 ±1.08 / 94.72 ms │     no change │
│ QQuery 12 │     286.33 / 289.44 ±2.75 / 292.73 ms │                       268.90 / 276.38 ±6.48 / 286.13 ms │     no change │
│ QQuery 13 │    488.73 / 504.46 ±13.08 / 518.93 ms │                       421.91 / 430.24 ±6.35 / 436.81 ms │ +1.17x faster │
│ QQuery 14 │     291.33 / 295.56 ±2.51 / 298.63 ms │                       263.35 / 269.11 ±8.25 / 285.48 ms │ +1.10x faster │
│ QQuery 15 │    283.45 / 295.51 ±12.63 / 318.35 ms │                       239.50 / 243.51 ±6.14 / 255.70 ms │ +1.21x faster │
│ QQuery 16 │     625.08 / 632.46 ±5.94 / 640.52 ms │                      542.98 / 582.59 ±56.92 / 694.33 ms │ +1.09x faster │
│ QQuery 17 │     633.39 / 644.88 ±9.18 / 659.73 ms │                      566.97 / 587.26 ±14.71 / 612.78 ms │ +1.10x faster │
│ QQuery 18 │ 1359.34 / 1398.33 ±24.08 / 1433.67 ms │                   1082.10 / 1131.79 ±40.20 / 1174.91 ms │ +1.24x faster │
│ QQuery 19 │       31.89 / 41.82 ±15.27 / 71.68 ms │                         29.73 / 36.61 ±11.26 / 59.02 ms │ +1.14x faster │
│ QQuery 20 │    494.40 / 523.95 ±26.34 / 564.16 ms │                      492.38 / 510.41 ±20.13 / 543.07 ms │     no change │
│ QQuery 21 │     586.09 / 596.59 ±8.07 / 609.15 ms │                      587.13 / 615.28 ±28.07 / 665.34 ms │     no change │
│ QQuery 22 │ 1040.79 / 1093.01 ±28.44 / 1127.09 ms │                   1050.59 / 1072.54 ±13.10 / 1086.82 ms │     no change │
│ QQuery 23 │ 2794.05 / 2862.68 ±47.95 / 2929.47 ms │                  2861.04 / 2985.73 ±111.42 / 3181.48 ms │     no change │
│ QQuery 24 │        50.92 / 54.04 ±2.86 / 57.50 ms │                          54.82 / 65.22 ±6.27 / 72.31 ms │  1.21x slower │
│ QQuery 25 │     113.49 / 115.55 ±1.54 / 117.19 ms │                       116.55 / 120.53 ±4.33 / 128.76 ms │     no change │
│ QQuery 26 │        50.65 / 53.98 ±3.74 / 61.05 ms │                          51.27 / 56.75 ±6.49 / 68.89 ms │  1.05x slower │
│ QQuery 27 │     656.64 / 665.41 ±9.50 / 679.85 ms │                      664.18 / 705.87 ±23.33 / 729.39 ms │  1.06x slower │
│ QQuery 28 │ 2874.68 / 2935.40 ±51.59 / 3019.01 ms │                   2915.98 / 2963.98 ±50.35 / 3054.52 ms │     no change │
│ QQuery 29 │        44.31 / 45.09 ±0.89 / 46.76 ms │                         44.28 / 54.50 ±19.17 / 92.82 ms │  1.21x slower │
│ QQuery 30 │     315.79 / 322.04 ±6.47 / 333.30 ms │                      276.58 / 288.19 ±11.64 / 309.45 ms │ +1.12x faster │
│ QQuery 31 │     356.08 / 365.40 ±5.22 / 371.92 ms │                       309.71 / 317.66 ±7.17 / 330.22 ms │ +1.15x faster │
│ QQuery 32 │ 1226.32 / 1239.45 ±14.12 / 1266.55 ms │                     946.30 / 988.81 ±30.26 / 1022.54 ms │ +1.25x faster │
│ QQuery 33 │ 1522.76 / 1617.31 ±95.84 / 1800.86 ms │                   1353.98 / 1410.54 ±45.07 / 1487.67 ms │ +1.15x faster │
│ QQuery 34 │ 1445.09 / 1497.13 ±49.14 / 1582.83 ms │                   1379.58 / 1412.33 ±29.79 / 1455.73 ms │ +1.06x faster │
│ QQuery 35 │    310.46 / 333.37 ±28.54 / 389.60 ms │                     230.34 / 312.22 ±138.34 / 588.26 ms │ +1.07x faster │
│ QQuery 36 │        74.38 / 77.69 ±3.04 / 83.26 ms │                          72.23 / 73.79 ±2.29 / 78.35 ms │ +1.05x faster │
│ QQuery 37 │        42.02 / 44.62 ±2.91 / 50.25 ms │                          40.66 / 41.93 ±1.08 / 43.58 ms │ +1.06x faster │
│ QQuery 38 │        50.65 / 57.22 ±6.97 / 68.94 ms │                         46.65 / 59.84 ±15.50 / 86.55 ms │     no change │
│ QQuery 39 │     151.39 / 160.53 ±5.69 / 168.69 ms │                       157.48 / 166.14 ±7.45 / 175.47 ms │     no change │
│ QQuery 40 │        18.39 / 22.79 ±4.60 / 29.96 ms │                          20.51 / 26.86 ±7.67 / 39.56 ms │  1.18x slower │
│ QQuery 41 │        17.33 / 19.38 ±2.20 / 23.57 ms │                          20.78 / 23.22 ±3.95 / 31.09 ms │  1.20x slower │
│ QQuery 42 │        16.51 / 19.63 ±4.91 / 29.34 ms │                          18.39 / 21.75 ±5.55 / 32.76 ms │  1.11x slower │
└───────────┴───────────────────────────────────────┴─────────────────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                                                      ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                                                      │ 20403.56ms │
│ Total Time (gabotechs_repartition-coalesce-batches-berfore-channels)   │ 19325.20ms │
│ Average Time (HEAD)                                                    │   474.50ms │
│ Average Time (gabotechs_repartition-coalesce-batches-berfore-channels) │   449.42ms │
│ Queries Faster                                                         │         21 │
│ Queries Slower                                                         │          7 │
│ Queries with No Change                                                 │         15 │
│ Queries with Failure                                                   │          0 │
└────────────────────────────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	105.0s
Peak memory	31.3 GiB
Avg memory	25.0 GiB
CPU user	1096.5s
CPU sys	68.3s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	100.0s
Peak memory	33.2 GiB
Avg memory	25.5 GiB
CPU user	1008.9s
CPU sys	71.5s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-05-04T19:35:18Z

🤖 Benchmark completed (GKE) | trigger

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

CPU Details (lscpu)

