Defer task spawning in SortPreservingMergeExec to first poll by Dandandan · Pull Request #21328 · apache/datafusion

Dandandan · 2026-04-02T19:28:18Z

Which issue does this PR close?

N/A

Rationale for this change

SortPreservingMergeExec::execute() eagerly executes all input partitions and spawns buffered tasks immediately, even before any batch is polled from the output stream. This wastes resources when streams are created but never consumed (e.g., query cancelled before first poll), and creates an unnecessary burst of concurrent tasks.

What changes are included in this PR?

Introduces a LazySortPreservingMergeStream that defers executing input partitions, spawning buffered tasks, and building the streaming merge until the first poll_next() call. The single-partition and zero-partition paths are unchanged.

How are these changes tested?

All existing tests pass (17/17), covering:

Multi-partition merge with various overlap patterns
Drop cancellation
Fetch limits
Congestion handling
Stable sort / round-robin tie breaking
Metrics

🤖 Generated with Claude Code

Previously, SortPreservingMergeExec eagerly executed all input partitions and spawned buffered tasks in execute(). This meant that even if the output stream was never polled, all tasks would be spawned. This changes the multi-partition path to use a lazy stream that defers spawning and building the streaming merge until first poll_next(). Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Dandandan · 2026-04-02T19:29:11Z

run benchmarks

adriangbot · 2026-04-02T19:29:42Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4180026900-670-wmbtx 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 lazy-sort-preserving-merge (6eff550) to 4084a18 (merge-base) diff using: tpcds
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-02T19:31:22Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4180026900-671-cf8zj 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 lazy-sort-preserving-merge (6eff550) to 4084a18 (merge-base) diff using: tpch
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-02T19:31:26Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4180026900-669-n59rl 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 lazy-sort-preserving-merge (6eff550) to 4084a18 (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-02T19:41:22Z

🤖 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 lazy-sort-preserving-merge
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Query     ┃                           HEAD ┃     lazy-sort-preserving-merge ┃    Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ QQuery 1  │ 45.64 / 46.21 ±0.73 / 47.65 ms │ 45.51 / 46.17 ±0.85 / 47.83 ms │ no change │
│ QQuery 2  │ 21.26 / 21.68 ±0.43 / 22.50 ms │ 21.32 / 21.54 ±0.15 / 21.74 ms │ no change │
│ QQuery 3  │ 31.86 / 32.17 ±0.17 / 32.35 ms │ 31.97 / 32.81 ±1.05 / 34.85 ms │ no change │
│ QQuery 4  │ 20.66 / 21.58 ±0.69 / 22.57 ms │ 20.57 / 21.69 ±1.10 / 23.21 ms │ no change │
│ QQuery 5  │ 48.61 / 50.56 ±1.27 / 51.75 ms │ 48.18 / 49.26 ±1.26 / 51.68 ms │ no change │
│ QQuery 6  │ 17.40 / 17.61 ±0.16 / 17.88 ms │ 17.23 / 17.69 ±0.37 / 18.37 ms │ no change │
│ QQuery 7  │ 54.42 / 55.97 ±1.02 / 57.47 ms │ 54.33 / 56.74 ±1.57 / 59.17 ms │ no change │
│ QQuery 8  │ 48.83 / 49.03 ±0.18 / 49.36 ms │ 48.64 / 48.87 ±0.17 / 49.14 ms │ no change │
│ QQuery 9  │ 53.92 / 55.48 ±0.95 / 56.84 ms │ 54.97 / 55.26 ±0.37 / 55.87 ms │ no change │
│ QQuery 10 │ 72.46 / 73.76 ±1.21 / 76.03 ms │ 70.77 / 73.62 ±1.77 / 76.31 ms │ no change │
│ QQuery 11 │ 14.25 / 14.42 ±0.17 / 14.67 ms │ 14.30 / 14.84 ±0.57 / 15.84 ms │ no change │
│ QQuery 12 │ 28.18 / 28.33 ±0.09 / 28.43 ms │ 28.02 / 28.41 ±0.31 / 28.92 ms │ no change │
│ QQuery 13 │ 39.36 / 40.09 ±0.57 / 40.74 ms │ 38.61 / 39.47 ±0.75 / 40.65 ms │ no change │
│ QQuery 14 │ 28.61 / 29.13 ±0.40 / 29.84 ms │ 28.85 / 29.10 ±0.19 / 29.36 ms │ no change │
│ QQuery 15 │ 33.94 / 34.36 ±0.40 / 35.07 ms │ 33.96 / 34.82 ±0.57 / 35.49 ms │ no change │
│ QQuery 16 │ 16.46 / 16.79 ±0.40 / 17.54 ms │ 15.93 / 16.54 ±0.35 / 16.98 ms │ no change │
│ QQuery 17 │ 74.11 / 74.76 ±0.68 / 75.98 ms │ 73.41 / 74.18 ±0.53 / 75.08 ms │ no change │
│ QQuery 18 │ 77.89 / 79.93 ±1.61 / 81.94 ms │ 78.89 / 80.51 ±0.86 / 81.33 ms │ no change │
│ QQuery 19 │ 37.54 / 37.87 ±0.23 / 38.14 ms │ 37.80 / 38.17 ±0.27 / 38.51 ms │ no change │
│ QQuery 20 │ 40.91 / 41.55 ±0.57 / 42.50 ms │ 40.55 / 41.32 ±0.49 / 41.80 ms │ no change │
│ QQuery 21 │ 64.27 / 66.27 ±1.05 / 67.14 ms │ 64.62 / 67.78 ±2.84 / 72.99 ms │ no change │
│ QQuery 22 │ 17.99 / 18.35 ±0.28 / 18.69 ms │ 18.03 / 18.56 ±0.64 / 19.78 ms │ no change │
└───────────┴────────────────────────────────┴────────────────────────────────┴───────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                         ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                         │ 905.89ms │
│ Total Time (lazy-sort-preserving-merge)   │ 907.38ms │
│ Average Time (HEAD)                       │  41.18ms │
│ Average Time (lazy-sort-preserving-merge) │  41.24ms │
│ Queries Faster                            │        0 │
│ Queries Slower                            │        0 │
│ Queries with No Change                    │       22 │
│ Queries with Failure                      │        0 │
└───────────────────────────────────────────┴──────────┘

Resource Usage

tpch — base (merge-base)

Metric	Value
Wall time	4.8s
Peak memory	4.1 GiB
Avg memory	3.6 GiB
CPU user	33.7s
CPU sys	3.0s
Disk read	0 B
Disk write	136.0 KiB

tpch — branch

Metric	Value
Wall time	4.8s
Peak memory	4.1 GiB
Avg memory	3.6 GiB
CPU user	33.6s
CPU sys	3.0s
Disk read	0 B
Disk write	72.0 KiB

