feat: add projection support to SortMergeJoinExec by Dandandan · Pull Request #24517 · apache/datafusion

Dandandan · 2026-08-20T11:16:54Z

Which issue does this PR close?

Closes Add projection support to SortMergeJoinExec #24518

Rationale for this change

SortMergeJoinExec emits every column of both inputs.
We can add projection support to only produce the output columns (saving some compute / copy).

select * from t1 right join t2 on t1.c3 = t2.c3   -- c3 needs a cast

ProjectionExec: expr=[c1@0 as c1, ..., c4@8 as c4]
  SortMergeJoinExec: join_type=Right, on=[(CAST(t1.c3 AS Decimal128(10, 2))@4, c3@2)]

The cast column is only there to be joined on, and every operator above the join
carries it until the projection removes it.

What changes are included in this PR?

SortMergeJoinExec takes an optional projection, set with with_projection, the same
shape as HashJoinExec's.

try_swapping_with_projection embeds the projection into the join when it cannot push
it into the children, so the query above becomes:

SortMergeJoinExec: join_type=Right, on=[(CAST(t1.c3 AS Decimal128(10, 2))@4, c3@2)], projection=[c1@0, c2@1, c3@2, c4@3, c1@5, c2@6, c3@7, c4@8]

The change doesn't bring a big speedup but helps aligning with other join types and helping simplify join optimization in other areas (e.g. join enumeration).

Are these changes tested?

Yes:

a projection pushdown test for a projection that interleaves the two sides, which
the existing pushdown cannot handle
a serialization round trip
existing sqllogictests, whose plans lose a ProjectionExec in four places

Are there any user-facing changes?

Explain / proto changes.

Dandandan · 2026-08-20T11:19:21Z

run benchmarks run tpch tpcds tpch10

env:
    PREFER_HASH_JOIN: false

adriangbot · 2026-08-20T11:22:08Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5355154782-1793-z2clk 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (65415ac) to 5091b42 (merge-base) diff

Run configuration

run benchmark run
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-08-20T11:22:15Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5355154782-1794-cl4xq 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (65415ac) to 5091b42 (merge-base) diff

Run configuration

run benchmark tpch
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-08-20T11:22:20Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5355154782-1796-fxftw 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (65415ac) to 5091b42 (merge-base) diff

Run configuration

run benchmark tpch10
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-08-20T11:22:25Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5355154782-1795-6cscf 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (65415ac) to 5091b42 (merge-base) diff

Run configuration

run benchmark tpcds
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

A sort merge join emitted every column it joined, so a query selecting a subset needed a ProjectionExec above it, and one joining on an expression kept that expression in the output. Projection pushdown could only push into the children when each side's columns stayed together, and gave up otherwise. SortMergeJoinExec now carries an optional projection, as HashJoinExec does, and projection pushdown embeds into the join whatever it cannot push into the children. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012YiAABcW4WSqij31zz2P6c

github-actions · 2026-08-20T11:37:35Z

Thank you for opening this pull request!

Reviewer note: cargo-semver-checks reported the current version number is not SemVer-compatible with the changes in this pull request (compared against the base branch).

Details

     Cloning apache/main
    Building datafusion v55.0.0 (current)
       Built [  80.657s] (current)
     Parsing datafusion v55.0.0 (current)
      Parsed [   0.035s] (current)
    Building datafusion v55.0.0 (baseline)
       Built [  63.308s] (baseline)
     Parsing datafusion v55.0.0 (baseline)
      Parsed [   0.035s] (baseline)
    Checking datafusion v55.0.0 -> v55.0.0 (no change; assume patch)
     Checked [   0.589s] 223 checks: 223 pass, 31 skip
     Summary no semver update required
    Finished [ 146.875s] datafusion
    Building datafusion-physical-plan v55.0.0 (current)
       Built [  41.394s] (current)
     Parsing datafusion-physical-plan v55.0.0 (current)
      Parsed [   0.152s] (current)
    Building datafusion-physical-plan v55.0.0 (baseline)
       Built [  40.435s] (baseline)
     Parsing datafusion-physical-plan v55.0.0 (baseline)
      Parsed [   0.153s] (baseline)
    Checking datafusion-physical-plan v55.0.0 -> v55.0.0 (no change; assume patch)
     Checked [   0.641s] 223 checks: 223 pass, 31 skip
     Summary no semver update required
    Finished [  84.660s] datafusion-physical-plan
    Building datafusion-proto v55.0.0 (current)
       Built [  56.289s] (current)
     Parsing datafusion-proto v55.0.0 (current)
      Parsed [   0.017s] (current)
    Building datafusion-proto v55.0.0 (baseline)
       Built [  56.981s] (baseline)
     Parsing datafusion-proto v55.0.0 (baseline)
      Parsed [   0.019s] (baseline)
    Checking datafusion-proto v55.0.0 -> v55.0.0 (no change; assume patch)
     Checked [   0.122s] 223 checks: 223 pass, 31 skip
     Summary no semver update required
    Finished [ 115.043s] datafusion-proto
    Building datafusion-proto-models v55.0.0 (current)
       Built [  26.737s] (current)
     Parsing datafusion-proto-models v55.0.0 (current)
      Parsed [   0.132s] (current)
    Building datafusion-proto-models v55.0.0 (baseline)
       Built [  26.555s] (baseline)
     Parsing datafusion-proto-models v55.0.0 (baseline)
      Parsed [   0.130s] (baseline)
    Checking datafusion-proto-models v55.0.0 -> v55.0.0 (no change; assume patch)
     Checked [   1.692s] 223 checks: 222 pass, 1 fail, 0 warn, 31 skip

--- failure constructible_struct_adds_field: struct exhaustively constructible through public API adds field ---

Description:
A pub struct that could be exhaustively constructed with a literal using only public API has a new pub field, breaking existing exhaustive literals.
        ref: https://doc.rust-lang.org/reference/expressions/struct-expr.html
       impl: https://github.com/obi1kenobi/cargo-semver-checks/tree/v0.50.0/src/lints/constructible_struct_adds_field.ron

Failed in:
  field SortMergeJoinExecNode.projection in /home/runner/work/datafusion/datafusion/datafusion/proto-models/src/generated/prost.rs:2563
  field SortMergeJoinExecNode.projection in /home/runner/work/datafusion/datafusion/datafusion/proto-models/src/generated/prost.rs:2563

     Summary semver requires new major version: 1 major and 0 minor checks failed
    Finished [  56.916s] datafusion-proto-models
    Building datafusion-sqllogictest v55.0.0 (current)
       Built [ 103.794s] (current)
     Parsing datafusion-sqllogictest v55.0.0 (current)
      Parsed [   0.021s] (current)
    Building datafusion-sqllogictest v55.0.0 (baseline)
       Built [ 103.719s] (baseline)
     Parsing datafusion-sqllogictest v55.0.0 (baseline)
      Parsed [   0.021s] (baseline)
    Checking datafusion-sqllogictest v55.0.0 -> v55.0.0 (no change; assume patch)
     Checked [   0.089s] 223 checks: 223 pass, 31 skip
     Summary no semver update required
    Finished [ 210.956s] datafusion-sqllogictest

adriangbot · 2026-08-20T11:37:39Z

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

Benchmarks requested: run

Kubernetes message

Job has reached the specified backoff limit

File an issue against this benchmark runner

adriangbot · 2026-08-20T11:37:48Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing feat/sort-merge-join-projection (65415ac) to 5091b42 (merge-base) diff

Run configuration

run benchmark tpch
env:
  PREFER_HASH_JOIN: "false"

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 feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┓
┃ Query     ┃     HEAD ┃ feat_sort-merge-join-projection ┃       Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━┩
│ QQuery 1  │ 39.29 ms │                        39.00 ms │    no change │
│ QQuery 2  │ 19.91 ms │                        19.18 ms │    no change │
│ QQuery 3  │ 32.11 ms │                        30.69 ms │    no change │
│ QQuery 4  │ 18.33 ms │                        17.58 ms │    no change │
│ QQuery 5  │ 38.81 ms │                        39.26 ms │    no change │
│ QQuery 6  │ 16.68 ms │                        16.52 ms │    no change │
│ QQuery 7  │ 46.37 ms │                        45.90 ms │    no change │
│ QQuery 8  │ 43.38 ms │                        43.30 ms │    no change │
│ QQuery 9  │ 50.36 ms │                        50.99 ms │    no change │
│ QQuery 10 │ 43.41 ms │                        43.72 ms │    no change │
│ QQuery 11 │ 14.05 ms │                        15.77 ms │ 1.12x slower │
│ QQuery 12 │ 24.96 ms │                        26.62 ms │ 1.07x slower │
│ QQuery 13 │ 32.79 ms │                        32.85 ms │    no change │
│ QQuery 14 │ 23.91 ms │                        24.26 ms │    no change │
│ QQuery 15 │ 32.15 ms │                        31.77 ms │    no change │
│ QQuery 16 │ 14.14 ms │                        14.26 ms │    no change │
│ QQuery 17 │ 75.52 ms │                        75.82 ms │    no change │
│ QQuery 18 │ 62.82 ms │                        62.11 ms │    no change │
│ QQuery 19 │ 33.95 ms │                        33.38 ms │    no change │
│ QQuery 20 │ 33.19 ms │                        33.09 ms │    no change │
│ QQuery 21 │ 58.15 ms │                        57.55 ms │    no change │
│ QQuery 22 │ 14.69 ms │                        14.61 ms │    no change │
└───────────┴──────────┴─────────────────────────────────┴──────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                              ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 768.96ms │
│ Total Time (feat_sort-merge-join-projection)   │ 768.24ms │
│ Average Time (HEAD)                            │  34.95ms │
│ Average Time (feat_sort-merge-join-projection) │  34.92ms │
│ Queries Faster                                 │        0 │
│ Queries Slower                                 │        2 │
│ Queries with No Change                         │       20 │
│ Queries with Failure                           │        0 │
└────────────────────────────────────────────────┴──────────┘