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

Details

Comparing HEAD and gabotechs_repartition-coalesce-batches-berfore-channels
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                      HEAD ┃ gabotechs_repartition-coalesce-batches-berfore-channels ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │            13.79 / 14.18 ±0.65 / 15.46 ms │                          13.72 / 14.19 ±0.69 / 15.57 ms │     no change │
│ QQuery 2  │         159.22 / 160.09 ±0.61 / 160.74 ms │                       160.90 / 162.32 ±0.82 / 163.14 ms │     no change │
│ QQuery 3  │         117.84 / 119.51 ±1.38 / 121.35 ms │                       117.74 / 118.88 ±0.88 / 120.37 ms │     no change │
│ QQuery 4  │     1318.19 / 1375.95 ±62.45 / 1477.81 ms │                   1342.78 / 1411.19 ±50.60 / 1498.56 ms │     no change │
│ QQuery 5  │         202.08 / 203.76 ±1.18 / 205.00 ms │                       204.12 / 205.77 ±1.54 / 207.72 ms │     no change │
│ QQuery 6  │         139.79 / 141.55 ±2.63 / 146.77 ms │                       144.42 / 145.12 ±1.05 / 147.20 ms │     no change │
│ QQuery 7  │         414.19 / 418.61 ±4.50 / 427.00 ms │                       430.53 / 433.52 ±2.82 / 438.10 ms │     no change │
│ QQuery 8  │         120.58 / 121.74 ±0.81 / 122.95 ms │                       122.07 / 123.38 ±1.53 / 126.34 ms │     no change │
│ QQuery 9  │         124.62 / 136.64 ±8.09 / 148.23 ms │                       126.56 / 135.82 ±7.00 / 143.82 ms │     no change │
│ QQuery 10 │         112.02 / 112.72 ±0.57 / 113.50 ms │                       110.68 / 113.28 ±2.98 / 118.65 ms │     no change │
│ QQuery 11 │         882.55 / 890.18 ±7.52 / 903.32 ms │                     905.46 / 984.28 ±68.02 / 1092.79 ms │  1.11x slower │
│ QQuery 12 │            48.04 / 49.41 ±1.08 / 51.34 ms │                          48.95 / 49.24 ±0.22 / 49.58 ms │     no change │
│ QQuery 13 │         392.20 / 395.44 ±2.01 / 398.43 ms │                       392.91 / 396.89 ±4.05 / 404.66 ms │     no change │
│ QQuery 14 │     1125.26 / 1146.61 ±13.71 / 1159.96 ms │                    1106.68 / 1121.48 ±9.38 / 1133.16 ms │     no change │
│ QQuery 15 │            39.50 / 40.60 ±0.98 / 42.44 ms │                          35.07 / 39.22 ±4.05 / 46.71 ms │     no change │
│ QQuery 16 │            14.02 / 14.51 ±0.27 / 14.79 ms │                          13.45 / 13.71 ±0.21 / 13.98 ms │ +1.06x faster │
│ QQuery 17 │         328.18 / 333.26 ±3.65 / 338.99 ms │                       312.42 / 318.04 ±5.43 / 326.68 ms │     no change │
│ QQuery 18 │         159.79 / 164.81 ±3.36 / 169.33 ms │                       158.20 / 161.08 ±3.20 / 167.26 ms │     no change │
│ QQuery 19 │         157.68 / 159.11 ±1.45 / 161.44 ms │                       158.56 / 162.12 ±4.13 / 170.18 ms │     no change │
│ QQuery 20 │            19.47 / 19.65 ±0.16 / 19.86 ms │                          19.77 / 20.05 ±0.29 / 20.61 ms │     no change │
│ QQuery 21 │            25.17 / 26.30 ±1.74 / 29.74 ms │                          25.20 / 25.48 ±0.21 / 25.69 ms │     no change │
│ QQuery 22 │         480.66 / 484.05 ±3.82 / 491.28 ms │                       486.80 / 490.94 ±3.30 / 494.88 ms │     no change │
│ QQuery 23 │     1302.63 / 1326.79 ±14.30 / 1346.95 ms │                   1325.29 / 1409.07 ±58.96 / 1465.86 ms │  1.06x slower │
│ QQuery 24 │        930.00 / 961.11 ±21.90 / 997.00 ms │                     960.05 / 990.04 ±19.02 / 1014.09 ms │     no change │
│ QQuery 25 │         451.82 / 456.44 ±3.63 / 462.79 ms │                       437.49 / 444.04 ±4.63 / 449.00 ms │     no change │
│ QQuery 26 │         105.03 / 107.61 ±1.55 / 109.72 ms │                       103.94 / 107.03 ±1.68 / 108.74 ms │     no change │
│ QQuery 27 │            14.44 / 14.61 ±0.18 / 14.91 ms │                          12.95 / 13.13 ±0.16 / 13.33 ms │ +1.11x faster │
│ QQuery 28 │         165.93 / 170.40 ±4.81 / 179.77 ms │                       160.11 / 162.64 ±3.45 / 169.40 ms │     no change │
│ QQuery 29 │         375.71 / 388.95 ±7.18 / 397.50 ms │                       379.60 / 388.42 ±8.76 / 402.66 ms │     no change │
│ QQuery 30 │            53.26 / 53.64 ±0.27 / 54.03 ms │                          53.15 / 53.44 ±0.23 / 53.65 ms │     no change │
│ QQuery 31 │         164.72 / 167.69 ±2.52 / 170.73 ms │                       170.37 / 173.12 ±1.43 / 174.34 ms │     no change │
│ QQuery 32 │            21.64 / 22.27 ±0.41 / 22.71 ms │                          24.11 / 24.48 ±0.28 / 24.98 ms │  1.10x slower │
│ QQuery 33 │         151.70 / 154.37 ±3.55 / 161.29 ms │                       158.11 / 161.19 ±3.81 / 168.69 ms │     no change │
│ QQuery 34 │            12.93 / 13.52 ±0.36 / 14.05 ms │                          13.99 / 14.19 ±0.21 / 14.57 ms │     no change │
│ QQuery 35 │         112.93 / 116.09 ±3.43 / 122.35 ms │                       116.75 / 120.92 ±4.20 / 128.22 ms │     no change │
│ QQuery 36 │            14.89 / 15.39 ±0.39 / 15.94 ms │                          16.56 / 17.02 ±0.24 / 17.26 ms │  1.11x slower │
│ QQuery 37 │            12.57 / 12.93 ±0.39 / 13.61 ms │                          12.79 / 13.46 ±0.56 / 14.23 ms │     no change │
│ QQuery 38 │            95.94 / 97.71 ±1.30 / 99.44 ms │                        97.52 / 100.85 ±2.48 / 105.07 ms │     no change │
│ QQuery 39 │         172.15 / 176.18 ±5.46 / 186.99 ms │                       162.15 / 178.28 ±8.13 / 183.26 ms │     no change │
│ QQuery 40 │         137.32 / 145.69 ±6.56 / 155.25 ms │                       128.22 / 133.99 ±6.78 / 147.13 ms │ +1.09x faster │
│ QQuery 41 │            21.74 / 22.04 ±0.30 / 22.57 ms │                          20.83 / 21.71 ±0.45 / 22.03 ms │     no change │
│ QQuery 42 │         114.64 / 117.43 ±3.63 / 124.59 ms │                       113.36 / 117.11 ±6.06 / 129.20 ms │     no change │
│ QQuery 43 │            11.70 / 11.82 ±0.10 / 11.95 ms │                          11.26 / 11.50 ±0.13 / 11.61 ms │     no change │
│ QQuery 44 │            22.73 / 25.65 ±2.38 / 28.80 ms │                          25.33 / 25.78 ±0.50 / 26.70 ms │     no change │
│ QQuery 45 │            76.73 / 77.41 ±0.43 / 77.90 ms │                          73.84 / 77.16 ±2.42 / 80.64 ms │     no change │
│ QQuery 46 │            14.12 / 14.39 ±0.27 / 14.90 ms │                          14.29 / 14.46 ±0.11 / 14.59 ms │     no change │
│ QQuery 47 │        731.41 / 774.81 ±44.60 / 855.73 ms │                      727.84 / 775.07 ±59.99 / 885.87 ms │     no change │
│ QQuery 48 │         278.94 / 283.01 ±5.10 / 292.88 ms │                       286.00 / 295.22 ±9.32 / 312.92 ms │     no change │
│ QQuery 49 │         288.04 / 291.22 ±2.46 / 295.19 ms │                       288.61 / 292.09 ±2.84 / 296.70 ms │     no change │
│ QQuery 50 │        253.24 / 267.90 ±10.73 / 286.33 ms │                       253.67 / 264.38 ±6.84 / 274.41 ms │     no change │
│ QQuery 51 │         213.75 / 220.68 ±6.74 / 231.66 ms │                       211.30 / 215.72 ±5.96 / 227.23 ms │     no change │
│ QQuery 52 │         115.18 / 116.78 ±2.56 / 121.87 ms │                       113.72 / 115.73 ±2.19 / 119.82 ms │     no change │
│ QQuery 53 │         111.66 / 113.76 ±1.71 / 116.79 ms │                       110.96 / 113.06 ±2.54 / 118.04 ms │     no change │