File an issue against this benchmark runner

adriangbot · 2026-04-02T19:46:52Z

🤖 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 lazy-sort-preserving-merge
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃               lazy-sort-preserving-merge ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │           43.80 / 44.54 ±0.77 / 45.90 ms │           43.30 / 44.04 ±0.73 / 45.35 ms │     no change │
│ QQuery 2  │        145.43 / 147.10 ±0.92 / 148.05 ms │        145.34 / 145.94 ±0.56 / 146.84 ms │     no change │
│ QQuery 3  │        114.40 / 114.94 ±0.56 / 115.80 ms │        114.85 / 115.84 ±1.17 / 118.09 ms │     no change │
│ QQuery 4  │    1406.12 / 1423.75 ±14.25 / 1445.73 ms │    1359.98 / 1396.12 ±21.59 / 1421.76 ms │     no change │
│ QQuery 5  │        172.36 / 174.33 ±1.58 / 176.85 ms │        173.75 / 175.10 ±1.17 / 177.19 ms │     no change │
│ QQuery 6  │    1001.85 / 1045.05 ±24.70 / 1070.16 ms │    1036.19 / 1049.60 ±13.95 / 1075.63 ms │     no change │
│ QQuery 7  │        354.77 / 356.64 ±1.70 / 359.35 ms │        356.08 / 361.25 ±6.25 / 373.57 ms │     no change │
│ QQuery 8  │        114.77 / 117.62 ±1.50 / 119.11 ms │        115.44 / 117.19 ±1.17 / 119.02 ms │     no change │
│ QQuery 9  │        101.31 / 105.32 ±3.01 / 108.91 ms │        101.29 / 105.15 ±2.84 / 108.66 ms │     no change │
│ QQuery 10 │        107.99 / 108.89 ±0.74 / 110.13 ms │        108.34 / 109.64 ±0.71 / 110.33 ms │     no change │
│ QQuery 11 │        971.48 / 978.68 ±5.84 / 985.21 ms │      977.52 / 990.67 ±11.72 / 1009.66 ms │     no change │
│ QQuery 12 │           45.45 / 46.92 ±1.10 / 48.26 ms │           47.07 / 48.05 ±0.82 / 49.13 ms │     no change │
│ QQuery 13 │        400.81 / 403.26 ±1.68 / 405.32 ms │        405.73 / 409.09 ±2.09 / 411.79 ms │     no change │
│ QQuery 14 │     1005.54 / 1020.34 ±7.94 / 1028.69 ms │     1022.64 / 1033.90 ±7.82 / 1044.45 ms │     no change │
│ QQuery 15 │           15.63 / 17.14 ±1.12 / 19.06 ms │           15.76 / 17.16 ±0.92 / 18.43 ms │     no change │
│ QQuery 16 │           40.94 / 41.78 ±0.63 / 42.77 ms │           41.13 / 42.29 ±1.02 / 43.78 ms │     no change │
│ QQuery 17 │        242.16 / 243.17 ±1.14 / 245.24 ms │        241.00 / 243.96 ±2.43 / 247.42 ms │     no change │
│ QQuery 18 │        129.70 / 130.98 ±0.80 / 132.09 ms │        131.23 / 132.38 ±1.11 / 133.83 ms │     no change │
│ QQuery 19 │        155.95 / 157.91 ±1.30 / 159.29 ms │        155.84 / 158.28 ±1.39 / 160.12 ms │     no change │
│ QQuery 20 │           13.51 / 14.37 ±0.66 / 15.42 ms │           13.93 / 14.76 ±0.86 / 16.41 ms │     no change │
│ QQuery 21 │           19.85 / 20.30 ±0.42 / 20.94 ms │           19.99 / 20.58 ±0.33 / 20.89 ms │     no change │
│ QQuery 22 │        484.54 / 492.08 ±6.13 / 503.30 ms │        490.14 / 496.14 ±5.48 / 503.35 ms │     no change │
│ QQuery 23 │        894.16 / 900.57 ±4.98 / 908.26 ms │        906.89 / 918.68 ±7.82 / 930.18 ms │     no change │
│ QQuery 24 │        416.94 / 419.75 ±2.86 / 425.09 ms │        421.12 / 423.69 ±1.76 / 426.47 ms │     no change │
│ QQuery 25 │        356.74 / 358.48 ±1.22 / 360.44 ms │        355.88 / 358.22 ±1.58 / 359.99 ms │     no change │
│ QQuery 26 │           80.29 / 82.85 ±2.19 / 85.57 ms │           83.59 / 85.12 ±1.12 / 86.46 ms │     no change │
│ QQuery 27 │        350.37 / 351.99 ±1.36 / 353.83 ms │        351.67 / 354.58 ±2.31 / 358.49 ms │     no change │
│ QQuery 28 │        149.05 / 150.47 ±1.47 / 152.81 ms │        151.02 / 152.36 ±0.92 / 153.64 ms │     no change │
│ QQuery 29 │        297.64 / 300.32 ±2.03 / 303.30 ms │        300.53 / 303.73 ±2.20 / 307.10 ms │     no change │
│ QQuery 30 │           43.24 / 44.52 ±1.14 / 46.56 ms │           43.34 / 44.42 ±0.61 / 45.24 ms │     no change │
│ QQuery 31 │        171.29 / 172.57 ±1.22 / 174.63 ms │        167.75 / 171.19 ±2.20 / 174.18 ms │     no change │
│ QQuery 32 │           57.49 / 58.62 ±1.23 / 60.88 ms │           57.48 / 58.62 ±1.27 / 60.77 ms │     no change │
│ QQuery 33 │        142.86 / 144.07 ±1.51 / 146.85 ms │        141.35 / 142.47 ±0.59 / 143.01 ms │     no change │
│ QQuery 34 │        106.49 / 107.79 ±0.88 / 109.17 ms │        107.49 / 108.68 ±1.35 / 111.14 ms │     no change │
│ QQuery 35 │        109.05 / 111.17 ±1.38 / 113.32 ms │        108.07 / 110.33 ±1.66 / 112.54 ms │     no change │
│ QQuery 36 │        212.58 / 220.59 ±4.24 / 225.19 ms │        212.77 / 219.68 ±5.18 / 224.77 ms │     no change │
│ QQuery 37 │        177.52 / 180.34 ±2.98 / 185.48 ms │        177.45 / 180.71 ±2.82 / 185.34 ms │     no change │
│ QQuery 38 │           84.48 / 87.74 ±3.26 / 93.32 ms │           85.90 / 87.95 ±2.47 / 92.24 ms │     no change │
│ QQuery 39 │        125.11 / 128.17 ±2.88 / 133.01 ms │        123.86 / 126.63 ±1.85 / 129.32 ms │     no change │
│ QQuery 40 │        113.52 / 119.91 ±6.77 / 132.98 ms │        111.77 / 117.93 ±5.97 / 129.15 ms │     no change │
│ QQuery 41 │           15.13 / 15.70 ±0.66 / 16.91 ms │           14.21 / 15.10 ±0.87 / 16.42 ms │     no change │
│ QQuery 42 │        107.51 / 108.93 ±1.31 / 111.10 ms │        108.07 / 108.98 ±1.04 / 110.94 ms │     no change │
│ QQuery 43 │           83.97 / 85.61 ±1.11 / 87.47 ms │           84.22 / 84.91 ±0.56 / 85.77 ms │     no change │
│ QQuery 44 │           11.36 / 12.45 ±1.53 / 15.48 ms │           13.59 / 14.12 ±0.46 / 14.91 ms │  1.13x slower │
│ QQuery 45 │           53.43 / 55.27 ±1.45 / 57.40 ms │           52.28 / 53.81 ±1.36 / 55.93 ms │     no change │
│ QQuery 46 │        236.18 / 238.95 ±1.80 / 240.98 ms │        231.31 / 238.14 ±5.91 / 248.87 ms │     no change │
│ QQuery 47 │       743.91 / 756.94 ±10.17 / 773.80 ms │        772.41 / 777.60 ±3.50 / 782.86 ms │     no change │
│ QQuery 48 │        291.43 / 293.79 ±1.96 / 296.49 ms │        290.46 / 302.24 ±6.45 / 309.59 ms │     no change │
│ QQuery 49 │        253.43 / 255.27 ±1.24 / 257.34 ms │        254.34 / 257.09 ±2.13 / 260.71 ms │     no change │
│ QQuery 50 │        230.72 / 236.61 ±5.46 / 245.73 ms │        229.02 / 236.46 ±4.99 / 244.75 ms │     no change │
│ QQuery 51 │        182.68 / 187.00 ±2.57 / 190.48 ms │        184.06 / 186.47 ±1.45 / 188.03 ms │     no change │