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

Comparing HEAD and feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                           HEAD ┃ feat_sort-merge-join-projection ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │ 39.29 / 40.80 ±1.62 / 43.05 ms │  39.00 / 39.95 ±1.36 / 42.66 ms │     no change │
│ QQuery 2  │ 19.91 / 20.46 ±0.58 / 21.55 ms │  19.18 / 19.57 ±0.21 / 19.79 ms │     no change │
│ QQuery 3  │ 32.11 / 33.42 ±2.16 / 37.69 ms │  30.69 / 34.66 ±3.38 / 38.68 ms │     no change │
│ QQuery 4  │ 18.33 / 18.60 ±0.29 / 19.13 ms │  17.58 / 18.07 ±0.39 / 18.65 ms │     no change │
│ QQuery 5  │ 38.81 / 41.70 ±1.52 / 43.01 ms │  39.26 / 41.71 ±1.85 / 44.99 ms │     no change │
│ QQuery 6  │ 16.68 / 16.96 ±0.20 / 17.20 ms │  16.52 / 16.87 ±0.36 / 17.44 ms │     no change │
│ QQuery 7  │ 46.37 / 48.66 ±1.76 / 51.14 ms │  45.90 / 48.30 ±3.48 / 55.16 ms │     no change │
│ QQuery 8  │ 43.38 / 43.79 ±0.46 / 44.64 ms │  43.30 / 43.56 ±0.23 / 43.99 ms │     no change │
│ QQuery 9  │ 50.36 / 52.31 ±1.71 / 55.50 ms │  50.99 / 52.54 ±0.92 / 53.36 ms │     no change │
│ QQuery 10 │ 43.41 / 44.49 ±1.39 / 47.24 ms │  43.72 / 45.41 ±1.14 / 46.88 ms │     no change │
│ QQuery 11 │ 14.05 / 14.26 ±0.14 / 14.46 ms │  15.77 / 16.03 ±0.22 / 16.38 ms │  1.12x slower │
│ QQuery 12 │ 24.96 / 25.63 ±0.91 / 27.41 ms │  26.62 / 27.90 ±2.06 / 32.00 ms │  1.09x slower │
│ QQuery 13 │ 32.79 / 34.60 ±0.92 / 35.31 ms │  32.85 / 34.46 ±1.63 / 36.60 ms │     no change │
│ QQuery 14 │ 23.91 / 24.32 ±0.30 / 24.71 ms │  24.26 / 24.42 ±0.16 / 24.70 ms │     no change │
│ QQuery 15 │ 32.15 / 34.16 ±1.40 / 35.84 ms │  31.77 / 32.05 ±0.26 / 32.43 ms │ +1.07x faster │
│ QQuery 16 │ 14.14 / 14.42 ±0.24 / 14.72 ms │  14.26 / 14.43 ±0.18 / 14.78 ms │     no change │
│ QQuery 17 │ 75.52 / 76.45 ±0.79 / 77.72 ms │  75.82 / 76.03 ±0.17 / 76.32 ms │     no change │
│ QQuery 18 │ 62.82 / 65.21 ±1.62 / 67.85 ms │  62.11 / 63.89 ±1.35 / 66.26 ms │     no change │
│ QQuery 19 │ 33.95 / 34.31 ±0.31 / 34.87 ms │  33.38 / 33.74 ±0.30 / 34.28 ms │     no change │
│ QQuery 20 │ 33.19 / 33.66 ±0.36 / 34.17 ms │  33.09 / 33.46 ±0.36 / 34.08 ms │     no change │
│ QQuery 21 │ 58.15 / 59.42 ±1.18 / 61.50 ms │  57.55 / 58.73 ±0.82 / 59.90 ms │     no change │
│ QQuery 22 │ 14.69 / 14.89 ±0.19 / 15.21 ms │  14.61 / 15.09 ±0.37 / 15.70 ms │     no change │
└───────────┴────────────────────────────────┴─────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                              ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 792.50ms │
│ Total Time (feat_sort-merge-join-projection)   │ 790.86ms │
│ Average Time (HEAD)                            │  36.02ms │
│ Average Time (feat_sort-merge-join-projection) │  35.95ms │
│ Queries Faster                                 │        1 │
│ Queries Slower                                 │        2 │
│ Queries with No Change                         │       19 │
│ Queries with Failure                           │        0 │
└────────────────────────────────────────────────┴──────────┘

Resource Usage

tpch — base (merge-base)

Metric	Value
Wall time	5.0s
Peak memory	1.2 GiB
Avg memory	510.6 MiB
CPU user	23.0s
CPU sys	1.9s
Peak spill	0 B

tpch — branch

Metric	Value
Wall time	5.0s
Peak memory	1.1 GiB
Avg memory	502.1 MiB
CPU user	22.9s
CPU sys	1.9s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-08-20T11:39:06Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing feat/sort-merge-join-projection (65415ac) to 5091b42 (merge-base) diff

Run configuration

run benchmark tpch10
env:
  PREFER_HASH_JOIN: "false"

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 feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf10.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃      HEAD ┃ feat_sort-merge-join-projection ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │ 313.13 ms │                       315.96 ms │     no change │
│ QQuery 2  │ 104.23 ms │                       103.43 ms │     no change │
│ QQuery 3  │ 238.63 ms │                       240.26 ms │     no change │
│ QQuery 4  │ 115.48 ms │                       117.76 ms │     no change │
│ QQuery 5  │ 366.56 ms │                       365.20 ms │     no change │
│ QQuery 6  │ 125.42 ms │                       126.58 ms │     no change │
│ QQuery 7  │ 493.03 ms │                       496.55 ms │     no change │
│ QQuery 8  │ 389.42 ms │                       386.10 ms │     no change │
│ QQuery 9  │ 587.40 ms │                       571.02 ms │     no change │
│ QQuery 10 │ 307.22 ms │                       310.03 ms │     no change │
│ QQuery 11 │  89.31 ms │                        87.67 ms │     no change │
│ QQuery 12 │ 181.37 ms │                       184.46 ms │     no change │
│ QQuery 13 │ 328.73 ms │                       312.74 ms │     no change │
│ QQuery 14 │ 176.44 ms │                       176.05 ms │     no change │
│ QQuery 15 │ 310.62 ms │                       307.16 ms │     no change │
│ QQuery 16 │  66.65 ms │                        65.53 ms │     no change │
│ QQuery 17 │ 671.98 ms │                       661.11 ms │     no change │
│ QQuery 18 │ 751.70 ms │                       749.06 ms │     no change │
│ QQuery 19 │ 247.27 ms │                       252.16 ms │     no change │
│ QQuery 20 │ 292.53 ms │                       283.79 ms │     no change │
│ QQuery 21 │ 675.06 ms │                       674.36 ms │     no change │
│ QQuery 22 │  64.98 ms │                        59.29 ms │ +1.10x faster │
└───────────┴───────────┴─────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 6897.17ms │
│ Total Time (feat_sort-merge-join-projection)   │ 6846.27ms │
│ Average Time (HEAD)                            │  313.51ms │
│ Average Time (feat_sort-merge-join-projection) │  311.19ms │
│ Queries Faster                                 │         1 │
│ Queries Slower                                 │         0 │
│ Queries with No Change                         │        21 │
│ Queries with Failure                           │         0 │
└────────────────────────────────────────────────┴───────────┘

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

Comparing HEAD and feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf10.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Query     ┃                               HEAD ┃    feat_sort-merge-join-projection ┃    Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ QQuery 1  │  313.13 / 314.65 ±1.09 / 316.37 ms │  315.96 / 317.75 ±0.91 / 318.35 ms │ no change │
│ QQuery 2  │  104.23 / 106.05 ±2.07 / 109.94 ms │  103.43 / 108.22 ±2.93 / 112.25 ms │ no change │
│ QQuery 3  │  238.63 / 243.70 ±5.06 / 251.36 ms │  240.26 / 242.03 ±1.34 / 243.74 ms │ no change │
│ QQuery 4  │  115.48 / 117.72 ±1.30 / 119.33 ms │  117.76 / 119.02 ±1.26 / 121.21 ms │ no change │
│ QQuery 5  │  366.56 / 373.10 ±6.56 / 383.83 ms │  365.20 / 372.30 ±4.74 / 378.72 ms │ no change │
│ QQuery 6  │  125.42 / 127.17 ±2.29 / 131.68 ms │  126.58 / 128.33 ±3.12 / 134.56 ms │ no change │
│ QQuery 7  │  493.03 / 503.03 ±9.25 / 519.36 ms │  496.55 / 508.43 ±7.27 / 518.86 ms │ no change │
│ QQuery 8  │  389.42 / 395.09 ±7.35 / 409.13 ms │ 386.10 / 393.56 ±11.57 / 416.57 ms │ no change │
│ QQuery 9  │  587.40 / 591.53 ±3.68 / 598.29 ms │  571.02 / 580.17 ±7.01 / 592.14 ms │ no change │
│ QQuery 10 │  307.22 / 317.92 ±7.21 / 329.87 ms │  310.03 / 315.88 ±7.24 / 329.74 ms │ no change │
│ QQuery 11 │    89.31 / 94.35 ±6.85 / 107.68 ms │     87.67 / 90.87 ±3.13 / 96.21 ms │ no change │
│ QQuery 12 │  181.37 / 189.34 ±7.15 / 202.65 ms │  184.46 / 191.51 ±7.67 / 205.55 ms │ no change │
│ QQuery 13 │  328.73 / 334.37 ±4.06 / 341.04 ms │  312.74 / 320.89 ±6.08 / 328.50 ms │ no change │
│ QQuery 14 │  176.44 / 179.70 ±4.57 / 188.57 ms │  176.05 / 178.26 ±2.91 / 183.94 ms │ no change │
│ QQuery 15 │  310.62 / 315.58 ±2.79 / 319.27 ms │  307.16 / 310.27 ±2.27 / 314.09 ms │ no change │
│ QQuery 16 │     66.65 / 68.69 ±2.07 / 72.44 ms │     65.53 / 68.37 ±2.65 / 73.36 ms │ no change │
│ QQuery 17 │ 671.98 / 686.66 ±13.76 / 711.73 ms │  661.11 / 673.12 ±6.99 / 681.86 ms │ no change │
│ QQuery 18 │ 751.70 / 775.88 ±18.57 / 800.04 ms │ 749.06 / 764.92 ±15.96 / 793.52 ms │ no change │
│ QQuery 19 │ 247.27 / 262.53 ±17.49 / 291.72 ms │ 252.16 / 274.37 ±19.00 / 298.52 ms │ no change │
│ QQuery 20 │ 292.53 / 301.43 ±12.31 / 325.62 ms │  283.79 / 292.96 ±7.04 / 299.27 ms │ no change │
│ QQuery 21 │ 675.06 / 691.06 ±10.69 / 702.67 ms │  674.36 / 688.24 ±9.26 / 698.35 ms │ no change │
│ QQuery 22 │     64.98 / 66.42 ±1.50 / 69.07 ms │     59.29 / 64.63 ±4.64 / 70.62 ms │ no change │
└───────────┴────────────────────────────────────┴────────────────────────────────────┴───────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 7055.98ms │
│ Total Time (feat_sort-merge-join-projection)   │ 7004.11ms │
│ Average Time (HEAD)                            │  320.73ms │
│ Average Time (feat_sort-merge-join-projection) │  318.37ms │
│ Queries Faster                                 │         0 │
│ Queries Slower                                 │         0 │
│ Queries with No Change                         │        22 │
│ Queries with Failure                           │         0 │
└────────────────────────────────────────────────┴───────────┘

Resource Usage

tpch10 — base (merge-base)

Metric	Value
Wall time	40.0s
Peak memory	5.0 GiB
Avg memory	1.4 GiB
CPU user	357.0s
CPU sys	21.0s
Peak spill	0 B

tpch10 — branch

Metric	Value
Wall time	40.0s
Peak memory	4.8 GiB
Avg memory	1.4 GiB
CPU user	357.0s
CPU sys	20.9s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-08-20T11:41:17Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing feat/sort-merge-join-projection (65415ac) to 5091b42 (merge-base) diff

Run configuration

run benchmark tpcds
env:
  PREFER_HASH_JOIN: "false"