│ QQuery 54 │         164.30 / 167.26 ±2.40 / 170.02 ms │                       163.56 / 166.63 ±3.56 / 171.47 ms │     no change │
│ QQuery 55 │         112.51 / 112.94 ±0.43 / 113.77 ms │                       110.23 / 111.19 ±1.15 / 113.46 ms │     no change │
│ QQuery 56 │         153.08 / 154.93 ±1.43 / 157.26 ms │                       152.45 / 154.43 ±1.52 / 156.73 ms │     no change │
│ QQuery 57 │         199.80 / 201.21 ±1.05 / 202.98 ms │                       198.71 / 199.57 ±0.74 / 200.69 ms │     no change │
│ QQuery 58 │         334.82 / 343.15 ±5.01 / 350.49 ms │                       364.45 / 375.97 ±6.33 / 382.11 ms │  1.10x slower │
│ QQuery 59 │         204.38 / 208.72 ±5.26 / 218.97 ms │                       204.79 / 209.03 ±4.38 / 217.18 ms │     no change │
│ QQuery 60 │         154.14 / 155.97 ±1.99 / 159.78 ms │                       160.85 / 162.05 ±1.04 / 163.54 ms │     no change │
│ QQuery 61 │            18.52 / 18.76 ±0.19 / 19.06 ms │                          20.47 / 23.25 ±4.86 / 32.94 ms │  1.24x slower │
│ QQuery 62 │         874.30 / 879.48 ±3.80 / 883.80 ms │                      868.38 / 919.09 ±41.03 / 992.29 ms │     no change │
│ QQuery 63 │         108.29 / 111.32 ±5.02 / 121.26 ms │                       107.73 / 111.71 ±4.80 / 121.12 ms │     no change │
│ QQuery 64 │         802.22 / 808.36 ±4.36 / 813.99 ms │                      849.98 / 868.70 ±12.56 / 886.93 ms │  1.07x slower │
│ QQuery 65 │         252.86 / 257.87 ±5.15 / 267.80 ms │                       252.40 / 256.26 ±6.33 / 268.90 ms │     no change │
│ QQuery 66 │        229.57 / 244.92 ±15.83 / 264.32 ms │                      231.68 / 246.19 ±11.33 / 258.93 ms │     no change │
│ QQuery 67 │        300.25 / 310.78 ±11.38 / 329.29 ms │                       307.34 / 316.35 ±9.83 / 335.02 ms │     no change │
│ QQuery 68 │            15.24 / 15.70 ±0.47 / 16.43 ms │                          14.75 / 15.20 ±0.74 / 16.66 ms │     no change │
│ QQuery 69 │         110.68 / 115.17 ±6.03 / 126.88 ms │                       106.64 / 109.93 ±4.23 / 118.28 ms │     no change │
│ QQuery 70 │        336.71 / 351.53 ±13.81 / 374.01 ms │                       328.43 / 339.25 ±6.31 / 344.51 ms │     no change │
│ QQuery 71 │         149.28 / 152.45 ±4.95 / 162.29 ms │                       143.25 / 145.22 ±3.24 / 151.62 ms │     no change │
│ QQuery 72 │         678.31 / 685.68 ±5.48 / 692.56 ms │                       582.43 / 592.45 ±6.79 / 601.57 ms │ +1.16x faster │
│ QQuery 73 │            13.80 / 13.96 ±0.10 / 14.12 ms │                          12.16 / 12.34 ±0.14 / 12.59 ms │ +1.13x faster │
│ QQuery 74 │        615.37 / 657.27 ±38.33 / 709.72 ms │                      566.56 / 619.37 ±49.26 / 680.73 ms │ +1.06x faster │
│ QQuery 75 │         314.34 / 316.92 ±3.23 / 323.28 ms │                       318.45 / 321.59 ±2.46 / 324.50 ms │     no change │
│ QQuery 76 │         168.88 / 172.74 ±3.73 / 177.98 ms │                       183.68 / 187.64 ±3.17 / 191.17 ms │  1.09x slower │
│ QQuery 77 │         211.02 / 215.85 ±4.58 / 221.49 ms │                       214.43 / 218.00 ±5.79 / 229.54 ms │     no change │
│ QQuery 78 │         422.73 / 431.60 ±6.74 / 442.86 ms │                      441.92 / 449.64 ±10.52 / 469.35 ms │     no change │
│ QQuery 79 │         267.50 / 270.45 ±2.15 / 273.47 ms │                      235.79 / 254.88 ±14.64 / 271.11 ms │ +1.06x faster │
│ QQuery 80 │         408.13 / 412.89 ±3.65 / 416.46 ms │                       405.21 / 409.29 ±3.24 / 413.94 ms │     no change │
│ QQuery 81 │            39.39 / 39.91 ±0.47 / 40.76 ms │                          35.84 / 37.39 ±1.44 / 40.11 ms │ +1.07x faster │
│ QQuery 82 │            48.71 / 52.08 ±5.82 / 63.70 ms │                          46.67 / 50.59 ±7.22 / 65.02 ms │     no change │
│ QQuery 83 │            67.04 / 68.31 ±1.16 / 70.20 ms │                          60.27 / 68.77 ±4.26 / 71.40 ms │     no change │
│ QQuery 84 │            51.50 / 51.91 ±0.27 / 52.28 ms │                          51.54 / 51.90 ±0.36 / 52.57 ms │     no change │
│ QQuery 85 │         166.24 / 169.24 ±3.53 / 175.86 ms │                       163.73 / 165.54 ±1.74 / 168.80 ms │     no change │
│ QQuery 86 │            44.06 / 44.83 ±0.67 / 45.68 ms │                          43.03 / 44.18 ±0.72 / 45.17 ms │     no change │
│ QQuery 87 │               3.72 / 3.83 ±0.15 / 4.12 ms │                             3.63 / 3.74 ±0.15 / 4.04 ms │     no change │
│ QQuery 88 │         119.90 / 124.71 ±5.02 / 134.37 ms │                       120.39 / 124.85 ±5.12 / 134.85 ms │     no change │
│ QQuery 89 │         124.33 / 125.45 ±0.88 / 126.91 ms │                       124.45 / 127.30 ±3.29 / 133.70 ms │     no change │
│ QQuery 90 │            30.22 / 32.62 ±4.06 / 40.72 ms │                          29.03 / 29.95 ±0.64 / 30.60 ms │ +1.09x faster │
│ QQuery 91 │            63.58 / 63.91 ±0.31 / 64.32 ms │                          63.91 / 65.12 ±0.79 / 66.02 ms │     no change │
│ QQuery 92 │            62.90 / 64.68 ±1.09 / 65.89 ms │                          63.55 / 64.56 ±0.73 / 65.55 ms │     no change │
│ QQuery 93 │         225.28 / 232.97 ±4.40 / 237.74 ms │                       227.93 / 230.86 ±2.22 / 233.93 ms │     no change │
│ QQuery 94 │            73.88 / 75.76 ±2.21 / 80.07 ms │                          70.69 / 73.22 ±1.36 / 74.26 ms │     no change │
│ QQuery 95 │         194.33 / 198.11 ±2.15 / 200.72 ms │                       180.76 / 183.61 ±1.83 / 186.45 ms │ +1.08x faster │
│ QQuery 96 │            75.75 / 76.64 ±0.81 / 78.15 ms │                          74.64 / 75.98 ±0.80 / 77.13 ms │     no change │
│ QQuery 97 │         135.21 / 142.28 ±5.12 / 149.26 ms │                       138.39 / 142.01 ±4.30 / 150.31 ms │     no change │
│ QQuery 98 │         158.43 / 162.57 ±3.29 / 168.17 ms │                       168.36 / 169.52 ±1.65 / 172.77 ms │     no change │
│ QQuery 99 │ 10762.08 / 10988.98 ±150.74 / 11187.23 ms │               10702.60 / 10921.40 ±162.49 / 11131.60 ms │     no change │
└───────────┴───────────────────────────────────────────┴─────────────────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                                                      ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                                                      │ 33773.60ms │
│ Total Time (gabotechs_repartition-coalesce-batches-berfore-channels)   │ 33915.99ms │
│ Average Time (HEAD)                                                    │   341.15ms │
│ Average Time (gabotechs_repartition-coalesce-batches-berfore-channels) │   342.59ms │
│ Queries Faster                                                         │         10 │
│ Queries Slower                                                         │          8 │
│ Queries with No Change                                                 │         81 │
│ Queries with Failure                                                   │          0 │
└────────────────────────────────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	170.0s
Peak memory	6.3 GiB
Avg memory	5.6 GiB
CPU user	351.5s
CPU sys	14.9s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	170.0s
Peak memory	6.2 GiB
Avg memory	5.6 GiB
CPU user	315.9s
CPU sys	12.7s
Peak spill	0 B