│ QQuery 52 │        109.28 / 110.15 ±0.59 / 110.96 ms │        108.01 / 109.00 ±1.32 / 111.61 ms │     no change │
│ QQuery 53 │        103.70 / 104.50 ±0.43 / 104.87 ms │        103.28 / 104.53 ±0.83 / 105.83 ms │     no change │
│ QQuery 54 │        148.34 / 149.56 ±0.90 / 151.13 ms │        148.26 / 150.42 ±1.88 / 153.92 ms │     no change │
│ QQuery 55 │        107.86 / 109.14 ±1.15 / 110.69 ms │        108.27 / 109.80 ±0.91 / 110.98 ms │     no change │
│ QQuery 56 │        142.07 / 142.59 ±0.66 / 143.89 ms │        143.05 / 144.48 ±1.30 / 146.28 ms │     no change │
│ QQuery 57 │        174.00 / 176.55 ±1.76 / 179.11 ms │        175.65 / 177.09 ±1.36 / 179.50 ms │     no change │
│ QQuery 58 │        299.52 / 303.69 ±2.47 / 306.97 ms │       295.49 / 318.38 ±15.51 / 339.42 ms │     no change │
│ QQuery 59 │        198.44 / 201.47 ±2.48 / 204.97 ms │        198.28 / 201.62 ±1.83 / 203.81 ms │     no change │
│ QQuery 60 │        142.57 / 145.48 ±1.61 / 147.50 ms │        143.90 / 146.02 ±1.94 / 149.03 ms │     no change │
│ QQuery 61 │        172.80 / 174.38 ±1.17 / 176.18 ms │        171.46 / 174.37 ±1.93 / 176.81 ms │     no change │
│ QQuery 62 │        913.45 / 923.46 ±6.20 / 931.00 ms │      926.05 / 947.25 ±28.35 / 1002.39 ms │     no change │
│ QQuery 63 │        105.20 / 106.92 ±1.74 / 109.62 ms │        104.81 / 106.65 ±1.40 / 108.55 ms │     no change │
│ QQuery 64 │        706.62 / 711.87 ±3.78 / 716.66 ms │        704.93 / 711.09 ±4.92 / 718.08 ms │     no change │
│ QQuery 65 │        261.02 / 263.21 ±1.52 / 265.28 ms │        254.54 / 260.72 ±4.60 / 268.64 ms │     no change │
│ QQuery 66 │        248.14 / 259.56 ±5.77 / 263.38 ms │        251.54 / 265.50 ±9.44 / 276.98 ms │     no change │
│ QQuery 67 │        314.65 / 321.55 ±6.75 / 333.84 ms │        310.09 / 317.86 ±5.79 / 325.36 ms │     no change │
│ QQuery 68 │        284.82 / 289.33 ±3.70 / 294.81 ms │        282.92 / 287.53 ±3.83 / 292.25 ms │     no change │
│ QQuery 69 │        105.33 / 106.46 ±0.83 / 107.55 ms │        103.94 / 106.36 ±1.59 / 108.60 ms │     no change │
│ QQuery 70 │       342.96 / 356.60 ±13.13 / 379.01 ms │       338.60 / 346.62 ±10.86 / 368.04 ms │     no change │
│ QQuery 71 │        138.85 / 139.75 ±0.73 / 141.07 ms │        136.65 / 138.32 ±1.63 / 140.33 ms │     no change │
│ QQuery 72 │        738.58 / 750.82 ±9.99 / 766.63 ms │       703.10 / 724.11 ±12.14 / 736.33 ms │     no change │
│ QQuery 73 │        105.02 / 107.78 ±2.56 / 111.86 ms │        102.07 / 104.26 ±1.25 / 105.93 ms │     no change │
│ QQuery 74 │        616.80 / 634.20 ±9.61 / 645.04 ms │        608.66 / 615.92 ±7.01 / 627.80 ms │     no change │
│ QQuery 75 │        279.77 / 281.00 ±1.70 / 284.29 ms │        278.94 / 280.58 ±1.48 / 282.62 ms │     no change │
│ QQuery 76 │        132.89 / 135.11 ±1.16 / 136.13 ms │        133.63 / 134.92 ±1.55 / 137.94 ms │     no change │
│ QQuery 77 │        189.38 / 190.85 ±1.13 / 192.56 ms │        187.46 / 190.51 ±2.10 / 193.58 ms │     no change │
│ QQuery 78 │        359.37 / 363.36 ±2.98 / 367.15 ms │        351.47 / 356.17 ±3.56 / 361.16 ms │     no change │
│ QQuery 79 │        248.33 / 250.30 ±1.33 / 252.28 ms │        234.21 / 236.80 ±2.19 / 240.36 ms │ +1.06x faster │
│ QQuery 80 │        331.41 / 334.29 ±2.46 / 337.83 ms │        327.77 / 328.89 ±1.04 / 330.67 ms │     no change │
│ QQuery 81 │           27.04 / 27.77 ±0.50 / 28.46 ms │           27.64 / 28.33 ±0.46 / 28.85 ms │     no change │
│ QQuery 82 │        201.88 / 206.36 ±3.33 / 210.30 ms │        200.68 / 204.76 ±3.59 / 210.84 ms │     no change │
│ QQuery 83 │           40.28 / 41.27 ±0.74 / 42.45 ms │           39.96 / 41.66 ±0.94 / 42.81 ms │     no change │
│ QQuery 84 │           48.80 / 49.23 ±0.34 / 49.67 ms │           50.08 / 51.24 ±1.01 / 52.45 ms │     no change │
│ QQuery 85 │        151.02 / 153.94 ±2.26 / 157.59 ms │        151.91 / 152.90 ±1.10 / 155.00 ms │     no change │
│ QQuery 86 │           40.63 / 41.67 ±0.75 / 42.90 ms │           39.35 / 40.02 ±0.61 / 41.05 ms │     no change │
│ QQuery 87 │           85.17 / 89.64 ±3.60 / 95.65 ms │           87.39 / 89.11 ±2.24 / 93.51 ms │     no change │
│ QQuery 88 │         99.87 / 100.75 ±0.58 / 101.42 ms │         98.83 / 100.06 ±0.65 / 100.76 ms │     no change │
│ QQuery 89 │        119.57 / 120.72 ±1.13 / 122.46 ms │        118.54 / 120.09 ±0.91 / 121.30 ms │     no change │
│ QQuery 90 │           25.15 / 25.45 ±0.32 / 26.04 ms │           23.84 / 24.55 ±0.94 / 26.40 ms │     no change │
│ QQuery 91 │           62.35 / 64.96 ±1.40 / 66.45 ms │           65.44 / 67.38 ±1.24 / 69.34 ms │     no change │
│ QQuery 92 │           58.37 / 59.13 ±0.61 / 60.23 ms │           58.93 / 59.24 ±0.24 / 59.49 ms │     no change │
│ QQuery 93 │        196.49 / 197.48 ±0.51 / 197.97 ms │        193.84 / 195.01 ±1.47 / 197.70 ms │     no change │
│ QQuery 94 │           62.58 / 63.22 ±0.90 / 64.99 ms │           62.15 / 63.36 ±1.06 / 65.15 ms │     no change │
│ QQuery 95 │        135.65 / 137.41 ±1.56 / 140.17 ms │        133.45 / 136.50 ±1.53 / 137.34 ms │     no change │
│ QQuery 96 │           73.35 / 74.76 ±0.94 / 76.24 ms │           73.98 / 75.04 ±0.73 / 76.15 ms │     no change │
│ QQuery 97 │        129.92 / 133.53 ±2.16 / 136.55 ms │        132.55 / 133.54 ±0.83 / 134.99 ms │     no change │
│ QQuery 98 │        153.89 / 157.07 ±2.57 / 160.05 ms │        153.25 / 158.44 ±2.98 / 161.39 ms │     no change │
│ QQuery 99 │ 10840.72 / 10888.12 ±33.72 / 10927.00 ms │ 10810.18 / 10843.75 ±23.28 / 10872.78 ms │     no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                         ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                         │ 34167.95ms │
│ Total Time (lazy-sort-preserving-merge)   │ 34147.82ms │
│ Average Time (HEAD)                       │   345.13ms │
│ Average Time (lazy-sort-preserving-merge) │   344.93ms │
│ Queries Faster                            │          1 │
│ Queries Slower                            │          1 │
│ Queries with No Change                    │         97 │
│ Queries with Failure                      │          0 │
└───────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	171.2s
Peak memory	5.5 GiB
Avg memory	4.5 GiB
CPU user	275.1s
CPU sys	18.1s
Disk read	0 B
Disk write	639.0 MiB