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 feat_sort-merge-join-projection
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ feat_sort-merge-join-projection ┃       Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━┩
│ QQuery 1  │   23.89 ms │                        23.69 ms │    no change │
│ QQuery 2  │   79.67 ms │                        79.00 ms │    no change │
│ QQuery 3  │   50.92 ms │                        50.42 ms │    no change │
│ QQuery 4  │  797.77 ms │                       798.78 ms │    no change │
│ QQuery 5  │  105.60 ms │                       103.93 ms │    no change │
│ QQuery 6  │   69.85 ms │                        70.24 ms │    no change │
│ QQuery 7  │  122.44 ms │                       120.42 ms │    no change │
│ QQuery 8  │   66.28 ms │                        66.24 ms │    no change │
│ QQuery 9  │   50.38 ms │                        50.87 ms │    no change │
│ QQuery 10 │  139.84 ms │                       135.21 ms │    no change │
│ QQuery 11 │  418.68 ms │                       423.78 ms │    no change │
│ QQuery 12 │   32.21 ms │                        32.43 ms │    no change │
│ QQuery 13 │  124.39 ms │                       125.63 ms │    no change │
│ QQuery 14 │  818.42 ms │                       821.67 ms │    no change │
│ QQuery 15 │   55.00 ms │                        55.32 ms │    no change │
│ QQuery 16 │   78.68 ms │                        80.52 ms │    no change │
│ QQuery 17 │  168.30 ms │                       167.53 ms │    no change │
│ QQuery 18 │  148.88 ms │                       148.89 ms │    no change │
│ QQuery 19 │   75.97 ms │                        75.52 ms │    no change │
│ QQuery 20 │   44.91 ms │                        44.51 ms │    no change │
│ QQuery 21 │  387.94 ms │                       388.17 ms │    no change │
│ QQuery 22 │   99.03 ms │                        99.87 ms │    no change │
│ QQuery 23 │  556.80 ms │                       554.00 ms │    no change │
│ QQuery 24 │  345.72 ms │                       345.23 ms │    no change │
│ QQuery 25 │  186.34 ms │                       188.55 ms │    no change │
│ QQuery 26 │   99.50 ms │                       103.01 ms │    no change │
│ QQuery 27 │  123.37 ms │                       124.40 ms │    no change │
│ QQuery 28 │   56.96 ms │                        57.86 ms │    no change │
│ QQuery 29 │  166.63 ms │                       170.53 ms │    no change │
│ QQuery 30 │   49.89 ms │                        53.33 ms │ 1.07x slower │
│ QQuery 31 │  238.99 ms │                       245.89 ms │    no change │
│ QQuery 32 │   67.63 ms │                        68.85 ms │    no change │
│ QQuery 33 │   71.38 ms │                        73.66 ms │    no change │
│ QQuery 34 │   69.54 ms │                        70.78 ms │    no change │
│ QQuery 35 │  123.74 ms │                       132.65 ms │ 1.07x slower │
│ QQuery 36 │   85.67 ms │                        88.59 ms │    no change │
│ QQuery 37 │   32.42 ms │                        34.41 ms │ 1.06x slower │
│ QQuery 38 │   98.46 ms │                       101.26 ms │    no change │
│ QQuery 39 │ 1630.45 ms │                      1650.15 ms │    no change │
│ QQuery 40 │  102.31 ms │                       107.54 ms │ 1.05x slower │
│ QQuery 41 │   11.35 ms │                        12.38 ms │ 1.09x slower │
│ QQuery 42 │   49.12 ms │                        51.02 ms │    no change │
│ QQuery 43 │   56.45 ms │                        57.63 ms │    no change │
│ QQuery 44 │   12.77 ms │                        13.70 ms │ 1.07x slower │
│ QQuery 45 │   46.67 ms │                        49.23 ms │ 1.05x slower │
│ QQuery 46 │  120.62 ms │                       122.69 ms │    no change │
│ QQuery 47 │  369.23 ms │                       367.75 ms │    no change │
│ QQuery 48 │  121.87 ms │                       128.84 ms │ 1.06x slower │
│ QQuery 49 │   79.07 ms │                        81.68 ms │    no change │
│ QQuery 50 │   86.40 ms │                        87.04 ms │    no change │
│ QQuery 51 │  122.62 ms │                       122.96 ms │    no change │
│ QQuery 52 │   49.61 ms │                        50.86 ms │    no change │
│ QQuery 53 │   53.48 ms │                        54.76 ms │    no change │
│ QQuery 54 │   96.76 ms │                       100.11 ms │    no change │
│ QQuery 55 │   47.07 ms │                        48.57 ms │    no change │
│ QQuery 56 │   71.08 ms │                        73.44 ms │    no change │
│ QQuery 57 │  220.19 ms │                       229.38 ms │    no change │
│ QQuery 58 │  186.12 ms │                       193.32 ms │    no change │
│ QQuery 59 │  119.70 ms │                       124.65 ms │    no change │
│ QQuery 60 │   70.63 ms │                        74.26 ms │ 1.05x slower │
│ QQuery 61 │  100.58 ms │                       103.99 ms │    no change │
│ QQuery 62 │   66.07 ms │                        68.22 ms │    no change │
│ QQuery 63 │   54.71 ms │                        55.59 ms │    no change │
│ QQuery 64 │  664.14 ms │                       678.51 ms │    no change │
│ QQuery 65 │  111.31 ms │                       111.86 ms │    no change │
│ QQuery 66 │  168.91 ms │                       170.63 ms │    no change │
│ QQuery 67 │  135.21 ms │                       134.14 ms │    no change │
│ QQuery 68 │  127.53 ms │                       129.56 ms │    no change │
│ QQuery 69 │  128.03 ms │                       129.45 ms │    no change │
│ QQuery 70 │  203.20 ms │                       208.33 ms │    no change │
│ QQuery 71 │   66.29 ms │                        69.43 ms │    no change │
│ QQuery 72 │ 6582.35 ms │                      6616.82 ms │    no change │
│ QQuery 73 │   70.97 ms │                        67.72 ms │    no change │
│ QQuery 74 │  280.47 ms │                       279.29 ms │    no change │
│ QQuery 75 │  216.39 ms │                       220.70 ms │    no change │
│ QQuery 76 │   38.78 ms │                        38.31 ms │    no change │
│ QQuery 77 │  106.35 ms │                       103.65 ms │    no change │
│ QQuery 78 │  288.02 ms │                       283.55 ms │    no change │
│ QQuery 79 │  125.31 ms │                       119.40 ms │    no change │
│ QQuery 80 │  228.96 ms │                       225.89 ms │    no change │
│ QQuery 81 │   46.46 ms │                        45.52 ms │    no change │
│ QQuery 82 │   45.76 ms │                        45.30 ms │    no change │
│ QQuery 83 │   54.82 ms │                        54.46 ms │    no change │
│ QQuery 84 │   45.21 ms │                        45.66 ms │    no change │
│ QQuery 85 │  134.37 ms │                       138.35 ms │    no change │
│ QQuery 86 │   26.86 ms │                        27.31 ms │    no change │
│ QQuery 87 │  100.05 ms │                       101.04 ms │    no change │
│ QQuery 88 │  164.12 ms │                       164.38 ms │    no change │
│ QQuery 89 │   58.20 ms │                        58.99 ms │    no change │
│ QQuery 90 │   26.67 ms │                        26.49 ms │    no change │
│ QQuery 91 │   52.89 ms │                        52.91 ms │    no change │
│ QQuery 92 │   49.04 ms │                        52.27 ms │ 1.07x slower │
│ QQuery 93 │  220.42 ms │                       218.64 ms │    no change │
│ QQuery 94 │   60.76 ms │                        59.80 ms │    no change │
│ QQuery 95 │  155.58 ms │                       153.11 ms │    no change │
│ QQuery 96 │   45.12 ms │                        45.43 ms │    no change │
│ QQuery 97 │   79.80 ms │                        80.09 ms │    no change │
│ QQuery 98 │   65.97 ms │                        64.80 ms │    no change │
│ QQuery 99 │  139.83 ms │                       140.33 ms │    no change │
└───────────┴────────────┴─────────────────────────────────┴──────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 21391.14ms │
│ Total Time (feat_sort-merge-join-projection)   │ 21567.51ms │
│ Average Time (HEAD)                            │   216.07ms │
│ Average Time (feat_sort-merge-join-projection) │   217.85ms │
│ Queries Faster                                 │          0 │
│ Queries Slower                                 │         10 │
│ Queries with No Change                         │         89 │
│ Queries with Failure                           │          0 │
└────────────────────────────────────────────────┴────────────┘