File an issue against this benchmark runner

gabotechs · 2026-05-04T19:43:38Z

run benchmarks

env:
  DATAFUSION_EXECUTION_TARGET_PARTITIONS: 256

adriangbot · 2026-05-04T19:45:52Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4373980023-2019-qhxhm 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-05-04T19:46:19Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4373980023-2020-4g4qq 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: tpcds
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-05-04T19:46:32Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4373980023-2021-bkfjb 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: tpch
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-05-04T19:57:55Z

🤖 Benchmark completed (GKE) | trigger

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

CPU Details (lscpu)

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

Details

Comparing HEAD and gabotechs_repartition-coalesce-batches-berfore-channels
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                               HEAD ┃ gabotechs_repartition-coalesce-batches-berfore-channels ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │     81.08 / 84.06 ±1.65 / 85.99 ms │                          80.95 / 83.92 ±1.91 / 86.82 ms │     no change │
│ QQuery 2  │     77.40 / 83.16 ±3.41 / 86.29 ms │                          82.28 / 84.42 ±1.29 / 85.65 ms │     no change │
│ QQuery 3  │ 380.72 / 404.84 ±21.00 / 439.20 ms │                      339.05 / 361.13 ±19.01 / 395.26 ms │ +1.12x faster │
│ QQuery 4  │     62.13 / 69.48 ±4.55 / 74.12 ms │                          67.85 / 68.46 ±0.54 / 69.47 ms │     no change │
│ QQuery 5  │ 641.00 / 692.54 ±34.21 / 734.60 ms │                      601.13 / 625.35 ±16.84 / 645.13 ms │ +1.11x faster │
│ QQuery 6  │     28.95 / 33.50 ±3.18 / 37.56 ms │                          28.30 / 29.87 ±1.28 / 32.15 ms │ +1.12x faster │
│ QQuery 7  │ 656.49 / 677.65 ±16.14 / 703.51 ms │                      616.89 / 648.76 ±25.71 / 685.81 ms │     no change │
│ QQuery 8  │ 633.58 / 665.74 ±23.16 / 699.52 ms │                      621.61 / 658.45 ±31.18 / 697.94 ms │     no change │
│ QQuery 9  │ 608.61 / 638.65 ±20.63 / 669.90 ms │                      577.22 / 594.91 ±16.49 / 626.07 ms │ +1.07x faster │
│ QQuery 10 │  119.28 / 128.39 ±6.38 / 138.74 ms │                       112.93 / 120.19 ±4.76 / 125.52 ms │ +1.07x faster │
│ QQuery 11 │     82.81 / 86.38 ±2.59 / 89.26 ms │                          78.55 / 85.89 ±3.91 / 89.20 ms │     no change │
│ QQuery 12 │ 661.09 / 685.12 ±16.92 / 711.09 ms │                      667.52 / 689.34 ±15.81 / 711.68 ms │     no change │
│ QQuery 13 │ 291.07 / 328.78 ±23.90 / 350.88 ms │                      287.39 / 315.76 ±19.41 / 343.61 ms │     no change │
│ QQuery 14 │     92.56 / 94.29 ±1.08 / 95.55 ms │                         84.78 / 92.73 ±5.40 / 101.78 ms │     no change │
│ QQuery 15 │  136.12 / 140.55 ±3.24 / 144.51 ms │                       132.89 / 134.90 ±1.55 / 137.02 ms │     no change │
│ QQuery 16 │  114.41 / 117.67 ±2.38 / 121.79 ms │                       100.97 / 102.85 ±1.24 / 104.60 ms │ +1.14x faster │
│ QQuery 17 │ 435.17 / 452.00 ±13.44 / 470.07 ms │                       330.08 / 338.10 ±6.29 / 345.87 ms │ +1.34x faster │
│ QQuery 18 │ 782.91 / 799.17 ±14.85 / 822.16 ms │                      473.85 / 494.66 ±18.38 / 519.04 ms │ +1.62x faster │
│ QQuery 19 │     56.13 / 57.18 ±0.97 / 58.70 ms │                          51.84 / 53.97 ±1.43 / 55.72 ms │ +1.06x faster │
│ QQuery 20 │ 371.21 / 407.23 ±25.75 / 432.51 ms │                      349.03 / 397.69 ±31.21 / 441.14 ms │     no change │
│ QQuery 21 │ 480.71 / 496.99 ±21.31 / 535.93 ms │                      415.58 / 447.98 ±18.79 / 473.48 ms │ +1.11x faster │
│ QQuery 22 │     71.43 / 76.58 ±5.98 / 87.80 ms │                          65.78 / 68.29 ±2.33 / 72.58 ms │ +1.12x faster │
└───────────┴────────────────────────────────────┴─────────────────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                                                      ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                                                      │ 7219.95ms │
│ Total Time (gabotechs_repartition-coalesce-batches-berfore-channels)   │ 6497.63ms │
│ Average Time (HEAD)                                                    │  328.18ms │
│ Average Time (gabotechs_repartition-coalesce-batches-berfore-channels) │  295.35ms │
│ Queries Faster                                                         │        11 │
│ Queries Slower                                                         │         0 │
│ Queries with No Change                                                 │        11 │
│ Queries with Failure                                                   │         0 │
└────────────────────────────────────────────────────────────────────────┴───────────┘

Resource Usage

tpch — base (merge-base)

Metric	Value
Wall time	40.0s
Peak memory	6.8 GiB
Avg memory	5.5 GiB
CPU user	312.4s
CPU sys	7.5s
Peak spill	0 B

tpch — branch

Metric	Value
Wall time	35.0s
Peak memory	6.7 GiB
Avg memory	5.4 GiB
CPU user	293.1s
CPU sys	6.9s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-05-04T20:01:57Z

🤖 Benchmark completed (GKE) | trigger

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

CPU Details (lscpu)