tpcds — branch

Metric	Value
Wall time	171.1s
Peak memory	5.8 GiB
Avg memory	4.7 GiB
CPU user	274.3s
CPU sys	18.4s
Disk read	0 B
Disk write	796.0 KiB

File an issue against this benchmark runner

adriangbot · 2026-04-02T19:46:57Z

🤖 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 lazy-sort-preserving-merge
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃            lazy-sort-preserving-merge ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.32 / 4.53 ±6.34 / 17.22 ms │          1.29 / 4.46 ±6.24 / 16.94 ms │     no change │
│ QQuery 1  │        14.26 / 14.32 ±0.06 / 14.43 ms │        14.31 / 14.63 ±0.35 / 15.22 ms │     no change │
│ QQuery 2  │        43.50 / 43.99 ±0.29 / 44.30 ms │        44.65 / 44.89 ±0.21 / 45.24 ms │     no change │
│ QQuery 3  │        42.60 / 44.10 ±1.05 / 45.64 ms │        44.74 / 46.10 ±1.17 / 48.20 ms │     no change │
│ QQuery 4  │     284.07 / 294.39 ±5.26 / 298.70 ms │     284.03 / 288.81 ±5.55 / 299.45 ms │     no change │
│ QQuery 5  │     343.27 / 349.08 ±4.20 / 356.08 ms │     343.96 / 346.46 ±2.99 / 352.17 ms │     no change │
│ QQuery 6  │           5.43 / 6.62 ±0.97 / 7.65 ms │          5.56 / 8.40 ±3.44 / 15.04 ms │  1.27x slower │
│ QQuery 7  │        16.40 / 16.80 ±0.32 / 17.36 ms │        16.61 / 17.13 ±0.59 / 18.26 ms │     no change │
│ QQuery 8  │     411.18 / 417.90 ±5.23 / 424.82 ms │    415.21 / 443.66 ±24.16 / 487.53 ms │  1.06x slower │
│ QQuery 9  │    621.24 / 633.96 ±10.59 / 650.55 ms │     636.44 / 644.05 ±6.47 / 655.07 ms │     no change │
│ QQuery 10 │        89.86 / 92.43 ±2.58 / 96.80 ms │        91.28 / 93.97 ±2.99 / 99.69 ms │     no change │
│ QQuery 11 │     102.96 / 103.79 ±0.62 / 104.50 ms │     104.08 / 106.06 ±2.49 / 110.93 ms │     no change │
│ QQuery 12 │     337.21 / 346.86 ±6.25 / 353.55 ms │     342.60 / 346.61 ±2.44 / 350.22 ms │     no change │
│ QQuery 13 │    446.13 / 456.41 ±11.22 / 477.92 ms │    457.53 / 469.03 ±11.93 / 485.25 ms │     no change │
│ QQuery 14 │     348.18 / 352.23 ±2.50 / 354.77 ms │     343.58 / 352.71 ±5.81 / 360.78 ms │     no change │
│ QQuery 15 │    361.82 / 372.69 ±10.23 / 390.58 ms │    356.76 / 375.85 ±20.43 / 411.72 ms │     no change │
│ QQuery 16 │    721.28 / 745.37 ±24.91 / 790.84 ms │    718.26 / 750.57 ±30.65 / 797.57 ms │     no change │
│ QQuery 17 │     718.07 / 725.05 ±4.61 / 732.25 ms │     709.67 / 720.86 ±6.87 / 729.55 ms │     no change │
│ QQuery 18 │ 1446.03 / 1469.53 ±19.14 / 1495.12 ms │ 1385.38 / 1472.36 ±49.24 / 1529.28 ms │     no change │
│ QQuery 19 │        34.89 / 42.01 ±7.77 / 55.18 ms │        34.89 / 39.20 ±6.16 / 51.40 ms │ +1.07x faster │
│ QQuery 20 │    710.86 / 728.65 ±16.02 / 753.45 ms │    714.91 / 724.05 ±10.57 / 744.00 ms │     no change │
│ QQuery 21 │     756.38 / 761.28 ±3.28 / 765.11 ms │     752.40 / 759.53 ±5.97 / 766.99 ms │     no change │
│ QQuery 22 │  1116.82 / 1125.07 ±6.94 / 1137.42 ms │  1120.38 / 1127.16 ±5.22 / 1132.06 ms │     no change │
│ QQuery 23 │ 3017.85 / 3046.09 ±29.48 / 3098.88 ms │ 3048.50 / 3085.51 ±26.50 / 3127.09 ms │     no change │
│ QQuery 24 │       93.49 / 99.75 ±3.45 / 102.69 ms │      98.60 / 100.83 ±1.85 / 103.55 ms │     no change │
│ QQuery 25 │     138.66 / 141.29 ±1.88 / 143.55 ms │     137.64 / 139.90 ±1.83 / 142.26 ms │     no change │
│ QQuery 26 │      97.38 / 100.15 ±1.74 / 102.67 ms │      97.34 / 100.87 ±2.06 / 103.55 ms │     no change │
│ QQuery 27 │     844.72 / 849.46 ±3.92 / 856.45 ms │     843.28 / 848.90 ±6.37 / 861.06 ms │     no change │
│ QQuery 28 │ 7671.11 / 7732.92 ±34.70 / 7774.83 ms │ 7685.77 / 7735.32 ±27.53 / 7770.18 ms │     no change │
│ QQuery 29 │        50.46 / 53.13 ±2.47 / 57.11 ms │        50.30 / 58.03 ±8.62 / 74.88 ms │  1.09x slower │
│ QQuery 30 │     354.21 / 363.29 ±6.86 / 375.08 ms │     359.46 / 365.45 ±4.53 / 371.06 ms │     no change │
│ QQuery 31 │     361.59 / 376.27 ±9.32 / 388.30 ms │    376.30 / 389.06 ±10.84 / 407.57 ms │     no change │
│ QQuery 32 │ 1028.28 / 1058.14 ±28.22 / 1106.02 ms │  1284.87 / 1292.00 ±4.47 / 1298.32 ms │  1.22x slower │
│ QQuery 33 │  1443.84 / 1455.00 ±8.74 / 1463.78 ms │ 1442.67 / 1502.95 ±40.15 / 1565.26 ms │     no change │
│ QQuery 34 │  1436.41 / 1447.06 ±9.23 / 1458.79 ms │ 1445.50 / 1463.29 ±12.64 / 1482.56 ms │     no change │
│ QQuery 35 │     383.94 / 386.50 ±1.30 / 387.56 ms │     380.85 / 388.68 ±4.94 / 394.35 ms │     no change │
│ QQuery 36 │     116.47 / 120.38 ±3.13 / 124.25 ms │     121.81 / 122.82 ±1.16 / 125.05 ms │     no change │
│ QQuery 37 │        46.13 / 49.62 ±2.02 / 51.84 ms │        49.31 / 50.12 ±0.68 / 51.35 ms │     no change │
│ QQuery 38 │        76.61 / 77.14 ±0.85 / 78.84 ms │        75.21 / 76.53 ±0.99 / 78.10 ms │     no change │
│ QQuery 39 │     203.54 / 213.22 ±6.50 / 223.27 ms │     213.54 / 221.07 ±4.50 / 225.31 ms │     no change │
│ QQuery 40 │        23.82 / 25.79 ±1.66 / 28.77 ms │        22.73 / 24.95 ±1.91 / 27.42 ms │     no change │
│ QQuery 41 │        19.13 / 20.56 ±0.98 / 22.07 ms │        18.94 / 20.07 ±0.79 / 21.33 ms │     no change │
│ QQuery 42 │        18.72 / 19.53 ±0.56 / 20.35 ms │        18.96 / 19.55 ±0.35 / 19.87 ms │     no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                         ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                         │ 26782.34ms │
│ Total Time (lazy-sort-preserving-merge)   │ 27202.41ms │
│ Average Time (HEAD)                       │   622.85ms │
│ Average Time (lazy-sort-preserving-merge) │   632.61ms │
│ Queries Faster                            │          1 │
│ Queries Slower                            │          4 │
│ Queries with No Change                    │         38 │
│ Queries with Failure                      │          0 │
└───────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	135.0s
Peak memory	41.0 GiB
Avg memory	31.5 GiB
CPU user	1276.4s
CPU sys	86.8s
Disk read	0 B
Disk write	4.5 GiB

clickbench_partitioned — branch

Metric	Value
Wall time	137.1s
Peak memory	41.0 GiB
Avg memory	31.8 GiB
CPU user	1277.3s
CPU sys	101.0s
Disk read	0 B
Disk write	532.0 KiB