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

Comparing HEAD and feat_sort-merge-join-projection
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃        feat_sort-merge-join-projection ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │         23.89 / 24.64 ±1.01 / 26.62 ms │         23.69 / 24.36 ±1.09 / 26.52 ms │     no change │
│ QQuery 2  │         79.67 / 80.34 ±0.70 / 81.23 ms │         79.00 / 79.26 ±0.31 / 79.86 ms │     no change │
│ QQuery 3  │         50.92 / 51.45 ±0.30 / 51.79 ms │         50.42 / 50.61 ±0.19 / 50.94 ms │     no change │
│ QQuery 4  │      797.77 / 808.57 ±8.39 / 818.87 ms │     798.78 / 808.21 ±10.30 / 827.75 ms │     no change │
│ QQuery 5  │      105.60 / 109.87 ±2.54 / 113.04 ms │      103.93 / 111.94 ±5.08 / 118.44 ms │     no change │
│ QQuery 6  │         69.85 / 70.13 ±0.21 / 70.40 ms │         70.24 / 70.67 ±0.29 / 71.13 ms │     no change │
│ QQuery 7  │      122.44 / 126.26 ±4.17 / 134.05 ms │      120.42 / 126.62 ±5.20 / 136.13 ms │     no change │
│ QQuery 8  │         66.28 / 66.70 ±0.36 / 67.32 ms │         66.24 / 66.65 ±0.28 / 67.11 ms │     no change │
│ QQuery 9  │         50.38 / 52.76 ±1.37 / 54.36 ms │         50.87 / 52.85 ±2.27 / 56.04 ms │     no change │
│ QQuery 10 │      139.84 / 146.47 ±9.57 / 165.50 ms │      135.21 / 142.72 ±7.80 / 156.02 ms │     no change │
│ QQuery 11 │     418.68 / 434.97 ±14.96 / 459.23 ms │      423.78 / 432.67 ±5.57 / 440.98 ms │     no change │
│ QQuery 12 │         32.21 / 35.46 ±5.95 / 47.36 ms │         32.43 / 34.96 ±4.40 / 43.74 ms │     no change │
│ QQuery 13 │      124.39 / 127.99 ±4.54 / 136.94 ms │      125.63 / 126.63 ±0.90 / 128.03 ms │     no change │
│ QQuery 14 │      818.42 / 828.59 ±5.66 / 834.10 ms │      821.67 / 836.59 ±9.29 / 849.95 ms │     no change │
│ QQuery 15 │         55.00 / 59.72 ±6.14 / 70.55 ms │         55.32 / 58.50 ±3.90 / 64.48 ms │     no change │
│ QQuery 16 │         78.68 / 81.63 ±2.57 / 85.93 ms │         80.52 / 83.12 ±2.64 / 88.20 ms │     no change │
│ QQuery 17 │      168.30 / 174.12 ±8.61 / 191.13 ms │      167.53 / 171.28 ±3.09 / 175.63 ms │     no change │
│ QQuery 18 │      148.88 / 153.43 ±4.91 / 162.59 ms │      148.89 / 154.77 ±4.92 / 161.45 ms │     no change │
│ QQuery 19 │         75.97 / 77.08 ±0.97 / 78.53 ms │         75.52 / 76.90 ±1.13 / 78.46 ms │     no change │
│ QQuery 20 │         44.91 / 45.69 ±0.94 / 47.51 ms │         44.51 / 44.87 ±0.30 / 45.40 ms │     no change │
│ QQuery 21 │      387.94 / 393.93 ±3.82 / 399.12 ms │      388.17 / 393.42 ±3.92 / 398.48 ms │     no change │
│ QQuery 22 │       99.03 / 103.72 ±6.41 / 115.85 ms │       99.87 / 106.81 ±7.48 / 118.44 ms │     no change │
│ QQuery 23 │      556.80 / 566.39 ±9.70 / 585.06 ms │      554.00 / 568.33 ±9.17 / 579.04 ms │     no change │
│ QQuery 24 │      345.72 / 352.85 ±7.05 / 365.76 ms │      345.23 / 352.48 ±3.92 / 357.12 ms │     no change │
│ QQuery 25 │      186.34 / 189.13 ±1.67 / 191.08 ms │      188.55 / 195.01 ±6.81 / 205.70 ms │     no change │
│ QQuery 26 │      99.50 / 110.36 ±11.72 / 133.08 ms │      103.01 / 108.96 ±5.49 / 118.83 ms │     no change │
│ QQuery 27 │      123.37 / 125.75 ±2.52 / 130.53 ms │      124.40 / 126.47 ±2.43 / 130.75 ms │     no change │
│ QQuery 28 │         56.96 / 60.74 ±1.95 / 62.30 ms │         57.86 / 63.50 ±3.96 / 70.27 ms │     no change │
│ QQuery 29 │      166.63 / 169.22 ±2.04 / 172.08 ms │      170.53 / 176.05 ±7.86 / 191.58 ms │     no change │
│ QQuery 30 │         49.89 / 53.77 ±5.43 / 64.44 ms │         53.33 / 54.62 ±1.00 / 56.05 ms │     no change │
│ QQuery 31 │      238.99 / 243.52 ±5.54 / 253.59 ms │     245.89 / 253.67 ±10.97 / 275.37 ms │     no change │
│ QQuery 32 │         67.63 / 68.20 ±0.47 / 68.83 ms │         68.85 / 72.09 ±5.00 / 82.00 ms │  1.06x slower │
│ QQuery 33 │         71.38 / 72.70 ±1.04 / 74.32 ms │         73.66 / 75.85 ±2.77 / 81.31 ms │     no change │
│ QQuery 34 │         69.54 / 69.80 ±0.33 / 70.42 ms │         70.78 / 71.60 ±0.75 / 72.68 ms │     no change │
│ QQuery 35 │      123.74 / 129.35 ±3.82 / 133.47 ms │      132.65 / 137.22 ±4.50 / 143.41 ms │  1.06x slower │
│ QQuery 36 │         85.67 / 86.87 ±1.49 / 89.76 ms │         88.59 / 90.29 ±1.26 / 92.40 ms │     no change │
│ QQuery 37 │         32.42 / 32.83 ±0.29 / 33.13 ms │         34.41 / 38.31 ±5.08 / 47.58 ms │  1.17x slower │
│ QQuery 38 │       98.46 / 101.96 ±5.14 / 112.17 ms │      101.26 / 102.61 ±1.72 / 105.98 ms │     no change │
│ QQuery 39 │   1630.45 / 1637.43 ±5.49 / 1644.36 ms │  1650.15 / 1668.00 ±12.19 / 1687.69 ms │     no change │
│ QQuery 40 │      102.31 / 103.55 ±0.75 / 104.50 ms │      107.54 / 113.42 ±4.40 / 119.38 ms │  1.10x slower │
│ QQuery 41 │         11.35 / 11.55 ±0.23 / 12.00 ms │         12.38 / 12.42 ±0.03 / 12.46 ms │  1.08x slower │
│ QQuery 42 │         49.12 / 51.89 ±2.03 / 55.43 ms │         51.02 / 53.02 ±2.88 / 58.71 ms │     no change │
│ QQuery 43 │         56.45 / 57.75 ±1.60 / 60.80 ms │         57.63 / 60.46 ±3.82 / 67.89 ms │     no change │
│ QQuery 44 │         12.77 / 12.90 ±0.18 / 13.23 ms │         13.70 / 13.82 ±0.08 / 13.93 ms │  1.07x slower │
│ QQuery 45 │         46.67 / 46.93 ±0.32 / 47.56 ms │         49.23 / 50.04 ±1.08 / 52.16 ms │  1.07x slower │
│ QQuery 46 │      120.62 / 126.96 ±5.68 / 135.04 ms │      122.69 / 128.47 ±4.52 / 134.99 ms │     no change │
│ QQuery 47 │      369.23 / 377.09 ±6.75 / 387.99 ms │      367.75 / 380.67 ±8.63 / 391.40 ms │     no change │
│ QQuery 48 │      121.87 / 127.95 ±3.78 / 132.49 ms │      128.84 / 133.65 ±3.54 / 139.02 ms │     no change │
│ QQuery 49 │         79.07 / 82.55 ±4.40 / 91.24 ms │         81.68 / 86.00 ±3.58 / 89.82 ms │     no change │
│ QQuery 50 │         86.40 / 87.87 ±1.37 / 89.68 ms │         87.04 / 89.07 ±2.24 / 92.89 ms │     no change │
│ QQuery 51 │      122.62 / 126.55 ±6.92 / 140.38 ms │      122.96 / 126.18 ±4.51 / 134.92 ms │     no change │
│ QQuery 52 │         49.61 / 50.16 ±0.47 / 50.97 ms │         50.86 / 52.71 ±2.32 / 57.20 ms │  1.05x slower │
│ QQuery 53 │         53.48 / 54.36 ±0.56 / 55.00 ms │         54.76 / 55.51 ±0.79 / 56.90 ms │     no change │
│ QQuery 54 │       96.76 / 103.37 ±8.87 / 120.55 ms │      100.11 / 101.77 ±1.83 / 104.69 ms │     no change │
│ QQuery 55 │         47.07 / 47.36 ±0.44 / 48.23 ms │         48.57 / 49.61 ±0.92 / 51.19 ms │     no change │
│ QQuery 56 │         71.08 / 72.92 ±1.28 / 74.70 ms │         73.44 / 75.81 ±2.41 / 80.44 ms │     no change │
│ QQuery 57 │     220.19 / 226.88 ±10.95 / 248.72 ms │      229.38 / 236.40 ±8.67 / 252.20 ms │     no change │
│ QQuery 58 │      186.12 / 188.42 ±2.54 / 192.56 ms │      193.32 / 195.80 ±2.99 / 201.46 ms │     no change │
│ QQuery 59 │      119.70 / 124.20 ±7.20 / 138.54 ms │      124.65 / 125.50 ±0.45 / 125.88 ms │     no change │
│ QQuery 60 │         70.63 / 72.72 ±1.23 / 74.22 ms │         74.26 / 78.58 ±5.49 / 89.04 ms │  1.08x slower │
│ QQuery 61 │      100.58 / 108.51 ±8.58 / 124.10 ms │      103.99 / 107.83 ±3.46 / 112.12 ms │     no change │
│ QQuery 62 │         66.07 / 66.66 ±0.38 / 67.25 ms │         68.22 / 69.04 ±0.68 / 70.24 ms │     no change │
│ QQuery 63 │         54.71 / 57.00 ±2.06 / 60.21 ms │         55.59 / 60.43 ±5.45 / 67.86 ms │  1.06x slower │
│ QQuery 64 │      664.14 / 674.43 ±7.95 / 688.17 ms │      678.51 / 694.52 ±8.71 / 704.11 ms │     no change │
│ QQuery 65 │      111.31 / 113.13 ±2.56 / 118.21 ms │      111.86 / 113.97 ±1.68 / 116.10 ms │     no change │
│ QQuery 66 │      168.91 / 177.44 ±8.06 / 192.35 ms │      170.63 / 178.21 ±5.89 / 184.32 ms │     no change │
│ QQuery 67 │      135.21 / 140.94 ±5.91 / 151.21 ms │      134.14 / 138.40 ±3.25 / 143.57 ms │     no change │
│ QQuery 68 │      127.53 / 132.15 ±5.18 / 141.19 ms │     129.56 / 138.32 ±13.68 / 165.08 ms │     no change │
│ QQuery 69 │      128.03 / 132.53 ±3.83 / 139.15 ms │     129.45 / 135.86 ±11.32 / 158.46 ms │     no change │
│ QQuery 70 │      203.20 / 206.26 ±2.95 / 211.19 ms │      208.33 / 213.72 ±5.82 / 223.33 ms │     no change │
│ QQuery 71 │         66.29 / 70.41 ±3.68 / 77.02 ms │         69.43 / 71.54 ±2.50 / 75.97 ms │     no change │
│ QQuery 72 │ 6582.35 / 6821.03 ±162.84 / 7037.76 ms │ 6616.82 / 6909.49 ±156.86 / 7050.23 ms │     no change │
│ QQuery 73 │      70.97 / 132.14 ±81.27 / 289.82 ms │      67.72 / 135.25 ±73.89 / 268.49 ms │     no change │
│ QQuery 74 │     280.47 / 327.63 ±74.58 / 476.24 ms │     279.29 / 334.75 ±92.30 / 518.65 ms │     no change │
│ QQuery 75 │      216.39 / 226.79 ±6.08 / 232.44 ms │      220.70 / 225.82 ±3.87 / 230.37 ms │     no change │
│ QQuery 76 │         38.78 / 39.94 ±1.66 / 43.20 ms │         38.31 / 42.45 ±3.40 / 48.56 ms │  1.06x slower │
│ QQuery 77 │      106.35 / 109.11 ±2.65 / 113.05 ms │      103.65 / 108.24 ±3.55 / 113.57 ms │     no change │
│ QQuery 78 │      288.02 / 296.18 ±9.62 / 313.94 ms │      283.55 / 291.17 ±4.64 / 296.60 ms │     no change │
│ QQuery 79 │      125.31 / 128.15 ±2.39 / 131.25 ms │      119.40 / 128.72 ±7.89 / 142.23 ms │     no change │
│ QQuery 80 │      228.96 / 231.32 ±1.89 / 234.38 ms │     225.89 / 239.06 ±10.64 / 256.53 ms │     no change │
│ QQuery 81 │       46.46 / 58.19 ±21.81 / 101.78 ms │         45.52 / 49.08 ±3.53 / 54.78 ms │ +1.19x faster │
│ QQuery 82 │         45.76 / 46.57 ±0.93 / 48.29 ms │         45.30 / 46.26 ±0.50 / 46.70 ms │     no change │
│ QQuery 83 │         54.82 / 58.57 ±2.76 / 63.07 ms │         54.46 / 59.29 ±5.28 / 66.41 ms │     no change │
│ QQuery 84 │         45.21 / 46.34 ±0.76 / 47.23 ms │         45.66 / 48.05 ±2.61 / 52.96 ms │     no change │
│ QQuery 85 │      134.37 / 141.77 ±4.78 / 149.12 ms │      138.35 / 140.97 ±2.14 / 144.43 ms │     no change │
│ QQuery 86 │         26.86 / 28.45 ±1.78 / 31.73 ms │         27.31 / 28.42 ±0.85 / 29.78 ms │     no change │
│ QQuery 87 │      100.05 / 103.82 ±6.39 / 116.53 ms │      101.04 / 105.13 ±3.22 / 108.80 ms │     no change │
│ QQuery 88 │      164.12 / 167.80 ±2.37 / 170.16 ms │      164.38 / 167.07 ±2.35 / 171.04 ms │     no change │
│ QQuery 89 │         58.20 / 61.25 ±2.39 / 65.57 ms │         58.99 / 60.48 ±1.23 / 62.48 ms │     no change │
│ QQuery 90 │         26.67 / 28.38 ±1.40 / 30.04 ms │         26.49 / 29.20 ±3.34 / 34.60 ms │     no change │
│ QQuery 91 │         52.89 / 53.67 ±1.17 / 55.98 ms │         52.91 / 53.97 ±1.39 / 56.67 ms │     no change │
│ QQuery 92 │         49.04 / 56.40 ±6.93 / 65.36 ms │         52.27 / 56.30 ±4.28 / 62.29 ms │     no change │
│ QQuery 93 │     220.42 / 233.63 ±11.26 / 252.72 ms │      218.64 / 226.51 ±7.67 / 240.23 ms │     no change │
│ QQuery 94 │         60.76 / 67.17 ±6.34 / 79.15 ms │         59.80 / 64.41 ±6.10 / 76.26 ms │     no change │
│ QQuery 95 │     155.58 / 167.22 ±12.33 / 190.99 ms │     153.11 / 172.86 ±11.98 / 184.48 ms │     no change │
│ QQuery 96 │         45.12 / 48.35 ±2.97 / 53.81 ms │         45.43 / 47.10 ±1.06 / 48.61 ms │     no change │
│ QQuery 97 │         79.80 / 81.92 ±2.42 / 86.58 ms │         80.09 / 82.25 ±1.67 / 85.18 ms │     no change │
│ QQuery 98 │         65.97 / 68.63 ±3.28 / 74.77 ms │         64.80 / 67.89 ±2.02 / 70.96 ms │     no change │
│ QQuery 99 │      139.83 / 145.37 ±6.28 / 157.30 ms │      140.33 / 150.46 ±7.87 / 163.45 ms │     no change │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 22133.57ms │
│ Total Time (feat_sort-merge-join-projection)   │ 22420.89ms │
│ Average Time (HEAD)                            │   223.57ms │
│ Average Time (feat_sort-merge-join-projection) │   226.47ms │
│ Queries Faster                                 │          1 │
│ Queries Slower                                 │         11 │
│ Queries with No Change                         │         87 │
│ Queries with Failure                           │          0 │
└────────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	115.0s
Peak memory	25.1 GiB
Avg memory	4.1 GiB
CPU user	617.2s
CPU sys	30.6s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	115.0s
Peak memory	25.4 GiB
Avg memory	4.3 GiB
CPU user	625.5s
CPU sys	31.5s
Peak spill	0 B