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

Details

Comparing HEAD and gabotechs_repartition-coalesce-batches-berfore-channels
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃ gabotechs_repartition-coalesce-batches-berfore-channels ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.56 / 5.11 ±6.88 / 18.87 ms │                            1.68 / 5.20 ±6.84 / 18.88 ms │     no change │
│ QQuery 1  │        23.35 / 25.87 ±1.81 / 28.40 ms │                          23.59 / 25.94 ±1.43 / 27.78 ms │     no change │
│ QQuery 2  │        49.25 / 51.11 ±1.24 / 52.88 ms │                          49.29 / 49.91 ±0.58 / 50.75 ms │     no change │
│ QQuery 3  │        39.42 / 41.33 ±1.66 / 43.82 ms │                          40.68 / 42.23 ±1.93 / 46.00 ms │     no change │
│ QQuery 4  │     365.86 / 373.34 ±3.93 / 377.23 ms │                       259.15 / 274.15 ±9.58 / 285.94 ms │ +1.36x faster │
│ QQuery 5  │     397.96 / 403.97 ±7.19 / 417.98 ms │                       335.46 / 338.66 ±2.94 / 342.29 ms │ +1.19x faster │
│ QQuery 6  │        22.21 / 23.65 ±0.83 / 24.65 ms │                          22.59 / 23.81 ±1.30 / 26.25 ms │     no change │
│ QQuery 7  │        55.15 / 55.83 ±0.70 / 57.07 ms │                          53.34 / 54.65 ±0.78 / 55.40 ms │     no change │
│ QQuery 8  │     521.18 / 523.30 ±2.43 / 526.99 ms │                       361.02 / 367.67 ±8.02 / 383.16 ms │ +1.42x faster │
│ QQuery 9  │     407.52 / 412.56 ±5.12 / 422.37 ms │                      414.74 / 434.76 ±14.19 / 453.97 ms │  1.05x slower │
│ QQuery 10 │     149.19 / 154.10 ±4.25 / 159.92 ms │                       147.57 / 149.96 ±1.52 / 151.64 ms │     no change │
│ QQuery 11 │     167.96 / 171.41 ±2.75 / 175.08 ms │                       161.24 / 165.93 ±4.23 / 172.86 ms │     no change │
│ QQuery 12 │     405.92 / 408.13 ±2.45 / 412.78 ms │                       326.62 / 338.26 ±8.20 / 350.41 ms │ +1.21x faster │
│ QQuery 13 │     783.07 / 795.02 ±8.96 / 807.76 ms │                      591.90 / 605.64 ±15.44 / 633.09 ms │ +1.31x faster │
│ QQuery 14 │     405.98 / 413.96 ±6.80 / 425.40 ms │                      316.61 / 357.19 ±34.76 / 406.19 ms │ +1.16x faster │
│ QQuery 15 │    434.68 / 457.14 ±26.78 / 508.16 ms │                       276.36 / 290.83 ±8.16 / 299.29 ms │ +1.57x faster │
│ QQuery 16 │    907.45 / 921.78 ±12.63 / 943.36 ms │                       642.44 / 650.28 ±6.93 / 660.33 ms │ +1.42x faster │
│ QQuery 17 │    911.76 / 927.91 ±18.36 / 961.73 ms │                      594.80 / 639.07 ±25.30 / 671.94 ms │ +1.45x faster │
│ QQuery 18 │ 1900.11 / 1944.04 ±40.29 / 2010.63 ms │                   1253.89 / 1289.05 ±30.06 / 1330.55 ms │ +1.51x faster │
│ QQuery 19 │      46.97 / 75.05 ±24.91 / 111.80 ms │                        41.61 / 82.03 ±50.33 / 178.66 ms │  1.09x slower │
│ QQuery 20 │    535.60 / 552.25 ±16.23 / 578.14 ms │                      536.65 / 573.16 ±23.15 / 598.92 ms │     no change │
│ QQuery 21 │    650.85 / 672.65 ±16.03 / 693.22 ms │                      641.17 / 672.74 ±28.75 / 718.88 ms │     no change │
│ QQuery 22 │ 1122.46 / 1156.24 ±18.77 / 1174.22 ms │                   1121.66 / 1151.89 ±26.10 / 1190.80 ms │     no change │
│ QQuery 23 │ 1720.66 / 1855.56 ±79.94 / 1936.98 ms │                  1697.27 / 1916.99 ±182.35 / 2207.42 ms │     no change │
│ QQuery 24 │        60.35 / 66.40 ±7.96 / 81.92 ms │                         57.37 / 66.78 ±13.64 / 93.62 ms │     no change │
│ QQuery 25 │     125.20 / 129.97 ±3.20 / 134.53 ms │                       127.67 / 133.34 ±5.83 / 144.01 ms │     no change │
│ QQuery 26 │        56.41 / 61.39 ±3.96 / 68.26 ms │                          59.04 / 61.03 ±1.55 / 62.99 ms │     no change │
│ QQuery 27 │    695.49 / 722.54 ±19.02 / 745.35 ms │                      700.60 / 720.31 ±11.35 / 732.66 ms │     no change │
│ QQuery 28 │ 3017.92 / 3125.77 ±69.70 / 3193.06 ms │                   3058.82 / 3112.94 ±53.75 / 3208.85 ms │     no change │
│ QQuery 29 │        57.17 / 60.34 ±1.96 / 62.03 ms │                        57.97 / 79.78 ±17.52 / 100.45 ms │  1.32x slower │
│ QQuery 30 │    437.78 / 450.69 ±13.76 / 476.57 ms │                       354.71 / 361.95 ±4.50 / 367.09 ms │ +1.25x faster │
│ QQuery 31 │    580.55 / 596.01 ±10.81 / 609.36 ms │                      424.66 / 444.21 ±11.79 / 454.99 ms │ +1.34x faster │
│ QQuery 32 │ 2445.76 / 2517.42 ±50.26 / 2592.13 ms │                   1389.33 / 1448.60 ±43.31 / 1516.73 ms │ +1.74x faster │
│ QQuery 33 │ 1829.39 / 1875.24 ±26.35 / 1906.43 ms │                   1620.21 / 1652.28 ±34.16 / 1699.87 ms │ +1.13x faster │
│ QQuery 34 │ 1881.78 / 1935.42 ±48.50 / 2022.17 ms │                   1606.08 / 1648.12 ±36.77 / 1715.81 ms │ +1.17x faster │
│ QQuery 35 │    413.55 / 429.55 ±12.10 / 447.04 ms │                      270.16 / 321.21 ±55.02 / 414.72 ms │ +1.34x faster │
│ QQuery 36 │     119.52 / 122.34 ±1.54 / 124.21 ms │                       111.93 / 118.17 ±4.50 / 122.84 ms │     no change │
│ QQuery 37 │        76.94 / 80.36 ±3.63 / 87.24 ms │                          75.28 / 79.74 ±3.69 / 85.71 ms │     no change │
│ QQuery 38 │        83.20 / 87.97 ±2.74 / 90.70 ms │                          79.04 / 82.69 ±2.08 / 85.05 ms │ +1.06x faster │
│ QQuery 39 │     197.62 / 206.14 ±9.17 / 219.14 ms │                       206.62 / 220.14 ±9.68 / 233.49 ms │  1.07x slower │
│ QQuery 40 │        56.43 / 63.67 ±7.47 / 77.63 ms │                          55.25 / 60.67 ±5.94 / 71.03 ms │     no change │
│ QQuery 41 │        54.63 / 55.48 ±0.84 / 57.02 ms │                          55.17 / 58.12 ±4.86 / 67.80 ms │     no change │
│ QQuery 42 │        53.69 / 57.95 ±3.47 / 62.06 ms │                          55.37 / 57.77 ±1.51 / 60.08 ms │     no change │
└───────────┴───────────────────────────────────────┴─────────────────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                                                      ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                                                      │ 25065.94ms │
│ Total Time (gabotechs_repartition-coalesce-batches-berfore-channels)   │ 21227.86ms │
│ Average Time (HEAD)                                                    │   582.93ms │
│ Average Time (gabotechs_repartition-coalesce-batches-berfore-channels) │   493.67ms │
│ Queries Faster                                                         │         17 │
│ Queries Slower                                                         │          4 │
│ Queries with No Change                                                 │         22 │
│ Queries with Failure                                                   │          0 │
└────────────────────────────────────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	130.0s
Peak memory	34.6 GiB
Avg memory	27.4 GiB
CPU user	1298.6s
CPU sys	92.9s
Peak spill	0 B

clickbench_partitioned — branch

Metric	Value
Wall time	110.0s
Peak memory	34.7 GiB
Avg memory	27.7 GiB
CPU user	1052.4s
CPU sys	90.7s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-05-04T20:06:52Z

🤖 Benchmark completed (GKE) | trigger

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

CPU Details (lscpu)