File an issue against this benchmark runner

Dandandan · 2026-04-02T19:53:33Z

run benchmarks

adriangbot · 2026-04-02T19:56:18Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4180142002-672-fb76c 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 lazy-sort-preserving-merge (6eff550) to 4084a18 (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-02T19:56:20Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4180142002-673-xf9zw 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 lazy-sort-preserving-merge (6eff550) to 4084a18 (merge-base) diff using: tpcds
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-02T19:56:21Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4180142002-674-fflk2 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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 lazy-sort-preserving-merge (6eff550) to 4084a18 (merge-base) diff using: tpch
Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-04-02T20:04:44Z

🤖 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 lazy-sort-preserving-merge
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┓
┃ Query     ┃                           HEAD ┃     lazy-sort-preserving-merge ┃       Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━┩
│ QQuery 1  │ 45.15 / 45.71 ±0.78 / 47.19 ms │ 45.21 / 45.72 ±0.67 / 47.03 ms │    no change │
│ QQuery 2  │ 20.70 / 21.07 ±0.27 / 21.40 ms │ 20.76 / 20.93 ±0.26 / 21.44 ms │    no change │
│ QQuery 3  │ 31.46 / 31.69 ±0.20 / 31.95 ms │ 31.45 / 33.08 ±2.30 / 37.61 ms │    no change │
│ QQuery 4  │ 20.48 / 20.99 ±0.51 / 21.95 ms │ 20.81 / 22.16 ±1.24 / 23.90 ms │ 1.06x slower │
│ QQuery 5  │ 47.23 / 48.26 ±0.94 / 49.87 ms │ 47.42 / 49.55 ±1.69 / 51.08 ms │    no change │
│ QQuery 6  │ 17.04 / 17.19 ±0.17 / 17.52 ms │ 17.06 / 18.02 ±1.13 / 20.05 ms │    no change │
│ QQuery 7  │ 54.27 / 55.03 ±0.84 / 56.60 ms │ 53.77 / 54.61 ±0.76 / 55.59 ms │    no change │
│ QQuery 8  │ 47.66 / 48.22 ±0.39 / 48.80 ms │ 47.51 / 47.93 ±0.27 / 48.33 ms │    no change │
│ QQuery 9  │ 55.29 / 55.70 ±0.25 / 56.08 ms │ 53.37 / 54.35 ±0.56 / 55.03 ms │    no change │
│ QQuery 10 │ 70.33 / 71.83 ±1.31 / 74.24 ms │ 70.30 / 71.01 ±0.52 / 71.90 ms │    no change │
│ QQuery 11 │ 13.61 / 13.92 ±0.27 / 14.40 ms │ 13.74 / 13.89 ±0.15 / 14.13 ms │    no change │
│ QQuery 12 │ 27.79 / 28.57 ±0.81 / 29.90 ms │ 27.68 / 27.95 ±0.26 / 28.28 ms │    no change │
│ QQuery 13 │ 38.05 / 39.58 ±1.32 / 41.79 ms │ 37.72 / 38.82 ±0.77 / 39.68 ms │    no change │
│ QQuery 14 │ 28.18 / 28.42 ±0.21 / 28.76 ms │ 28.25 / 28.81 ±0.59 / 29.95 ms │    no change │
│ QQuery 15 │ 33.47 / 33.98 ±0.79 / 35.54 ms │ 33.38 / 33.62 ±0.18 / 33.84 ms │    no change │
│ QQuery 16 │ 15.82 / 16.20 ±0.35 / 16.74 ms │ 15.59 / 16.18 ±0.39 / 16.73 ms │    no change │
│ QQuery 17 │ 71.79 / 72.31 ±0.30 / 72.59 ms │ 72.14 / 73.12 ±0.58 / 73.76 ms │    no change │
│ QQuery 18 │ 76.41 / 77.63 ±1.52 / 80.53 ms │ 76.90 / 78.12 ±0.80 / 79.27 ms │    no change │
│ QQuery 19 │ 37.01 / 37.72 ±0.89 / 39.42 ms │ 37.10 / 37.44 ±0.34 / 38.06 ms │    no change │
│ QQuery 20 │ 39.25 / 40.71 ±1.15 / 42.33 ms │ 39.82 / 40.48 ±0.51 / 41.00 ms │    no change │
│ QQuery 21 │ 62.19 / 63.92 ±1.54 / 66.72 ms │ 62.15 / 63.53 ±1.60 / 66.48 ms │    no change │
│ QQuery 22 │ 17.72 / 18.09 ±0.44 / 18.95 ms │ 17.20 / 17.83 ±0.67 / 19.06 ms │    no change │
└───────────┴────────────────────────────────┴────────────────────────────────┴──────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                         ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                         │ 886.73ms │
│ Total Time (lazy-sort-preserving-merge)   │ 887.16ms │
│ Average Time (HEAD)                       │  40.31ms │
│ Average Time (lazy-sort-preserving-merge) │  40.33ms │
│ Queries Faster                            │        0 │
│ Queries Slower                            │        1 │
│ Queries with No Change                    │       21 │
│ Queries with Failure                      │        0 │
└───────────────────────────────────────────┴──────────┘

Resource Usage

tpch — base (merge-base)

Metric	Value
Wall time	4.7s
Peak memory	4.0 GiB
Avg memory	3.6 GiB
CPU user	33.0s
CPU sys	2.8s
Disk read	0 B
Disk write	136.0 KiB

tpch — branch

Metric	Value
Wall time	4.7s
Peak memory	4.0 GiB
Avg memory	3.6 GiB
CPU user	33.1s
CPU sys	2.8s
Disk read	0 B
Disk write	60.0 KiB