File an issue against this benchmark runner

Dandandan · 2026-08-20T11:45:02Z

run benchmarks run tpch tpcds tpch10

env:
    PREFER_HASH_JOIN: false

adriangbot · 2026-08-20T11:47:10Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5355405346-1799-r87fx 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (2d55b49) to 40c208e (merge-base) diff

Run configuration

run benchmark tpcds
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-08-20T11:48:02Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5355405346-1800-b5bwt 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (2d55b49) to 40c208e (merge-base) diff

Run configuration

run benchmark tpch10
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-08-20T11:48:02Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5355405346-1798-ft7q4 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (2d55b49) to 40c208e (merge-base) diff

Run configuration

run benchmark tpch
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-08-20T11:48:05Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5355405346-1797-f6bzv 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (2d55b49) to 40c208e (merge-base) diff

Run configuration

run benchmark run
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-08-20T11:59:31Z

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

Benchmarks requested: run

Kubernetes message

Job has reached the specified backoff limit

File an issue against this benchmark runner

adriangbot · 2026-08-20T12:00:07Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing feat/sort-merge-join-projection (2d55b49) to 40c208e (merge-base) diff

Run configuration

run benchmark tpch
env:
  PREFER_HASH_JOIN: "false"

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 feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Query     ┃     HEAD ┃ feat_sort-merge-join-projection ┃    Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ QQuery 1  │ 37.93 ms │                        37.76 ms │ no change │
│ QQuery 2  │ 18.94 ms │                        19.07 ms │ no change │
│ QQuery 3  │ 30.89 ms │                        30.66 ms │ no change │
│ QQuery 4  │ 17.47 ms │                        17.26 ms │ no change │
│ QQuery 5  │ 39.36 ms │                        37.84 ms │ no change │
│ QQuery 6  │ 15.86 ms │                        15.96 ms │ no change │
│ QQuery 7  │ 44.54 ms │                        42.81 ms │ no change │
│ QQuery 8  │ 42.09 ms │                        41.85 ms │ no change │
│ QQuery 9  │ 49.06 ms │                        48.88 ms │ no change │
│ QQuery 10 │ 41.38 ms │                        41.54 ms │ no change │
│ QQuery 11 │ 13.02 ms │                        12.97 ms │ no change │
│ QQuery 12 │ 23.85 ms │                        23.75 ms │ no change │
│ QQuery 13 │ 31.24 ms │                        31.72 ms │ no change │
│ QQuery 14 │ 22.81 ms │                        22.89 ms │ no change │
│ QQuery 15 │ 30.24 ms │                        30.10 ms │ no change │
│ QQuery 16 │ 13.60 ms │                        13.59 ms │ no change │
│ QQuery 17 │ 69.08 ms │                        69.38 ms │ no change │
│ QQuery 18 │ 57.72 ms │                        58.10 ms │ no change │
│ QQuery 19 │ 32.32 ms │                        32.55 ms │ no change │
│ QQuery 20 │ 30.80 ms │                        31.07 ms │ no change │
│ QQuery 21 │ 54.22 ms │                        53.65 ms │ no change │
│ QQuery 22 │ 13.72 ms │                        13.51 ms │ no change │
└───────────┴──────────┴─────────────────────────────────┴───────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                              ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 730.15ms │
│ Total Time (feat_sort-merge-join-projection)   │ 726.90ms │
│ Average Time (HEAD)                            │  33.19ms │
│ Average Time (feat_sort-merge-join-projection) │  33.04ms │
│ Queries Faster                                 │        0 │
│ Queries Slower                                 │        0 │
│ Queries with No Change                         │       22 │
│ Queries with Failure                           │        0 │
└────────────────────────────────────────────────┴──────────┘

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

Comparing HEAD and feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┓
┃ Query     ┃                           HEAD ┃ feat_sort-merge-join-projection ┃       Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━┩
│ QQuery 1  │ 37.93 / 38.96 ±1.65 / 42.25 ms │  37.76 / 39.55 ±2.15 / 43.29 ms │    no change │
│ QQuery 2  │ 18.94 / 19.31 ±0.29 / 19.73 ms │  19.07 / 19.38 ±0.46 / 20.27 ms │    no change │
│ QQuery 3  │ 30.89 / 32.25 ±1.00 / 33.39 ms │  30.66 / 33.97 ±1.92 / 35.97 ms │ 1.05x slower │
│ QQuery 4  │ 17.47 / 17.68 ±0.24 / 18.14 ms │  17.26 / 17.44 ±0.12 / 17.56 ms │    no change │
│ QQuery 5  │ 39.36 / 40.18 ±0.96 / 42.05 ms │  37.84 / 40.85 ±1.97 / 43.15 ms │    no change │
│ QQuery 6  │ 15.86 / 16.02 ±0.08 / 16.09 ms │  15.96 / 16.15 ±0.16 / 16.39 ms │    no change │
│ QQuery 7  │ 44.54 / 45.95 ±1.32 / 47.55 ms │  42.81 / 44.80 ±1.25 / 46.14 ms │    no change │
│ QQuery 8  │ 42.09 / 42.74 ±1.23 / 45.20 ms │  41.85 / 42.09 ±0.12 / 42.20 ms │    no change │
│ QQuery 9  │ 49.06 / 49.85 ±0.77 / 51.19 ms │  48.88 / 50.08 ±0.89 / 51.34 ms │    no change │
│ QQuery 10 │ 41.38 / 41.93 ±0.68 / 43.26 ms │  41.54 / 42.18 ±0.96 / 44.09 ms │    no change │
│ QQuery 11 │ 13.02 / 13.66 ±0.78 / 15.18 ms │  12.97 / 13.34 ±0.50 / 14.29 ms │    no change │
│ QQuery 12 │ 23.85 / 24.56 ±1.04 / 26.62 ms │  23.75 / 24.13 ±0.23 / 24.38 ms │    no change │
│ QQuery 13 │ 31.24 / 34.17 ±2.31 / 37.47 ms │  31.72 / 32.73 ±0.84 / 33.71 ms │    no change │
│ QQuery 14 │ 22.81 / 23.65 ±1.26 / 26.14 ms │  22.89 / 23.11 ±0.12 / 23.23 ms │    no change │
│ QQuery 15 │ 30.24 / 30.62 ±0.49 / 31.59 ms │  30.10 / 30.40 ±0.31 / 30.92 ms │    no change │
│ QQuery 16 │ 13.60 / 13.71 ±0.13 / 13.91 ms │  13.59 / 13.67 ±0.05 / 13.73 ms │    no change │
│ QQuery 17 │ 69.08 / 69.52 ±0.48 / 70.34 ms │  69.38 / 70.20 ±1.06 / 72.10 ms │    no change │
│ QQuery 18 │ 57.72 / 59.02 ±0.71 / 59.83 ms │  58.10 / 60.22 ±1.32 / 62.22 ms │    no change │
│ QQuery 19 │ 32.32 / 32.44 ±0.13 / 32.67 ms │  32.55 / 32.84 ±0.32 / 33.29 ms │    no change │
│ QQuery 20 │ 30.80 / 31.33 ±0.58 / 32.45 ms │  31.07 / 31.28 ±0.18 / 31.51 ms │    no change │
│ QQuery 21 │ 54.22 / 56.24 ±1.55 / 58.44 ms │  53.65 / 56.20 ±1.70 / 58.10 ms │    no change │
│ QQuery 22 │ 13.72 / 13.86 ±0.10 / 13.98 ms │  13.51 / 13.75 ±0.15 / 13.95 ms │    no change │
└───────────┴────────────────────────────────┴─────────────────────────────────┴──────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                              ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 747.65ms │
│ Total Time (feat_sort-merge-join-projection)   │ 748.33ms │
│ Average Time (HEAD)                            │  33.98ms │
│ Average Time (feat_sort-merge-join-projection) │  34.02ms │
│ 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	5.0s
Peak memory	1.2 GiB
Avg memory	533.7 MiB
CPU user	21.2s
CPU sys	1.7s
Peak spill	0 B

tpch — branch

Metric	Value
Wall time	5.0s
Peak memory	1.2 GiB
Avg memory	514.1 MiB
CPU user	21.4s
CPU sys	1.6s
Peak spill	0 B

File an issue against this benchmark runner

codecov-commenter · 2026-08-20T12:01:33Z

Codecov Report

❌ Patch coverage is 74.13793% with 30 lines in your changes missing coverage. Please review.
✅ Project coverage is 81.30%. Comparing base (40c208e) to head (dd3a41e).
⚠️ Report is 6 commits behind head on main.