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

Details

Comparing HEAD and gabotechs_repartition-coalesce-batches-berfore-channels
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                      HEAD ┃ gabotechs_repartition-coalesce-batches-berfore-channels ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │            59.89 / 60.88 ±0.72 / 62.06 ms │                          57.67 / 58.72 ±0.65 / 59.66 ms │     no change │
│ QQuery 2  │         287.90 / 294.59 ±7.63 / 309.50 ms │                       293.62 / 298.18 ±3.69 / 302.38 ms │     no change │
│ QQuery 3  │         153.80 / 159.23 ±4.70 / 166.20 ms │                       152.32 / 158.94 ±4.31 / 163.80 ms │     no change │
│ QQuery 4  │     1584.03 / 1634.24 ±32.13 / 1675.55 ms │                   1577.42 / 1608.11 ±39.11 / 1679.48 ms │     no change │
│ QQuery 5  │        629.67 / 645.17 ±12.01 / 664.54 ms │                      622.07 / 671.39 ±29.87 / 705.61 ms │     no change │
│ QQuery 6  │        270.48 / 279.48 ±11.42 / 301.90 ms │                       270.68 / 278.75 ±5.46 / 284.94 ms │     no change │
│ QQuery 7  │      1226.87 / 1237.03 ±9.32 / 1252.20 ms │                   1221.31 / 1254.15 ±25.33 / 1294.88 ms │     no change │
│ QQuery 8  │         221.33 / 228.10 ±4.98 / 236.47 ms │                       208.70 / 220.59 ±6.84 / 229.31 ms │     no change │
│ QQuery 9  │        276.35 / 289.40 ±12.16 / 310.43 ms │                      268.77 / 288.58 ±21.61 / 330.69 ms │     no change │
│ QQuery 10 │         170.87 / 172.98 ±2.50 / 177.70 ms │                       158.84 / 170.55 ±6.10 / 176.61 ms │     no change │
│ QQuery 11 │     1111.47 / 1144.32 ±38.99 / 1197.94 ms │                   1089.10 / 1132.32 ±40.07 / 1188.62 ms │     no change │
│ QQuery 12 │         114.87 / 117.21 ±1.80 / 119.88 ms │                       115.88 / 120.01 ±4.10 / 126.97 ms │     no change │
│ QQuery 13 │         414.98 / 427.31 ±9.16 / 441.10 ms │                       415.47 / 422.04 ±3.64 / 426.45 ms │     no change │
│ QQuery 14 │     1518.98 / 1551.46 ±18.57 / 1575.32 ms │                   1533.70 / 1555.05 ±21.90 / 1593.12 ms │     no change │
│ QQuery 15 │        300.81 / 332.93 ±17.10 / 347.63 ms │                      301.26 / 324.30 ±18.47 / 344.00 ms │     no change │
│ QQuery 16 │            56.20 / 62.95 ±5.33 / 70.31 ms │                          63.64 / 68.39 ±3.42 / 73.95 ms │  1.09x slower │
│ QQuery 17 │     1190.34 / 1235.97 ±49.68 / 1323.06 ms │                   1184.02 / 1214.22 ±25.01 / 1260.05 ms │     no change │
│ QQuery 18 │        511.92 / 544.38 ±19.74 / 565.18 ms │                      519.01 / 550.72 ±27.28 / 592.76 ms │     no change │
│ QQuery 19 │         192.63 / 199.65 ±6.06 / 207.00 ms │                      202.79 / 213.03 ±11.67 / 235.46 ms │  1.07x slower │
│ QQuery 20 │           89.78 / 94.92 ±7.37 / 109.38 ms │                        96.60 / 101.30 ±2.92 / 105.71 ms │  1.07x slower │
│ QQuery 21 │            61.45 / 62.70 ±1.44 / 65.47 ms │                          68.36 / 71.87 ±3.02 / 75.43 ms │  1.15x slower │
│ QQuery 22 │         530.14 / 534.36 ±3.88 / 540.81 ms │                      537.43 / 549.10 ±10.44 / 561.77 ms │     no change │
│ QQuery 23 │     3727.18 / 3842.12 ±83.52 / 3960.75 ms │                   3509.55 / 3617.45 ±66.65 / 3713.93 ms │ +1.06x faster │
│ QQuery 24 │     2988.73 / 3068.66 ±52.31 / 3129.02 ms │                   2919.73 / 2945.93 ±30.63 / 3005.62 ms │     no change │
│ QQuery 25 │     1373.32 / 1409.02 ±25.15 / 1436.82 ms │                   1298.53 / 1331.25 ±26.91 / 1380.03 ms │ +1.06x faster │
│ QQuery 26 │         733.74 / 737.35 ±3.11 / 742.69 ms │                      695.63 / 718.96 ±20.02 / 744.62 ms │     no change │
│ QQuery 27 │            68.77 / 70.46 ±1.43 / 72.77 ms │                          70.87 / 71.93 ±0.76 / 72.63 ms │     no change │
│ QQuery 28 │         200.67 / 206.02 ±4.20 / 213.04 ms │                       196.30 / 207.24 ±8.43 / 215.39 ms │     no change │
│ QQuery 29 │     1319.35 / 1360.79 ±23.10 / 1390.90 ms │                   1306.85 / 1333.69 ±27.23 / 1381.65 ms │     no change │
│ QQuery 30 │         153.52 / 162.34 ±5.38 / 167.51 ms │                       160.03 / 163.18 ±1.57 / 163.99 ms │     no change │
│ QQuery 31 │         402.88 / 411.32 ±8.53 / 423.31 ms │                       421.61 / 431.59 ±9.59 / 449.06 ms │     no change │
│ QQuery 32 │            69.43 / 72.49 ±1.86 / 74.50 ms │                          74.26 / 75.20 ±1.03 / 76.94 ms │     no change │
│ QQuery 33 │         204.53 / 208.85 ±3.00 / 213.54 ms │                       213.32 / 216.55 ±4.59 / 225.55 ms │     no change │
│ QQuery 34 │            54.81 / 57.11 ±2.29 / 60.94 ms │                          58.56 / 62.60 ±5.27 / 72.70 ms │  1.10x slower │
│ QQuery 35 │         170.45 / 176.27 ±5.09 / 185.06 ms │                      156.97 / 170.63 ±12.36 / 191.76 ms │     no change │
│ QQuery 36 │          79.18 / 88.56 ±15.25 / 118.92 ms │                          80.71 / 84.45 ±2.69 / 88.81 ms │     no change │
│ QQuery 37 │            43.54 / 44.61 ±1.03 / 46.41 ms │                          44.29 / 45.91 ±1.47 / 48.33 ms │     no change │
│ QQuery 38 │        196.21 / 209.05 ±10.88 / 228.34 ms │                       192.63 / 203.05 ±7.19 / 213.78 ms │     no change │
│ QQuery 39 │        444.39 / 458.43 ±14.58 / 484.58 ms │                       460.17 / 470.19 ±5.16 / 474.81 ms │     no change │
│ QQuery 40 │        302.86 / 315.45 ±12.00 / 335.72 ms │                      257.04 / 271.97 ±10.01 / 285.36 ms │ +1.16x faster │
│ QQuery 41 │            87.12 / 90.21 ±2.60 / 93.57 ms │                          88.76 / 90.11 ±0.92 / 91.44 ms │     no change │
│ QQuery 42 │         143.68 / 150.38 ±5.72 / 159.25 ms │                       145.20 / 153.89 ±6.64 / 162.47 ms │     no change │
│ QQuery 43 │            47.70 / 47.98 ±0.28 / 48.45 ms │                          48.12 / 48.87 ±0.54 / 49.59 ms │     no change │
│ QQuery 44 │         138.84 / 145.21 ±5.55 / 154.50 ms │                       141.96 / 148.43 ±4.79 / 154.11 ms │     no change │
│ QQuery 45 │        372.77 / 399.33 ±21.27 / 431.64 ms │                       359.72 / 367.56 ±4.56 / 373.09 ms │ +1.09x faster │
│ QQuery 46 │            63.85 / 65.73 ±1.33 / 66.93 ms │                          62.23 / 64.09 ±1.23 / 65.52 ms │     no change │
│ QQuery 47 │      983.86 / 1053.01 ±40.81 / 1095.64 ms │                       978.31 / 988.67 ±6.28 / 996.08 ms │ +1.07x faster │
│ QQuery 48 │        309.82 / 320.98 ±12.14 / 341.86 ms │                       307.55 / 313.00 ±3.87 / 317.74 ms │     no change │
│ QQuery 49 │        568.16 / 593.58 ±16.52 / 613.49 ms │                       579.77 / 586.39 ±3.58 / 590.56 ms │     no change │
│ QQuery 50 │        819.00 / 840.04 ±11.72 / 854.10 ms │                      767.50 / 780.94 ±13.69 / 803.10 ms │ +1.08x faster │
│ QQuery 51 │         417.16 / 424.46 ±4.39 / 430.69 ms │                       391.70 / 398.37 ±8.24 / 414.33 ms │ +1.07x faster │
│ QQuery 52 │         147.31 / 154.71 ±6.29 / 162.34 ms │                       144.96 / 154.65 ±5.81 / 161.89 ms │     no change │
│ QQuery 53 │         177.13 / 179.01 ±2.61 / 184.12 ms │                       176.21 / 181.37 ±4.32 / 186.54 ms │     no change │