File an issue against this benchmark runner

adriangbot · 2026-04-02T20:10:34Z

🤖 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 lazy-sort-preserving-merge
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃               lazy-sort-preserving-merge ┃       Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━┩
│ QQuery 1  │           43.56 / 44.32 ±0.69 / 45.55 ms │           43.12 / 43.75 ±0.72 / 44.97 ms │    no change │
│ QQuery 2  │        146.49 / 146.81 ±0.29 / 147.18 ms │        146.45 / 147.29 ±0.63 / 148.06 ms │    no change │
│ QQuery 3  │        114.03 / 114.59 ±0.49 / 115.29 ms │        114.56 / 115.13 ±0.59 / 116.27 ms │    no change │
│ QQuery 4  │    1342.85 / 1374.40 ±18.31 / 1396.47 ms │    1326.82 / 1347.07 ±10.48 / 1356.54 ms │    no change │
│ QQuery 5  │        172.34 / 173.96 ±0.83 / 174.59 ms │        173.48 / 174.67 ±0.77 / 175.63 ms │    no change │
│ QQuery 6  │    1024.44 / 1054.63 ±26.54 / 1093.30 ms │    1000.99 / 1026.86 ±22.58 / 1063.44 ms │    no change │
│ QQuery 7  │        353.51 / 357.41 ±3.34 / 362.65 ms │        353.15 / 356.32 ±3.29 / 361.81 ms │    no change │
│ QQuery 8  │        115.91 / 116.95 ±0.76 / 118.09 ms │        117.61 / 118.16 ±0.50 / 119.04 ms │    no change │
│ QQuery 9  │        100.20 / 107.90 ±6.85 / 117.35 ms │        100.89 / 104.82 ±3.02 / 107.90 ms │    no change │
│ QQuery 10 │        107.22 / 108.21 ±0.59 / 109.06 ms │        109.46 / 110.72 ±0.80 / 111.83 ms │    no change │
│ QQuery 11 │        968.82 / 976.91 ±5.39 / 985.62 ms │        926.49 / 942.95 ±9.23 / 952.01 ms │    no change │
│ QQuery 12 │           45.01 / 46.00 ±0.63 / 46.80 ms │           44.99 / 46.41 ±1.87 / 49.99 ms │    no change │
│ QQuery 13 │        400.05 / 403.44 ±2.51 / 406.80 ms │        405.84 / 407.19 ±1.23 / 409.14 ms │    no change │
│ QQuery 14 │     1011.08 / 1027.17 ±8.09 / 1032.27 ms │     1008.18 / 1021.33 ±9.57 / 1037.40 ms │    no change │
│ QQuery 15 │           15.83 / 17.24 ±1.41 / 19.17 ms │           15.76 / 16.41 ±0.35 / 16.81 ms │    no change │
│ QQuery 16 │           40.92 / 42.66 ±1.07 / 43.85 ms │           41.01 / 41.85 ±0.67 / 43.00 ms │    no change │
│ QQuery 17 │        242.63 / 244.98 ±1.24 / 246.05 ms │        239.52 / 241.49 ±1.32 / 243.53 ms │    no change │
│ QQuery 18 │        128.38 / 129.83 ±0.83 / 130.82 ms │        130.44 / 131.73 ±0.72 / 132.45 ms │    no change │
│ QQuery 19 │        154.82 / 156.24 ±1.07 / 157.17 ms │        155.69 / 157.16 ±1.26 / 158.59 ms │    no change │
│ QQuery 20 │           14.04 / 14.37 ±0.34 / 15.00 ms │           13.34 / 14.06 ±0.57 / 15.01 ms │    no change │
│ QQuery 21 │           19.83 / 20.44 ±0.90 / 22.23 ms │           19.47 / 20.13 ±0.45 / 20.86 ms │    no change │
│ QQuery 22 │        488.22 / 491.30 ±2.46 / 493.55 ms │       482.07 / 493.81 ±13.60 / 519.20 ms │    no change │
│ QQuery 23 │        899.10 / 905.10 ±5.03 / 912.46 ms │       887.54 / 899.61 ±12.99 / 923.29 ms │    no change │
│ QQuery 24 │        414.71 / 418.76 ±3.81 / 423.55 ms │       417.28 / 435.76 ±32.59 / 500.89 ms │    no change │
│ QQuery 25 │        349.43 / 355.20 ±3.39 / 359.05 ms │        353.41 / 355.58 ±1.98 / 358.10 ms │    no change │
│ QQuery 26 │           82.62 / 83.45 ±0.80 / 84.76 ms │           83.09 / 84.04 ±0.71 / 85.24 ms │    no change │
│ QQuery 27 │        347.96 / 351.77 ±2.33 / 354.85 ms │        350.42 / 353.63 ±2.23 / 356.23 ms │    no change │
│ QQuery 28 │        149.68 / 150.21 ±0.53 / 151.02 ms │        149.42 / 150.53 ±0.70 / 151.48 ms │    no change │
│ QQuery 29 │        298.97 / 301.42 ±2.83 / 306.56 ms │        299.17 / 303.33 ±4.01 / 308.89 ms │    no change │
│ QQuery 30 │           44.38 / 45.31 ±1.26 / 47.64 ms │           44.94 / 45.86 ±0.82 / 47.29 ms │    no change │
│ QQuery 31 │        170.60 / 172.27 ±1.67 / 174.38 ms │        168.95 / 173.98 ±2.89 / 177.76 ms │    no change │
│ QQuery 32 │           56.84 / 58.18 ±0.77 / 59.14 ms │           57.22 / 58.10 ±0.68 / 58.83 ms │    no change │
│ QQuery 33 │        140.85 / 141.94 ±1.23 / 144.24 ms │        142.23 / 143.43 ±0.93 / 145.05 ms │    no change │
│ QQuery 34 │        107.11 / 108.50 ±1.13 / 110.24 ms │        107.34 / 108.35 ±1.29 / 110.70 ms │    no change │
│ QQuery 35 │        105.98 / 109.20 ±1.96 / 112.19 ms │        107.01 / 109.61 ±2.69 / 114.66 ms │    no change │
│ QQuery 36 │        211.11 / 220.15 ±5.10 / 225.30 ms │        217.79 / 220.32 ±2.91 / 224.84 ms │    no change │
│ QQuery 37 │        179.54 / 181.08 ±1.63 / 183.92 ms │        179.67 / 182.78 ±3.21 / 188.57 ms │    no change │
│ QQuery 38 │           85.89 / 88.95 ±2.64 / 93.27 ms │           84.86 / 88.51 ±3.93 / 95.15 ms │    no change │
│ QQuery 39 │        126.29 / 128.67 ±1.29 / 130.14 ms │        125.02 / 129.24 ±3.27 / 134.72 ms │    no change │
│ QQuery 40 │        111.57 / 116.76 ±4.45 / 124.80 ms │        112.38 / 116.79 ±5.19 / 126.53 ms │    no change │
│ QQuery 41 │           14.31 / 15.52 ±1.00 / 16.95 ms │           14.27 / 15.25 ±1.02 / 17.04 ms │    no change │
│ QQuery 42 │        109.14 / 110.04 ±0.67 / 111.18 ms │        108.86 / 110.54 ±1.18 / 112.34 ms │    no change │
│ QQuery 43 │           84.01 / 84.96 ±0.98 / 86.59 ms │           84.42 / 85.23 ±0.60 / 86.29 ms │    no change │
│ QQuery 44 │           11.43 / 11.76 ±0.23 / 12.05 ms │           13.51 / 14.25 ±0.83 / 15.47 ms │ 1.21x slower │
│ QQuery 45 │           52.92 / 53.95 ±1.09 / 55.68 ms │           53.07 / 54.29 ±1.09 / 56.21 ms │    no change │
│ QQuery 46 │        229.23 / 236.34 ±4.03 / 240.89 ms │        234.67 / 238.73 ±2.35 / 240.72 ms │    no change │
│ QQuery 47 │        709.07 / 720.49 ±8.41 / 732.17 ms │        712.71 / 721.99 ±5.99 / 730.11 ms │    no change │
│ QQuery 48 │        288.25 / 295.36 ±4.07 / 299.21 ms │        287.84 / 294.65 ±4.45 / 299.87 ms │    no change │
│ QQuery 49 │        257.61 / 258.79 ±0.86 / 259.70 ms │        255.23 / 257.43 ±2.15 / 260.60 ms │    no change │
│ QQuery 50 │        230.35 / 239.25 ±6.16 / 247.33 ms │        235.37 / 237.67 ±1.59 / 240.28 ms │    no change │
│ QQuery 51 │        180.72 / 185.66 ±2.97 / 190.06 ms │        184.70 / 188.34 ±1.95 / 190.34 ms │    no change │
│ QQuery 52 │        109.39 / 110.18 ±0.69 / 111.05 ms │        108.31 / 111.25 ±1.81 / 113.16 ms │    no change │