Files with missing lines	Patch %	Lines
...on/physical-plan/src/joins/sort_merge_join/exec.rs	85.14%	8 Missing and 7 partials ⚠️
datafusion/proto-models/src/generated/pbjson.rs	0.00%	15 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #24517      +/-   ##
==========================================
+ Coverage   81.29%   81.30%   +0.01%     
==========================================
  Files        1116     1117       +1     
  Lines      395460   395961     +501     
  Branches   395460   395961     +501     
==========================================
+ Hits       321486   321949     +463     
- Misses      55167    55185      +18     
- Partials    18807    18827      +20

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

adriangbot · 2026-08-20T12:01:44Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing feat/sort-merge-join-projection (2d55b49) to 40c208e (merge-base) diff

Run configuration

run benchmark tpch10
env:
  PREFER_HASH_JOIN: "false"

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 feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf10.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━┓
┃ Query     ┃      HEAD ┃ feat_sort-merge-join-projection ┃       Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━┩
│ QQuery 1  │ 306.40 ms │                       307.70 ms │    no change │
│ QQuery 2  │  99.19 ms │                       109.25 ms │ 1.10x slower │
│ QQuery 3  │ 226.34 ms │                       248.27 ms │ 1.10x slower │
│ QQuery 4  │ 111.40 ms │                       122.04 ms │ 1.10x slower │
│ QQuery 5  │ 350.49 ms │                       389.77 ms │ 1.11x slower │
│ QQuery 6  │ 122.35 ms │                       132.76 ms │ 1.09x slower │
│ QQuery 7  │ 479.14 ms │                       522.36 ms │ 1.09x slower │
│ QQuery 8  │ 370.59 ms │                       383.97 ms │    no change │
│ QQuery 9  │ 555.28 ms │                       550.90 ms │    no change │
│ QQuery 10 │ 297.54 ms │                       295.57 ms │    no change │
│ QQuery 11 │  86.59 ms │                        82.54 ms │    no change │
│ QQuery 12 │ 175.83 ms │                       184.53 ms │    no change │
│ QQuery 13 │ 290.34 ms │                       286.43 ms │    no change │
│ QQuery 14 │ 166.88 ms │                       167.25 ms │    no change │
│ QQuery 15 │ 298.17 ms │                       297.40 ms │    no change │
│ QQuery 16 │  64.75 ms │                        68.84 ms │ 1.06x slower │
│ QQuery 17 │ 619.31 ms │                       608.84 ms │    no change │
│ QQuery 18 │ 670.07 ms │                       690.02 ms │    no change │
│ QQuery 19 │ 239.83 ms │                       241.24 ms │    no change │
│ QQuery 20 │ 277.66 ms │                       273.63 ms │    no change │
│ QQuery 21 │ 644.81 ms │                       663.22 ms │    no change │
│ QQuery 22 │  57.06 ms │                        60.17 ms │ 1.05x slower │
└───────────┴───────────┴─────────────────────────────────┴──────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 6510.02ms │
│ Total Time (feat_sort-merge-join-projection)   │ 6686.70ms │
│ Average Time (HEAD)                            │  295.91ms │
│ Average Time (feat_sort-merge-join-projection) │  303.94ms │
│ Queries Faster                                 │         0 │
│ Queries Slower                                 │         8 │
│ Queries with No Change                         │        14 │
│ Queries with Failure                           │         0 │
└────────────────────────────────────────────────┴───────────┘

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

Comparing HEAD and feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf10.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                               HEAD ┃    feat_sort-merge-join-projection ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │  306.40 / 308.45 ±1.90 / 311.69 ms │  307.70 / 309.07 ±1.40 / 311.24 ms │     no change │
│ QQuery 2  │   99.19 / 104.79 ±4.15 / 110.53 ms │  109.25 / 114.32 ±4.20 / 120.68 ms │  1.09x slower │
│ QQuery 3  │  226.34 / 231.74 ±4.05 / 238.39 ms │  248.27 / 250.90 ±1.63 / 253.16 ms │  1.08x slower │
│ QQuery 4  │  111.40 / 113.27 ±1.57 / 115.85 ms │  122.04 / 126.00 ±2.28 / 128.58 ms │  1.11x slower │
│ QQuery 5  │  350.49 / 360.67 ±6.85 / 368.69 ms │  389.77 / 396.10 ±6.96 / 408.56 ms │  1.10x slower │
│ QQuery 6  │  122.35 / 125.71 ±3.76 / 132.39 ms │  132.76 / 135.23 ±2.48 / 139.52 ms │  1.08x slower │
│ QQuery 7  │ 479.14 / 506.51 ±31.40 / 547.80 ms │  522.36 / 534.45 ±6.63 / 540.50 ms │  1.06x slower │
│ QQuery 8  │  370.59 / 381.31 ±6.90 / 390.00 ms │  383.97 / 388.88 ±3.44 / 392.91 ms │     no change │
│ QQuery 9  │  555.28 / 563.72 ±8.57 / 576.35 ms │ 550.90 / 564.28 ±11.39 / 579.88 ms │     no change │
│ QQuery 10 │  297.54 / 304.94 ±5.69 / 313.96 ms │ 295.57 / 304.48 ±11.25 / 326.50 ms │     no change │
│ QQuery 11 │     86.59 / 89.55 ±3.05 / 93.86 ms │    82.54 / 92.37 ±8.22 / 105.81 ms │     no change │
│ QQuery 12 │  175.83 / 183.31 ±8.35 / 198.70 ms │  184.53 / 187.56 ±3.17 / 193.31 ms │     no change │
│ QQuery 13 │  290.34 / 298.74 ±8.91 / 315.30 ms │  286.43 / 297.28 ±6.80 / 307.21 ms │     no change │
│ QQuery 14 │  166.88 / 172.02 ±5.39 / 178.63 ms │  167.25 / 172.93 ±4.90 / 180.16 ms │     no change │
│ QQuery 15 │  298.17 / 301.66 ±4.64 / 310.83 ms │  297.40 / 308.38 ±9.02 / 319.18 ms │     no change │
│ QQuery 16 │     64.75 / 68.06 ±4.63 / 77.11 ms │     68.84 / 72.32 ±2.57 / 75.50 ms │  1.06x slower │
│ QQuery 17 │ 619.31 / 672.57 ±56.81 / 761.65 ms │ 608.84 / 631.73 ±25.45 / 677.16 ms │ +1.06x faster │
│ QQuery 18 │ 670.07 / 709.72 ±19.96 / 722.87 ms │  690.02 / 700.79 ±8.53 / 710.66 ms │     no change │
│ QQuery 19 │  239.83 / 248.58 ±9.98 / 261.18 ms │ 241.24 / 255.47 ±11.36 / 267.82 ms │     no change │
│ QQuery 20 │ 277.66 / 283.63 ±10.11 / 303.77 ms │ 273.63 / 292.69 ±13.38 / 308.91 ms │     no change │
│ QQuery 21 │  644.81 / 654.08 ±6.08 / 662.95 ms │ 663.22 / 692.73 ±18.93 / 720.63 ms │  1.06x slower │
│ QQuery 22 │     57.06 / 60.42 ±1.92 / 62.91 ms │     60.17 / 62.36 ±1.35 / 63.84 ms │     no change │
└───────────┴────────────────────────────────────┴────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 6743.44ms │
│ Total Time (feat_sort-merge-join-projection)   │ 6890.32ms │
│ Average Time (HEAD)                            │  306.52ms │
│ Average Time (feat_sort-merge-join-projection) │  313.20ms │
│ Queries Faster                                 │         1 │
│ Queries Slower                                 │         8 │
│ Queries with No Change                         │        13 │
│ Queries with Failure                           │         0 │
└────────────────────────────────────────────────┴───────────┘

Resource Usage

tpch10 — base (merge-base)

Metric	Value
Wall time	35.0s
Peak memory	5.3 GiB
Avg memory	1.5 GiB
CPU user	342.8s
CPU sys	18.7s
Peak spill	0 B

tpch10 — branch

Metric	Value
Wall time	35.0s
Peak memory	4.9 GiB
Avg memory	1.5 GiB
CPU user	345.7s
CPU sys	19.6s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-08-20T12:03:39Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing feat/sort-merge-join-projection (2d55b49) to 40c208e (merge-base) diff

Run configuration

run benchmark tpcds
env:
  PREFER_HASH_JOIN: "false"

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 feat_sort-merge-join-projection
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ feat_sort-merge-join-projection ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │   24.51 ms │                        24.14 ms │     no change │
│ QQuery 2  │   79.80 ms │                        79.48 ms │     no change │
│ QQuery 3  │   51.10 ms │                        50.88 ms │     no change │
│ QQuery 4  │  808.95 ms │                       804.67 ms │     no change │
│ QQuery 5  │  104.68 ms │                       104.37 ms │     no change │
│ QQuery 6  │   69.99 ms │                        70.15 ms │     no change │
│ QQuery 7  │  122.50 ms │                       122.17 ms │     no change │
│ QQuery 8  │   66.38 ms │                        66.76 ms │     no change │
│ QQuery 9  │   51.54 ms │                        51.88 ms │     no change │
│ QQuery 10 │  134.72 ms │                       135.32 ms │     no change │
│ QQuery 11 │  422.53 ms │                       418.28 ms │     no change │
│ QQuery 12 │   32.33 ms │                        32.51 ms │     no change │
│ QQuery 13 │  123.91 ms │                       124.22 ms │     no change │
│ QQuery 14 │  827.32 ms │                       822.48 ms │     no change │
│ QQuery 15 │   55.03 ms │                        55.29 ms │     no change │
│ QQuery 16 │   78.05 ms │                        77.99 ms │     no change │
│ QQuery 17 │  167.81 ms │                       169.06 ms │     no change │
│ QQuery 18 │  150.92 ms │                       152.36 ms │     no change │
│ QQuery 19 │   76.27 ms │                        75.88 ms │     no change │
│ QQuery 20 │   44.96 ms │                        44.54 ms │     no change │
│ QQuery 21 │  385.18 ms │                       386.77 ms │     no change │
│ QQuery 22 │   99.69 ms │                        99.33 ms │     no change │
│ QQuery 23 │  567.12 ms │                       557.05 ms │     no change │
│ QQuery 24 │  349.47 ms │                       349.16 ms │     no change │
│ QQuery 25 │  185.48 ms │                       187.32 ms │     no change │
│ QQuery 26 │  103.28 ms │                       102.09 ms │     no change │
│ QQuery 27 │  122.54 ms │                       123.87 ms │     no change │
│ QQuery 28 │   56.82 ms │                        57.79 ms │     no change │
│ QQuery 29 │  169.23 ms │                       167.39 ms │     no change │
│ QQuery 30 │   48.22 ms │                        48.18 ms │     no change │
│ QQuery 31 │  240.39 ms │                       241.44 ms │     no change │
│ QQuery 32 │   68.18 ms │                        68.43 ms │     no change │
│ QQuery 33 │   73.58 ms │                        72.70 ms │     no change │
│ QQuery 34 │   70.19 ms │                        70.49 ms │     no change │
│ QQuery 35 │  126.21 ms │                       126.86 ms │     no change │
│ QQuery 36 │   86.44 ms │                        86.65 ms │     no change │
│ QQuery 37 │   32.83 ms │                        32.80 ms │     no change │
│ QQuery 38 │   99.75 ms │                        99.12 ms │     no change │
│ QQuery 39 │ 1598.47 ms │                      1619.68 ms │     no change │
│ QQuery 40 │  102.47 ms │                       103.18 ms │     no change │
│ QQuery 41 │   11.54 ms │                        11.52 ms │     no change │
│ QQuery 42 │   50.27 ms │                        49.26 ms │     no change │
│ QQuery 43 │   56.19 ms │                        56.35 ms │     no change │
│ QQuery 44 │   12.52 ms │                        12.69 ms │     no change │
│ QQuery 45 │   46.64 ms │                        46.59 ms │     no change │
│ QQuery 46 │  119.52 ms │                       118.74 ms │     no change │
│ QQuery 47 │  362.56 ms │                       360.99 ms │     no change │
│ QQuery 48 │  123.39 ms │                       127.77 ms │     no change │
│ QQuery 49 │   80.57 ms │                        79.94 ms │     no change │
│ QQuery 50 │   87.92 ms │                        85.32 ms │     no change │
│ QQuery 51 │  124.18 ms │                       120.92 ms │     no change │
│ QQuery 52 │   50.30 ms │                        50.12 ms │     no change │
│ QQuery 53 │   53.99 ms │                        53.95 ms │     no change │
│ QQuery 54 │  100.16 ms │                        97.10 ms │     no change │
│ QQuery 55 │   47.24 ms │                        46.90 ms │     no change │
│ QQuery 56 │   72.43 ms │                        72.24 ms │     no change │
│ QQuery 57 │  217.77 ms │                       220.15 ms │     no change │
│ QQuery 58 │  188.84 ms │                       187.16 ms │     no change │
│ QQuery 59 │  120.74 ms │                       121.22 ms │     no change │
│ QQuery 60 │   72.51 ms │                        71.75 ms │     no change │
│ QQuery 61 │  101.35 ms │                       101.57 ms │     no change │
│ QQuery 62 │   65.18 ms │                        65.70 ms │     no change │
│ QQuery 63 │   54.18 ms │                        53.69 ms │     no change │
│ QQuery 64 │  662.31 ms │                       655.14 ms │     no change │
│ QQuery 65 │  111.84 ms │                       111.80 ms │     no change │
│ QQuery 66 │  169.02 ms │                       168.34 ms │     no change │
│ QQuery 67 │  134.74 ms │                       134.74 ms │     no change │
│ QQuery 68 │  127.73 ms │                       126.99 ms │     no change │
│ QQuery 69 │  127.62 ms │                       127.54 ms │     no change │
│ QQuery 70 │  205.75 ms │                       203.86 ms │     no change │
│ QQuery 71 │   64.78 ms │                        65.68 ms │     no change │
│ QQuery 72 │ 6518.71 ms │                      6525.09 ms │     no change │
│ QQuery 73 │   69.04 ms │                        72.74 ms │  1.05x slower │
│ QQuery 74 │  282.67 ms │                       277.86 ms │     no change │
│ QQuery 75 │  213.21 ms │                       215.16 ms │     no change │
│ QQuery 76 │   38.43 ms │                        39.02 ms │     no change │
│ QQuery 77 │  105.90 ms │                       106.63 ms │     no change │
│ QQuery 78 │  298.82 ms │                       282.46 ms │ +1.06x faster │
│ QQuery 79 │  119.79 ms │                       120.68 ms │     no change │
│ QQuery 80 │  230.34 ms │                       232.64 ms │     no change │
│ QQuery 81 │   46.08 ms │                        45.96 ms │     no change │
│ QQuery 82 │   45.53 ms │                        45.59 ms │     no change │
│ QQuery 83 │   55.52 ms │                        54.91 ms │     no change │
│ QQuery 84 │   47.08 ms │                        45.63 ms │     no change │
│ QQuery 85 │  140.82 ms │                       134.22 ms │     no change │
│ QQuery 86 │   26.59 ms │                        27.41 ms │     no change │
│ QQuery 87 │  101.43 ms │                        99.98 ms │     no change │
│ QQuery 88 │  164.71 ms │                       165.36 ms │     no change │
│ QQuery 89 │   59.29 ms │                        59.11 ms │     no change │
│ QQuery 90 │   26.53 ms │                        26.94 ms │     no change │
│ QQuery 91 │   53.57 ms │                        52.87 ms │     no change │
│ QQuery 92 │   50.42 ms │                        51.12 ms │     no change │
│ QQuery 93 │  220.22 ms │                       217.46 ms │     no change │
│ QQuery 94 │   59.70 ms │                        59.46 ms │     no change │
│ QQuery 95 │  153.97 ms │                       167.70 ms │  1.09x slower │
│ QQuery 96 │   45.38 ms │                        46.01 ms │     no change │
│ QQuery 97 │   80.57 ms │                        80.36 ms │     no change │
│ QQuery 98 │   64.92 ms │                        67.47 ms │     no change │
│ QQuery 99 │  143.63 ms │                       144.46 ms │     no change │
└───────────┴────────────┴─────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 21365.47ms │
│ Total Time (feat_sort-merge-join-projection)   │ 21352.77ms │
│ Average Time (HEAD)                            │   215.81ms │
│ Average Time (feat_sort-merge-join-projection) │   215.68ms │
│ Queries Faster                                 │          1 │
│ Queries Slower                                 │          2 │
│ Queries with No Change                         │         96 │
│ Queries with Failure                           │          0 │
└────────────────────────────────────────────────┴────────────┘