│ QQuery 54 │         345.14 / 349.29 ±4.53 / 357.12 ms │                       352.89 / 357.85 ±3.74 / 363.81 ms │     no change │
│ QQuery 55 │         146.35 / 149.01 ±3.06 / 154.58 ms │                       144.99 / 150.64 ±5.42 / 160.78 ms │     no change │
│ QQuery 56 │         205.43 / 209.29 ±2.74 / 212.61 ms │                       206.70 / 211.01 ±2.55 / 213.85 ms │     no change │
│ QQuery 57 │         419.91 / 428.02 ±7.54 / 441.09 ms │                       409.50 / 416.17 ±7.01 / 425.66 ms │     no change │
│ QQuery 58 │     1790.37 / 1803.50 ±20.16 / 1842.91 ms │                   1812.42 / 1831.01 ±11.02 / 1843.16 ms │     no change │
│ QQuery 59 │         281.46 / 292.89 ±8.33 / 303.95 ms │                       274.26 / 283.68 ±5.96 / 292.14 ms │     no change │
│ QQuery 60 │         207.20 / 216.56 ±7.99 / 230.65 ms │                       210.87 / 215.33 ±5.68 / 226.39 ms │     no change │
│ QQuery 61 │            37.18 / 39.37 ±2.14 / 42.89 ms │                          39.46 / 40.65 ±1.23 / 42.53 ms │     no change │
│ QQuery 62 │        936.90 / 957.05 ±13.95 / 978.68 ms │                      961.88 / 983.33 ±11.44 / 994.99 ms │     no change │
│ QQuery 63 │         179.10 / 185.20 ±6.77 / 198.28 ms │                       181.87 / 187.59 ±5.49 / 196.15 ms │     no change │
│ QQuery 64 │    2225.57 / 2353.22 ±149.43 / 2636.00 ms │                  2233.75 / 2370.45 ±107.76 / 2479.95 ms │     no change │
│ QQuery 65 │        364.92 / 378.15 ±14.96 / 405.16 ms │                      376.66 / 391.47 ±11.70 / 403.41 ms │     no change │
│ QQuery 66 │        318.15 / 341.98 ±17.06 / 365.22 ms │                       353.97 / 361.03 ±7.62 / 375.71 ms │  1.06x slower │
│ QQuery 67 │        356.54 / 375.45 ±15.19 / 401.66 ms │                       410.71 / 421.40 ±8.78 / 437.46 ms │  1.12x slower │
│ QQuery 68 │            65.67 / 71.47 ±5.79 / 78.57 ms │                          63.36 / 68.14 ±4.40 / 73.70 ms │     no change │
│ QQuery 69 │         163.78 / 169.70 ±6.03 / 178.65 ms │                       157.83 / 170.76 ±6.99 / 177.85 ms │     no change │
│ QQuery 70 │        459.61 / 486.97 ±13.96 / 498.89 ms │                      444.24 / 465.41 ±11.86 / 480.88 ms │     no change │
│ QQuery 71 │        185.76 / 198.92 ±10.18 / 216.77 ms │                       191.81 / 202.18 ±7.13 / 211.73 ms │     no change │
│ QQuery 72 │     5324.13 / 5369.91 ±34.64 / 5404.88 ms │                   5240.96 / 5308.18 ±40.10 / 5366.41 ms │     no change │
│ QQuery 73 │            58.61 / 68.53 ±9.50 / 85.76 ms │                          63.43 / 69.26 ±6.93 / 82.66 ms │     no change │
│ QQuery 74 │        743.87 / 805.02 ±47.12 / 877.56 ms │                      822.44 / 845.60 ±18.05 / 877.68 ms │  1.05x slower │
│ QQuery 75 │         632.19 / 641.11 ±6.40 / 651.87 ms │                       614.27 / 627.09 ±8.06 / 636.09 ms │     no change │
│ QQuery 76 │        711.96 / 874.20 ±88.08 / 945.14 ms │                    884.33 / 1028.80 ±89.95 / 1109.34 ms │  1.18x slower │
│ QQuery 77 │        391.18 / 439.09 ±28.44 / 474.44 ms │                      391.04 / 420.13 ±33.55 / 479.56 ms │     no change │
│ QQuery 78 │        900.81 / 916.61 ±13.85 / 936.65 ms │                      896.57 / 914.07 ±11.18 / 927.57 ms │     no change │
│ QQuery 79 │         286.28 / 298.06 ±9.66 / 314.96 ms │                       307.88 / 320.85 ±9.83 / 337.93 ms │  1.08x slower │
│ QQuery 80 │        836.43 / 856.13 ±19.27 / 879.60 ms │                      832.44 / 880.40 ±35.44 / 942.70 ms │     no change │
│ QQuery 81 │         140.46 / 149.48 ±9.20 / 164.14 ms │                      145.11 / 158.20 ±14.91 / 186.50 ms │  1.06x slower │
│ QQuery 82 │           83.45 / 90.19 ±9.86 / 109.72 ms │                          85.38 / 89.10 ±4.61 / 97.93 ms │     no change │
│ QQuery 83 │         616.56 / 630.55 ±9.53 / 641.25 ms │                       648.37 / 662.50 ±9.64 / 672.50 ms │  1.05x slower │
│ QQuery 84 │            76.57 / 79.40 ±3.68 / 86.49 ms │                          77.45 / 79.18 ±1.14 / 80.43 ms │     no change │
│ QQuery 85 │        451.63 / 488.93 ±25.97 / 532.43 ms │                      458.52 / 485.75 ±14.77 / 497.00 ms │     no change │
│ QQuery 86 │            92.82 / 94.58 ±1.88 / 97.99 ms │                         91.86 / 96.49 ±4.54 / 104.55 ms │     no change │
│ QQuery 87 │               3.86 / 3.98 ±0.13 / 4.23 ms │                             3.72 / 3.93 ±0.18 / 4.22 ms │     no change │
│ QQuery 88 │        207.16 / 217.29 ±10.14 / 234.88 ms │                       208.44 / 222.32 ±9.56 / 237.98 ms │     no change │
│ QQuery 89 │         201.47 / 209.53 ±5.95 / 219.66 ms │                       205.94 / 212.30 ±9.84 / 231.82 ms │     no change │
│ QQuery 90 │            57.36 / 60.62 ±3.19 / 66.12 ms │                          61.35 / 64.48 ±2.78 / 69.13 ms │  1.06x slower │
│ QQuery 91 │         107.52 / 109.68 ±1.35 / 111.20 ms │                       111.20 / 114.28 ±2.92 / 118.60 ms │     no change │
│ QQuery 92 │         112.61 / 117.48 ±3.09 / 121.08 ms │                       115.18 / 119.41 ±3.13 / 123.07 ms │     no change │
│ QQuery 93 │         514.12 / 521.99 ±5.10 / 529.46 ms │                       499.69 / 504.78 ±5.37 / 515.07 ms │     no change │
│ QQuery 94 │         136.70 / 139.67 ±2.07 / 141.98 ms │                       133.13 / 140.46 ±7.32 / 152.33 ms │     no change │
│ QQuery 95 │        804.86 / 853.73 ±37.08 / 906.07 ms │                      772.42 / 806.01 ±33.08 / 848.53 ms │ +1.06x faster │
│ QQuery 96 │           91.50 / 98.14 ±6.64 / 110.72 ms │                          94.21 / 96.39 ±1.97 / 98.97 ms │     no change │
│ QQuery 97 │         252.07 / 260.75 ±7.74 / 273.53 ms │                      237.74 / 247.91 ±12.42 / 272.10 ms │     no change │
│ QQuery 98 │         228.35 / 234.53 ±5.19 / 242.23 ms │                       234.71 / 239.54 ±5.19 / 246.45 ms │     no change │
│ QQuery 99 │ 11097.82 / 11473.16 ±219.00 / 11681.29 ms │               10982.87 / 11462.30 ±397.04 / 12127.31 ms │     no change │
└───────────┴───────────────────────────────────────────┴─────────────────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                                                      ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                                                      │ 63781.15ms │
│ Total Time (gabotechs_repartition-coalesce-batches-berfore-channels)   │ 63376.22ms │
│ Average Time (HEAD)                                                    │   644.25ms │
│ Average Time (gabotechs_repartition-coalesce-batches-berfore-channels) │   640.16ms │
│ Queries Faster                                                         │          8 │
│ Queries Slower                                                         │         13 │
│ Queries with No Change                                                 │         78 │
│ Queries with Failure                                                   │          0 │
└────────────────────────────────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	320.1s
Peak memory	11.9 GiB
Avg memory	6.1 GiB
CPU user	1551.3s
CPU sys	52.5s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	320.1s
Peak memory	12.4 GiB
Avg memory	6.3 GiB
CPU user	1430.2s
CPU sys	49.6s
Peak spill	0 B