│ QQuery 53 │        104.06 / 104.72 ±0.57 / 105.55 ms │        103.25 / 104.48 ±1.14 / 106.40 ms │    no change │
│ QQuery 54 │        149.52 / 152.16 ±2.15 / 155.30 ms │        147.47 / 149.39 ±1.35 / 151.33 ms │    no change │
│ QQuery 55 │        108.14 / 109.77 ±1.04 / 111.12 ms │        107.45 / 110.13 ±2.96 / 115.20 ms │    no change │
│ QQuery 56 │        143.04 / 143.83 ±0.88 / 145.14 ms │        140.89 / 144.07 ±1.98 / 147.09 ms │    no change │
│ QQuery 57 │        176.14 / 178.52 ±2.08 / 181.84 ms │        173.20 / 175.32 ±1.14 / 176.60 ms │    no change │
│ QQuery 58 │        289.36 / 298.02 ±6.31 / 304.76 ms │        294.24 / 301.66 ±4.79 / 307.53 ms │    no change │
│ QQuery 59 │        197.16 / 199.68 ±1.96 / 201.62 ms │        199.73 / 200.96 ±1.31 / 203.46 ms │    no change │
│ QQuery 60 │        143.54 / 145.67 ±2.07 / 149.18 ms │        143.38 / 145.77 ±1.48 / 147.25 ms │    no change │
│ QQuery 61 │        170.31 / 172.40 ±1.30 / 173.96 ms │        173.01 / 174.30 ±0.89 / 175.70 ms │    no change │
│ QQuery 62 │       894.19 / 921.78 ±22.66 / 954.93 ms │       894.84 / 933.72 ±26.49 / 969.86 ms │    no change │
│ QQuery 63 │        102.92 / 105.72 ±1.85 / 108.57 ms │        103.85 / 107.53 ±2.90 / 112.75 ms │    no change │
│ QQuery 64 │        696.89 / 700.50 ±2.78 / 704.27 ms │        698.94 / 703.20 ±2.50 / 705.73 ms │    no change │
│ QQuery 65 │        251.03 / 257.44 ±4.18 / 261.84 ms │        250.09 / 254.73 ±4.69 / 263.54 ms │    no change │
│ QQuery 66 │       232.20 / 249.57 ±10.22 / 261.93 ms │        242.47 / 254.11 ±8.22 / 265.61 ms │    no change │
│ QQuery 67 │        311.30 / 315.01 ±4.90 / 324.59 ms │        312.99 / 317.13 ±4.44 / 325.48 ms │    no change │
│ QQuery 68 │        280.54 / 283.45 ±1.76 / 286.03 ms │        279.24 / 280.82 ±0.90 / 281.62 ms │    no change │
│ QQuery 69 │        104.04 / 104.91 ±0.76 / 106.00 ms │        104.91 / 105.96 ±0.92 / 107.35 ms │    no change │
│ QQuery 70 │        345.57 / 355.37 ±7.16 / 363.81 ms │        338.87 / 345.49 ±7.79 / 360.41 ms │    no change │
│ QQuery 71 │        134.62 / 135.80 ±1.23 / 137.98 ms │        134.73 / 137.83 ±2.96 / 142.98 ms │    no change │
│ QQuery 72 │        715.22 / 722.92 ±6.36 / 732.08 ms │       713.49 / 728.48 ±12.77 / 745.27 ms │    no change │
│ QQuery 73 │        102.93 / 105.24 ±1.93 / 108.40 ms │        103.69 / 106.40 ±2.51 / 111.12 ms │    no change │
│ QQuery 74 │        584.91 / 588.17 ±2.77 / 592.24 ms │       593.56 / 624.85 ±22.06 / 661.16 ms │ 1.06x slower │
│ QQuery 75 │        274.68 / 277.16 ±2.24 / 280.03 ms │        279.42 / 281.52 ±1.65 / 284.33 ms │    no change │
│ QQuery 76 │        132.16 / 134.83 ±1.80 / 136.89 ms │        133.48 / 135.74 ±1.65 / 138.09 ms │    no change │
│ QQuery 77 │        187.88 / 189.28 ±1.35 / 191.60 ms │        188.69 / 191.09 ±2.28 / 194.53 ms │    no change │
│ QQuery 78 │        349.51 / 357.69 ±5.38 / 365.15 ms │        355.84 / 362.25 ±5.28 / 370.21 ms │    no change │
│ QQuery 79 │        234.54 / 237.01 ±1.48 / 239.09 ms │        237.84 / 245.15 ±6.69 / 253.85 ms │    no change │
│ QQuery 80 │        326.75 / 328.87 ±1.34 / 330.67 ms │        330.64 / 333.00 ±1.98 / 335.63 ms │    no change │
│ QQuery 81 │           26.13 / 27.77 ±0.97 / 29.07 ms │           26.44 / 27.43 ±0.83 / 28.76 ms │    no change │
│ QQuery 82 │        199.96 / 201.30 ±0.86 / 202.42 ms │        202.77 / 205.21 ±2.11 / 208.50 ms │    no change │
│ QQuery 83 │           39.40 / 40.66 ±1.36 / 43.27 ms │           39.47 / 40.63 ±0.73 / 41.54 ms │    no change │
│ QQuery 84 │           48.66 / 49.62 ±0.70 / 50.45 ms │           48.64 / 50.65 ±1.27 / 52.34 ms │    no change │
│ QQuery 85 │        147.77 / 149.84 ±1.64 / 152.80 ms │        152.40 / 154.33 ±1.90 / 156.72 ms │    no change │
│ QQuery 86 │           39.50 / 40.37 ±1.00 / 42.24 ms │           39.95 / 40.73 ±0.90 / 42.43 ms │    no change │
│ QQuery 87 │           87.65 / 90.64 ±2.74 / 95.10 ms │           86.38 / 90.05 ±3.68 / 95.92 ms │    no change │
│ QQuery 88 │          99.00 / 99.96 ±0.53 / 100.54 ms │         99.59 / 100.31 ±0.81 / 101.88 ms │    no change │
│ QQuery 89 │        118.35 / 119.33 ±0.81 / 120.54 ms │        120.24 / 121.01 ±0.55 / 121.72 ms │    no change │
│ QQuery 90 │           23.50 / 23.86 ±0.30 / 24.26 ms │           24.29 / 24.87 ±0.51 / 25.75 ms │    no change │
│ QQuery 91 │           63.30 / 63.82 ±0.36 / 64.26 ms │           63.36 / 65.44 ±1.37 / 67.17 ms │    no change │
│ QQuery 92 │           57.76 / 58.21 ±0.53 / 59.25 ms │           58.15 / 59.60 ±1.15 / 61.51 ms │    no change │
│ QQuery 93 │        189.73 / 192.11 ±2.05 / 195.21 ms │        194.82 / 195.94 ±1.15 / 198.01 ms │    no change │
│ QQuery 94 │           61.72 / 62.26 ±0.33 / 62.68 ms │           62.35 / 63.48 ±1.03 / 65.30 ms │    no change │
│ QQuery 95 │        133.02 / 135.20 ±1.52 / 136.97 ms │        134.42 / 137.68 ±2.08 / 140.59 ms │    no change │
│ QQuery 96 │           71.54 / 73.46 ±0.96 / 74.08 ms │           73.85 / 74.75 ±0.64 / 75.72 ms │    no change │
│ QQuery 97 │        128.89 / 130.65 ±1.89 / 134.30 ms │        132.86 / 134.63 ±1.30 / 136.15 ms │    no change │
│ QQuery 98 │        154.30 / 155.88 ±1.69 / 158.14 ms │        154.65 / 157.48 ±1.76 / 160.18 ms │    no change │
│ QQuery 99 │ 10745.75 / 10793.86 ±45.21 / 10878.95 ms │ 10783.20 / 10831.27 ±43.10 / 10889.72 ms │    no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴──────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                         ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                         │ 33813.26ms │
│ Total Time (lazy-sort-preserving-merge)   │ 33898.99ms │
│ Average Time (HEAD)                       │   341.55ms │
│ Average Time (lazy-sort-preserving-merge) │   342.41ms │
│ Queries Faster                            │          0 │
│ Queries Slower                            │          2 │
│ Queries with No Change                    │         97 │
│ Queries with Failure                      │          0 │
└───────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	169.4s
Peak memory	5.6 GiB
Avg memory	4.5 GiB
CPU user	272.0s
CPU sys	17.6s
Disk read	0 B
Disk write	703.0 MiB