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

Comparing HEAD and feat_sort-merge-join-projection
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃        feat_sort-merge-join-projection ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │         24.51 / 25.13 ±0.93 / 26.94 ms │         24.14 / 24.89 ±1.01 / 26.89 ms │     no change │
│ QQuery 2  │         79.80 / 80.20 ±0.39 / 80.90 ms │         79.48 / 79.94 ±0.41 / 80.59 ms │     no change │
│ QQuery 3  │         51.10 / 51.50 ±0.38 / 52.21 ms │         50.88 / 51.15 ±0.21 / 51.43 ms │     no change │
│ QQuery 4  │      808.95 / 818.83 ±7.74 / 831.73 ms │     804.67 / 821.78 ±14.73 / 840.59 ms │     no change │
│ QQuery 5  │      104.68 / 109.35 ±3.44 / 113.43 ms │      104.37 / 108.86 ±2.76 / 113.02 ms │     no change │
│ QQuery 6  │         69.99 / 70.29 ±0.28 / 70.72 ms │         70.15 / 70.52 ±0.27 / 70.92 ms │     no change │
│ QQuery 7  │      122.50 / 128.00 ±7.41 / 142.58 ms │      122.17 / 129.74 ±9.84 / 148.45 ms │     no change │
│ QQuery 8  │         66.38 / 66.67 ±0.18 / 66.91 ms │         66.76 / 67.05 ±0.32 / 67.66 ms │     no change │
│ QQuery 9  │         51.54 / 53.00 ±1.41 / 55.30 ms │         51.88 / 53.41 ±0.82 / 54.16 ms │     no change │
│ QQuery 10 │      134.72 / 141.92 ±8.09 / 157.51 ms │     135.32 / 141.97 ±11.62 / 165.19 ms │     no change │
│ QQuery 11 │     422.53 / 436.88 ±12.04 / 453.93 ms │     418.28 / 439.37 ±15.15 / 455.96 ms │     no change │
│ QQuery 12 │         32.33 / 37.05 ±5.46 / 46.28 ms │         32.51 / 37.08 ±6.29 / 49.27 ms │     no change │
│ QQuery 13 │      123.91 / 125.51 ±1.04 / 126.97 ms │      124.22 / 125.61 ±1.49 / 128.12 ms │     no change │
│ QQuery 14 │      827.32 / 832.20 ±4.29 / 838.62 ms │      822.48 / 836.85 ±8.99 / 846.45 ms │     no change │
│ QQuery 15 │         55.03 / 58.56 ±3.57 / 65.36 ms │         55.29 / 59.86 ±4.48 / 66.99 ms │     no change │
│ QQuery 16 │         78.05 / 80.89 ±2.46 / 84.26 ms │         77.99 / 80.30 ±2.15 / 83.01 ms │     no change │
│ QQuery 17 │      167.81 / 173.24 ±5.01 / 180.48 ms │      169.06 / 172.11 ±2.93 / 177.70 ms │     no change │
│ QQuery 18 │      150.92 / 160.23 ±8.62 / 174.69 ms │      152.36 / 158.29 ±7.12 / 170.81 ms │     no change │
│ QQuery 19 │         76.27 / 78.39 ±1.36 / 79.92 ms │         75.88 / 77.28 ±1.11 / 78.97 ms │     no change │
│ QQuery 20 │         44.96 / 45.24 ±0.14 / 45.37 ms │         44.54 / 45.30 ±0.47 / 45.92 ms │     no change │
│ QQuery 21 │      385.18 / 387.46 ±3.91 / 395.27 ms │      386.77 / 389.89 ±2.44 / 394.18 ms │     no change │
│ QQuery 22 │       99.69 / 105.46 ±6.73 / 118.44 ms │       99.33 / 105.63 ±5.77 / 112.69 ms │     no change │
│ QQuery 23 │      567.12 / 573.55 ±5.61 / 583.85 ms │     557.05 / 572.91 ±10.27 / 585.55 ms │     no change │
│ QQuery 24 │      349.47 / 353.71 ±5.16 / 363.39 ms │      349.16 / 355.91 ±7.02 / 369.14 ms │     no change │
│ QQuery 25 │      185.48 / 192.78 ±5.33 / 200.10 ms │      187.32 / 190.76 ±2.30 / 194.53 ms │     no change │
│ QQuery 26 │      103.28 / 108.39 ±4.61 / 116.11 ms │      102.09 / 108.03 ±3.55 / 111.37 ms │     no change │
│ QQuery 27 │      122.54 / 125.15 ±1.43 / 126.48 ms │      123.87 / 125.63 ±2.18 / 129.86 ms │     no change │
│ QQuery 28 │         56.82 / 60.00 ±2.39 / 62.27 ms │         57.79 / 61.47 ±1.89 / 63.02 ms │     no change │
│ QQuery 29 │      169.23 / 174.51 ±6.03 / 182.73 ms │      167.39 / 174.40 ±6.14 / 181.86 ms │     no change │
│ QQuery 30 │         48.22 / 48.91 ±0.57 / 49.68 ms │         48.18 / 49.37 ±1.49 / 52.11 ms │     no change │
│ QQuery 31 │      240.39 / 243.89 ±5.54 / 254.93 ms │      241.44 / 245.97 ±4.50 / 254.36 ms │     no change │
│ QQuery 32 │         68.18 / 71.00 ±5.14 / 81.28 ms │         68.43 / 70.61 ±3.89 / 78.39 ms │     no change │
│ QQuery 33 │         73.58 / 75.25 ±1.10 / 76.72 ms │         72.70 / 75.63 ±4.27 / 84.01 ms │     no change │
│ QQuery 34 │         70.19 / 70.62 ±0.38 / 71.25 ms │         70.49 / 71.38 ±0.95 / 73.24 ms │     no change │
│ QQuery 35 │      126.21 / 132.04 ±5.62 / 141.94 ms │      126.86 / 130.20 ±3.19 / 135.52 ms │     no change │
│ QQuery 36 │         86.44 / 88.79 ±2.14 / 92.02 ms │         86.65 / 88.48 ±1.43 / 90.51 ms │     no change │
│ QQuery 37 │         32.83 / 35.56 ±5.29 / 46.14 ms │         32.80 / 35.68 ±4.61 / 44.75 ms │     no change │
│ QQuery 38 │       99.75 / 102.06 ±3.72 / 109.45 ms │       99.12 / 101.87 ±3.82 / 109.40 ms │     no change │
│ QQuery 39 │  1598.47 / 1624.57 ±15.48 / 1646.39 ms │   1619.68 / 1629.51 ±6.22 / 1638.16 ms │     no change │
│ QQuery 40 │      102.47 / 106.22 ±5.01 / 116.10 ms │      103.18 / 107.77 ±7.63 / 123.00 ms │     no change │
│ QQuery 41 │         11.54 / 11.78 ±0.20 / 12.12 ms │         11.52 / 11.87 ±0.22 / 12.15 ms │     no change │
│ QQuery 42 │         50.27 / 52.32 ±3.10 / 58.44 ms │         49.26 / 50.71 ±1.20 / 52.40 ms │     no change │
│ QQuery 43 │         56.19 / 56.74 ±0.33 / 57.18 ms │         56.35 / 56.99 ±0.33 / 57.31 ms │     no change │
│ QQuery 44 │         12.52 / 12.97 ±0.31 / 13.44 ms │         12.69 / 12.82 ±0.12 / 13.02 ms │     no change │
│ QQuery 45 │         46.64 / 47.70 ±0.80 / 48.97 ms │         46.59 / 46.86 ±0.22 / 47.23 ms │     no change │
│ QQuery 46 │      119.52 / 123.72 ±4.82 / 132.66 ms │      118.74 / 121.50 ±4.09 / 129.58 ms │     no change │
│ QQuery 47 │      362.56 / 369.08 ±6.49 / 380.30 ms │     360.99 / 368.84 ±10.52 / 389.70 ms │     no change │
│ QQuery 48 │      123.39 / 132.20 ±7.87 / 146.96 ms │      127.77 / 132.90 ±5.39 / 142.46 ms │     no change │
│ QQuery 49 │         80.57 / 81.06 ±0.42 / 81.58 ms │         79.94 / 81.51 ±1.47 / 84.12 ms │     no change │
│ QQuery 50 │         87.92 / 89.53 ±2.00 / 92.67 ms │         85.32 / 87.29 ±1.65 / 89.42 ms │     no change │
│ QQuery 51 │      124.18 / 125.62 ±1.26 / 127.41 ms │      120.92 / 123.38 ±1.52 / 125.60 ms │     no change │
│ QQuery 52 │         50.30 / 50.85 ±0.45 / 51.55 ms │         50.12 / 50.73 ±0.45 / 51.38 ms │     no change │
│ QQuery 53 │         53.99 / 56.22 ±2.35 / 60.19 ms │         53.95 / 57.65 ±5.07 / 67.40 ms │     no change │
│ QQuery 54 │      100.16 / 101.29 ±1.08 / 103.30 ms │         97.10 / 97.81 ±0.47 / 98.44 ms │     no change │
│ QQuery 55 │         47.24 / 47.84 ±0.37 / 48.39 ms │         46.90 / 47.43 ±0.31 / 47.73 ms │     no change │
│ QQuery 56 │         72.43 / 74.27 ±1.42 / 76.46 ms │         72.24 / 74.08 ±1.19 / 75.14 ms │     no change │
│ QQuery 57 │      217.77 / 226.90 ±8.64 / 242.22 ms │      220.15 / 225.44 ±6.44 / 237.76 ms │     no change │
│ QQuery 58 │      188.84 / 193.68 ±6.24 / 205.50 ms │      187.16 / 191.53 ±4.75 / 200.15 ms │     no change │
│ QQuery 59 │      120.74 / 122.61 ±2.84 / 128.26 ms │      121.22 / 123.42 ±3.43 / 130.26 ms │     no change │
│ QQuery 60 │         72.51 / 74.64 ±2.64 / 79.70 ms │         71.75 / 74.52 ±1.88 / 77.06 ms │     no change │
│ QQuery 61 │     101.35 / 112.38 ±10.22 / 124.69 ms │      101.57 / 119.04 ±8.76 / 124.48 ms │  1.06x slower │
│ QQuery 62 │         65.18 / 69.38 ±6.21 / 81.70 ms │         65.70 / 70.63 ±6.43 / 82.98 ms │     no change │
│ QQuery 63 │         54.18 / 57.24 ±3.19 / 63.32 ms │         53.69 / 55.09 ±1.39 / 57.31 ms │     no change │
│ QQuery 64 │     662.31 / 677.73 ±10.75 / 692.88 ms │     655.14 / 674.99 ±14.72 / 699.65 ms │     no change │