File an issue against this benchmark runner

Dandandan · 2026-05-05T05:58:53Z

run benchmark tpch10

env:
  DATAFUSION_EXECUTION_TARGET_PARTITIONS: 256

Dandandan · 2026-05-05T06:00:40Z

+/// inside `inner` holds cloned `Arc`s pointing back at those shared
+/// resources.
+struct OutputChannels {
+    inner: HashMap<usize, OutputChannel>,


Probably no perf change but I think Vec<Option<OutputChannel>> is better

adriangbot · 2026-05-05T06:03:09Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4376856575-2023-d9wvg 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: tpch10
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-05-05T06:17:28Z

🤖 Benchmark completed (GKE) | trigger

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

CPU Details (lscpu)

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

Details

Comparing HEAD and gabotechs_repartition-coalesce-batches-berfore-channels
--------------------
Benchmark tpch_sf10.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃ gabotechs_repartition-coalesce-batches-berfore-channels ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │     367.67 / 370.56 ±2.75 / 374.78 ms │                       365.18 / 373.67 ±4.77 / 379.99 ms │     no change │
│ QQuery 2  │ 1004.79 / 1044.73 ±25.91 / 1077.76 ms │                      911.89 / 929.35 ±13.91 / 949.82 ms │ +1.12x faster │
│ QQuery 3  │ 1449.50 / 1482.28 ±26.02 / 1515.45 ms │                   1020.76 / 1037.48 ±17.38 / 1068.95 ms │ +1.43x faster │
│ QQuery 4  │    723.52 / 737.25 ±10.02 / 751.24 ms │                      434.60 / 469.21 ±19.98 / 496.75 ms │ +1.57x faster │
│ QQuery 5  │ 1761.69 / 1789.09 ±27.18 / 1828.80 ms │                   1226.71 / 1252.80 ±14.01 / 1265.94 ms │ +1.43x faster │
│ QQuery 6  │     144.57 / 145.98 ±2.30 / 150.51 ms │                       143.31 / 143.63 ±0.25 / 143.98 ms │     no change │
│ QQuery 7  │ 1974.80 / 2092.05 ±68.11 / 2169.52 ms │                   1449.65 / 1485.49 ±25.19 / 1515.01 ms │ +1.41x faster │
│ QQuery 8  │ 2244.63 / 2306.15 ±35.70 / 2345.41 ms │                   1728.81 / 1790.63 ±34.56 / 1828.12 ms │ +1.29x faster │
│ QQuery 9  │ 2464.34 / 2498.28 ±19.34 / 2518.71 ms │                   1684.52 / 1727.57 ±52.17 / 1821.02 ms │ +1.45x faster │
│ QQuery 10 │  1178.65 / 1188.35 ±7.41 / 1199.86 ms │                   1032.39 / 1063.70 ±28.55 / 1108.34 ms │ +1.12x faster │
│ QQuery 11 │    702.30 / 724.13 ±17.12 / 746.29 ms │                      592.85 / 617.01 ±24.95 / 657.93 ms │ +1.17x faster │
│ QQuery 12 │    625.06 / 645.32 ±16.86 / 664.50 ms │                      570.42 / 586.23 ±19.29 / 621.02 ms │ +1.10x faster │
│ QQuery 13 │    603.10 / 632.75 ±15.98 / 651.15 ms │                      528.48 / 545.38 ±10.40 / 561.09 ms │ +1.16x faster │
│ QQuery 14 │    484.56 / 503.27 ±11.11 / 515.79 ms │                      436.17 / 454.68 ±13.84 / 469.21 ms │ +1.11x faster │
│ QQuery 15 │     436.02 / 439.40 ±1.99 / 442.13 ms │                       418.67 / 420.91 ±2.24 / 425.06 ms │     no change │
│ QQuery 16 │    485.73 / 511.92 ±25.38 / 554.75 ms │                      428.91 / 447.49 ±13.36 / 470.72 ms │ +1.14x faster │
│ QQuery 17 │ 3045.31 / 3069.78 ±13.38 / 3085.75 ms │                   2151.55 / 2189.00 ±32.42 / 2242.67 ms │ +1.40x faster │
│ QQuery 18 │ 2469.94 / 2514.16 ±34.27 / 2549.66 ms │                    1490.67 / 1497.52 ±5.59 / 1504.71 ms │ +1.68x faster │
│ QQuery 19 │    669.29 / 695.50 ±20.46 / 728.66 ms │                      659.63 / 697.76 ±20.50 / 720.99 ms │     no change │
│ QQuery 20 │ 1457.36 / 1496.40 ±31.64 / 1552.13 ms │                   1392.10 / 1418.72 ±25.38 / 1449.70 ms │ +1.05x faster │
│ QQuery 21 │ 3155.81 / 3196.69 ±25.12 / 3230.72 ms │                   1917.15 / 1967.47 ±35.49 / 2026.96 ms │ +1.62x faster │
│ QQuery 22 │    463.16 / 487.37 ±22.66 / 524.17 ms │                      371.26 / 406.72 ±20.05 / 432.38 ms │ +1.20x faster │
└───────────┴───────────────────────────────────────┴─────────────────────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                                                      ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                                                      │ 28571.41ms │
│ Total Time (gabotechs_repartition-coalesce-batches-berfore-channels)   │ 21522.43ms │
│ Average Time (HEAD)                                                    │  1298.70ms │
│ Average Time (gabotechs_repartition-coalesce-batches-berfore-channels) │   978.29ms │
│ Queries Faster                                                         │         18 │
│ Queries Slower                                                         │          0 │
│ Queries with No Change                                                 │          4 │
│ Queries with Failure                                                   │          0 │
└────────────────────────────────────────────────────────────────────────┴────────────┘

Resource Usage

tpch10 — base (merge-base)

Metric	Value
Wall time	145.0s
Peak memory	14.1 GiB
Avg memory	11.1 GiB
CPU user	1507.0s
CPU sys	38.3s
Peak spill	0 B

tpch10 — branch

Metric	Value
Wall time	110.0s
Peak memory	20.4 GiB
Avg memory	11.5 GiB
CPU user	1102.2s
CPU sys	37.5s
Peak spill	0 B

File an issue against this benchmark runner

Samyak2 · 2026-05-05T06:33:14Z

This should fix #20491 too!

gabotechs · 2026-05-05T06:34:46Z

Ok, as it looks like it does what was intended, I'll open this for review and improve the PR description.

Dandandan · 2026-05-05T18:13:32Z

run benchmark tpch10

env:
  DATAFUSION_EXECUTION_TARGET_PARTITIONS: 2048

adriangbot · 2026-05-05T18:16:52Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4381836993-2030-67qvv 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: tpch10
Results will be posted here when complete

File an issue against this benchmark runner

Dandandan · 2026-05-05T19:49:15Z

run benchmark tpch10

env:
  DATAFUSION_EXECUTION_TARGET_PARTITIONS: 512

adriangbot · 2026-05-05T19:52:29Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4382478164-2031-w9s9p 6.12.68+ #1 SMP Wed Apr 1 02:23:28 UTC 2026 aarch64 GNU/Linux

CPU Details (lscpu)

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

Comparing gabotechs/repartition-coalesce-batches-berfore-channels (dcf6482) to b2fd2d3 (merge-base) diff using: tpch10
Results will be posted here when complete

File an issue against this benchmark runner

gabotechs added 3 commits May 4, 2026 18:16

Coalesce batches in RepartitionExec before sending them to the distri…

c533520

…butor channel, instead of after.

Adapt test for new coalescing behavior

21bc1a8

Remove clippy expectation

dcf6482

github-actions Bot added the physical-plan Changes to the physical-plan crate label May 4, 2026

Dandandan reviewed May 5, 2026

View reviewed changes

gabotechs marked this pull request as ready for review May 5, 2026 06:34

gabotechs added the performance Make DataFusion faster label May 5, 2026

Conversation

gabotechs commented May 4, 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?

Are these changes tested?

Are there any user-facing changes?

Uh oh!

gabotechs commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

gabotechs commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

gabotechs commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

adriangbot commented May 4, 2026

Uh oh!

Dandandan commented May 5, 2026

Uh oh!

Dandandan May 5, 2026

Choose a reason for hiding this comment

Uh oh!

adriangbot commented May 5, 2026

Uh oh!

adriangbot commented May 5, 2026

Uh oh!

Samyak2 commented May 5, 2026

Uh oh!

gabotechs commented May 5, 2026

Uh oh!

Dandandan commented May 5, 2026

Uh oh!

adriangbot commented May 5, 2026

Uh oh!

Dandandan commented May 5, 2026

Uh oh!

adriangbot commented May 5, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

gabotechs commented May 4, 2026 •

edited

Loading