tpcds — branch

Metric	Value
Wall time	169.8s
Peak memory	5.5 GiB
Avg memory	4.6 GiB
CPU user	272.1s
CPU sys	18.1s
Disk read	0 B
Disk write	148.0 KiB

File an issue against this benchmark runner

adriangbot · 2026-04-02T20:10:46Z

🤖 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 lazy-sort-preserving-merge
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃            lazy-sort-preserving-merge ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.28 / 4.56 ±6.40 / 17.35 ms │          1.33 / 4.62 ±6.45 / 17.51 ms │     no change │
│ QQuery 1  │        14.32 / 14.77 ±0.26 / 15.03 ms │        14.43 / 14.73 ±0.18 / 14.94 ms │     no change │
│ QQuery 2  │        44.78 / 45.12 ±0.31 / 45.67 ms │        44.80 / 44.99 ±0.20 / 45.32 ms │     no change │
│ QQuery 3  │        44.69 / 45.32 ±0.66 / 46.53 ms │        43.80 / 45.20 ±0.84 / 46.39 ms │     no change │
│ QQuery 4  │     292.74 / 300.41 ±6.35 / 309.33 ms │     293.46 / 303.78 ±9.47 / 319.34 ms │     no change │
│ QQuery 5  │     346.77 / 353.15 ±4.55 / 359.72 ms │     346.87 / 352.60 ±7.05 / 365.62 ms │     no change │
│ QQuery 6  │           5.92 / 6.64 ±0.58 / 7.54 ms │           5.09 / 6.84 ±1.12 / 8.50 ms │     no change │
│ QQuery 7  │        16.82 / 17.24 ±0.40 / 17.77 ms │        17.08 / 17.90 ±0.72 / 19.04 ms │     no change │
│ QQuery 8  │     416.34 / 423.89 ±4.29 / 428.17 ms │     421.55 / 429.08 ±6.37 / 438.74 ms │     no change │
│ QQuery 9  │     645.55 / 658.00 ±7.10 / 665.34 ms │    644.18 / 662.08 ±12.51 / 677.49 ms │     no change │
│ QQuery 10 │       91.10 / 94.53 ±3.54 / 101.35 ms │       94.32 / 96.67 ±2.77 / 101.97 ms │     no change │
│ QQuery 11 │     106.17 / 107.42 ±1.12 / 109.47 ms │     106.35 / 107.83 ±1.51 / 110.72 ms │     no change │
│ QQuery 12 │     343.79 / 351.30 ±5.19 / 359.45 ms │     349.30 / 354.82 ±3.91 / 359.46 ms │     no change │
│ QQuery 13 │     463.95 / 469.29 ±3.49 / 473.39 ms │    458.95 / 471.51 ±10.55 / 490.32 ms │     no change │
│ QQuery 14 │     348.51 / 355.80 ±4.87 / 361.33 ms │     350.91 / 356.20 ±4.46 / 362.28 ms │     no change │
│ QQuery 15 │    374.95 / 397.08 ±16.02 / 419.14 ms │     361.04 / 366.96 ±5.84 / 376.04 ms │ +1.08x faster │
│ QQuery 16 │    733.53 / 747.87 ±13.48 / 773.39 ms │     718.97 / 726.30 ±4.73 / 733.61 ms │     no change │
│ QQuery 17 │     718.30 / 722.31 ±4.40 / 730.28 ms │     714.86 / 727.88 ±8.63 / 740.01 ms │     no change │
│ QQuery 18 │ 1449.05 / 1486.25 ±35.15 / 1544.04 ms │  1406.57 / 1410.54 ±3.24 / 1416.33 ms │ +1.05x faster │
│ QQuery 19 │        36.78 / 38.16 ±1.46 / 40.15 ms │        36.72 / 40.40 ±3.56 / 45.15 ms │  1.06x slower │
│ QQuery 20 │    717.03 / 732.32 ±16.82 / 754.06 ms │    710.26 / 720.21 ±11.70 / 739.57 ms │     no change │
│ QQuery 21 │     759.28 / 766.56 ±5.98 / 777.25 ms │     755.43 / 758.58 ±2.35 / 760.95 ms │     no change │
│ QQuery 22 │  1129.70 / 1135.22 ±5.65 / 1145.59 ms │  1125.50 / 1130.81 ±3.57 / 1136.21 ms │     no change │
│ QQuery 23 │ 3063.67 / 3092.32 ±30.87 / 3147.47 ms │  3094.27 / 3106.40 ±6.19 / 3111.59 ms │     no change │
│ QQuery 24 │     101.24 / 103.39 ±1.31 / 105.09 ms │      97.60 / 101.16 ±2.25 / 103.55 ms │     no change │
│ QQuery 25 │     137.99 / 139.95 ±2.03 / 142.98 ms │     138.26 / 140.92 ±1.46 / 142.08 ms │     no change │
│ QQuery 26 │      97.32 / 102.20 ±2.89 / 106.16 ms │     100.80 / 104.18 ±1.86 / 105.77 ms │     no change │
│ QQuery 27 │     848.67 / 855.31 ±4.25 / 861.35 ms │     841.92 / 851.34 ±7.11 / 859.64 ms │     no change │
│ QQuery 28 │ 7726.17 / 7789.43 ±38.55 / 7840.38 ms │ 7697.56 / 7772.98 ±39.87 / 7816.02 ms │     no change │
│ QQuery 29 │        50.30 / 55.49 ±3.32 / 60.76 ms │        52.18 / 56.37 ±5.13 / 65.83 ms │     no change │
│ QQuery 30 │     364.43 / 367.96 ±4.09 / 375.75 ms │     365.84 / 376.41 ±7.90 / 387.67 ms │     no change │
│ QQuery 31 │     365.59 / 380.19 ±7.79 / 388.48 ms │    352.26 / 379.27 ±17.30 / 404.46 ms │     no change │
│ QQuery 32 │ 1194.60 / 1265.06 ±48.62 / 1318.77 ms │ 1038.85 / 1054.63 ±17.78 / 1087.37 ms │ +1.20x faster │
│ QQuery 33 │ 1471.45 / 1515.35 ±42.95 / 1594.90 ms │ 1463.32 / 1480.52 ±20.31 / 1519.49 ms │     no change │
│ QQuery 34 │  1479.15 / 1494.24 ±8.83 / 1506.62 ms │ 1466.78 / 1484.75 ±14.27 / 1503.75 ms │     no change │
│ QQuery 35 │     396.66 / 400.11 ±1.98 / 402.13 ms │     393.33 / 401.85 ±8.51 / 416.51 ms │     no change │
│ QQuery 36 │     115.41 / 122.18 ±3.78 / 126.87 ms │     111.74 / 123.56 ±6.49 / 131.46 ms │     no change │
│ QQuery 37 │        49.80 / 51.90 ±1.51 / 53.95 ms │        48.99 / 51.08 ±1.38 / 52.91 ms │     no change │
│ QQuery 38 │        75.55 / 76.40 ±0.82 / 77.82 ms │        74.68 / 75.72 ±0.84 / 77.25 ms │     no change │
│ QQuery 39 │     215.43 / 225.19 ±5.40 / 231.67 ms │     214.25 / 222.56 ±6.42 / 228.45 ms │     no change │
│ QQuery 40 │        25.02 / 26.50 ±1.15 / 27.70 ms │        22.99 / 25.63 ±1.71 / 27.89 ms │     no change │
│ QQuery 41 │        20.17 / 21.55 ±1.41 / 24.24 ms │        20.69 / 22.68 ±1.41 / 24.54 ms │  1.05x slower │
│ QQuery 42 │        19.98 / 20.41 ±0.38 / 20.96 ms │        20.24 / 20.67 ±0.30 / 21.10 ms │     no change │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                         ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                         │ 27378.37ms │
│ Total Time (lazy-sort-preserving-merge)   │ 27003.20ms │
│ Average Time (HEAD)                       │   636.71ms │
│ Average Time (lazy-sort-preserving-merge) │   627.98ms │
│ Queries Faster                            │          3 │
│ Queries Slower                            │          2 │
│ Queries with No Change                    │         38 │
│ Queries with Failure                      │          0 │
└───────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric	Value
Wall time	138.2s
Peak memory	40.5 GiB
Avg memory	31.4 GiB
CPU user	1289.2s
CPU sys	101.5s
Disk read	0 B
Disk write	5.4 GiB

clickbench_partitioned — branch

Metric	Value
Wall time	136.2s
Peak memory	42.6 GiB
Avg memory	34.8 GiB
CPU user	1290.4s
CPU sys	84.0s
Disk read	0 B
Disk write	120.0 KiB

File an issue against this benchmark runner

github-actions bot added the physical-plan Changes to the physical-plan crate label Apr 2, 2026

Dandandan mentioned this pull request Apr 2, 2026

perf: Implement physical execution of uncorrelated scalar subqueries #21240

Open

Conversation

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

Uh oh!

Which issue does this PR close?

Rationale for this change

What changes are included in this PR?

How are these changes tested?

Uh oh!

Dandandan commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

Dandandan commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

adriangbot commented Apr 2, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Dandandan commented Apr 2, 2026 •

edited

Loading