│ QQuery 65 │      111.84 / 114.96 ±6.03 / 127.03 ms │      111.80 / 116.06 ±5.63 / 126.92 ms │     no change │
│ QQuery 66 │      169.02 / 171.83 ±2.72 / 175.81 ms │      168.34 / 172.09 ±2.55 / 174.97 ms │     no change │
│ QQuery 67 │      134.74 / 136.01 ±0.91 / 137.07 ms │      134.74 / 135.30 ±0.44 / 135.99 ms │     no change │
│ QQuery 68 │      127.73 / 130.77 ±3.35 / 136.09 ms │     126.99 / 139.00 ±11.21 / 156.21 ms │  1.06x slower │
│ QQuery 69 │      127.62 / 133.86 ±8.81 / 150.24 ms │     127.54 / 137.10 ±11.06 / 157.78 ms │     no change │
│ QQuery 70 │      205.75 / 208.36 ±2.81 / 213.24 ms │      203.86 / 209.28 ±4.07 / 213.94 ms │     no change │
│ QQuery 71 │         64.78 / 66.80 ±1.14 / 68.19 ms │         65.68 / 66.86 ±1.10 / 68.84 ms │     no change │
│ QQuery 72 │ 6518.71 / 6800.31 ±206.81 / 7111.73 ms │ 6525.09 / 6807.77 ±185.79 / 7046.38 ms │     no change │
│ QQuery 73 │      69.04 / 116.19 ±77.69 / 269.89 ms │     72.74 / 195.86 ±147.14 / 444.29 ms │  1.69x slower │
│ QQuery 74 │    282.67 / 350.09 ±122.82 / 595.51 ms │     277.86 / 294.22 ±13.95 / 319.80 ms │ +1.19x faster │
│ QQuery 75 │     213.21 / 225.81 ±14.63 / 254.08 ms │      215.16 / 223.50 ±8.94 / 240.63 ms │     no change │
│ QQuery 76 │         38.43 / 38.82 ±0.53 / 39.85 ms │         39.02 / 39.91 ±0.61 / 40.87 ms │     no change │
│ QQuery 77 │      105.90 / 110.51 ±3.41 / 116.11 ms │      106.63 / 109.29 ±1.92 / 111.32 ms │     no change │
│ QQuery 78 │      298.82 / 305.27 ±5.06 / 310.56 ms │     282.46 / 294.06 ±10.48 / 311.87 ms │     no change │
│ QQuery 79 │      119.79 / 130.36 ±8.28 / 141.36 ms │      120.68 / 126.42 ±3.54 / 131.22 ms │     no change │
│ QQuery 80 │      230.34 / 236.81 ±7.70 / 251.63 ms │     232.64 / 241.92 ±11.83 / 264.74 ms │     no change │
│ QQuery 81 │         46.08 / 47.39 ±1.78 / 50.75 ms │         45.96 / 47.89 ±1.23 / 49.76 ms │     no change │
│ QQuery 82 │         45.53 / 48.87 ±3.44 / 53.52 ms │         45.59 / 47.12 ±1.76 / 50.51 ms │     no change │
│ QQuery 83 │         55.52 / 57.88 ±3.32 / 64.45 ms │         54.91 / 56.35 ±1.06 / 57.94 ms │     no change │
│ QQuery 84 │         47.08 / 50.58 ±3.72 / 57.26 ms │         45.63 / 51.19 ±8.03 / 66.99 ms │     no change │
│ QQuery 85 │      140.82 / 142.96 ±2.71 / 147.92 ms │      134.22 / 138.53 ±5.76 / 149.91 ms │     no change │
│ QQuery 86 │         26.59 / 28.20 ±1.58 / 31.12 ms │         27.41 / 28.05 ±0.72 / 29.42 ms │     no change │
│ QQuery 87 │      101.43 / 103.51 ±2.83 / 109.09 ms │       99.98 / 106.63 ±5.56 / 113.45 ms │     no change │
│ QQuery 88 │      164.71 / 169.40 ±3.22 / 172.85 ms │      165.36 / 169.46 ±3.21 / 175.04 ms │     no change │
│ QQuery 89 │         59.29 / 63.71 ±4.05 / 70.52 ms │         59.11 / 60.31 ±0.85 / 61.62 ms │ +1.06x faster │
│ QQuery 90 │         26.53 / 26.88 ±0.30 / 27.26 ms │         26.94 / 27.14 ±0.15 / 27.31 ms │     no change │
│ QQuery 91 │         53.57 / 59.21 ±7.57 / 73.92 ms │         52.87 / 55.62 ±2.90 / 60.01 ms │ +1.06x faster │
│ QQuery 92 │         50.42 / 52.71 ±2.51 / 57.00 ms │         51.12 / 56.60 ±5.54 / 65.17 ms │  1.07x slower │
│ QQuery 93 │      220.22 / 227.46 ±7.50 / 241.68 ms │      217.46 / 226.21 ±9.94 / 245.60 ms │     no change │
│ QQuery 94 │         59.70 / 69.99 ±5.77 / 77.06 ms │         59.46 / 65.45 ±3.39 / 69.43 ms │ +1.07x faster │
│ QQuery 95 │     153.97 / 169.42 ±11.09 / 182.34 ms │      167.70 / 174.23 ±6.38 / 182.25 ms │     no change │
│ QQuery 96 │         45.38 / 46.50 ±0.57 / 46.85 ms │         46.01 / 48.77 ±2.93 / 52.94 ms │     no change │
│ QQuery 97 │         80.57 / 84.07 ±2.01 / 86.75 ms │         80.36 / 84.58 ±3.02 / 88.20 ms │     no change │
│ QQuery 98 │         64.92 / 67.96 ±2.21 / 71.34 ms │         67.47 / 69.42 ±1.61 / 71.71 ms │     no change │
│ QQuery 99 │      143.63 / 149.28 ±5.01 / 157.73 ms │      144.46 / 149.02 ±4.05 / 154.30 ms │     no change │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 22163.48ms │
│ Total Time (feat_sort-merge-join-projection)   │ 22200.67ms │
│ Average Time (HEAD)                            │   223.87ms │
│ Average Time (feat_sort-merge-join-projection) │   224.25ms │
│ Queries Faster                                 │          4 │
│ Queries Slower                                 │          4 │
│ Queries with No Change                         │         91 │
│ Queries with Failure                           │          0 │
└────────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	115.0s
Peak memory	26.4 GiB
Avg memory	4.0 GiB
CPU user	622.2s
CPU sys	29.9s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	115.0s
Peak memory	26.8 GiB
Avg memory	3.9 GiB
CPU user	622.8s
CPU sys	30.6s
Peak spill	0 B

File an issue against this benchmark runner

jayzhan211

Thanks @Dandandan , a few comments are all somethings we have in HashJoin but not there for SMJ

Swapping a projected join left the indices pointing at the columns the other side now occupies, so it emitted the wrong columns. An empty projection also decoded as an absent one, which is a different output schema. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012YiAABcW4WSqij31zz2P6c

Dandandan · 2026-08-20T13:38:13Z

run benchmarks run tpch tpcds tpch10

env:
    PREFER_HASH_JOIN: false

adriangbot · 2026-08-20T13:41:35Z

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

Benchmarks requested: run

Kubernetes message

Job has reached the specified backoff limit

File an issue against this benchmark runner

adriangbot · 2026-08-20T13:41:39Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5356661936-1804-9qcfg 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (dc0e106) to 40c208e (merge-base) diff

Run configuration

run benchmark tpch10
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-08-20T13:41:39Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5356661936-1802-mh94c 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (dc0e106) to 40c208e (merge-base) diff

Run configuration

run benchmark tpch
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

adriangbot · 2026-08-20T13:41:41Z

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c5356661936-1803-cq4md 6.12.85+ #1 SMP Sat Jun 27 09:31:30 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 feat/sort-merge-join-projection (dc0e106) to 40c208e (merge-base) diff

Run configuration

run benchmark tpcds
env:
  PREFER_HASH_JOIN: "false"

Results will be posted here when complete

File an issue against this benchmark runner

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012YiAABcW4WSqij31zz2P6c

adriangbot · 2026-08-20T13:53:09Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing feat/sort-merge-join-projection (dc0e106) to 40c208e (merge-base) diff

Run configuration

run benchmark tpch
env:
  PREFER_HASH_JOIN: "false"

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 feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Query     ┃     HEAD ┃ feat_sort-merge-join-projection ┃    Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ QQuery 1  │ 37.95 ms │                        37.78 ms │ no change │
│ QQuery 2  │ 19.23 ms │                        18.84 ms │ no change │
│ QQuery 3  │ 30.06 ms │                        30.85 ms │ no change │
│ QQuery 4  │ 17.22 ms │                        17.25 ms │ no change │
│ QQuery 5  │ 37.32 ms │                        36.96 ms │ no change │
│ QQuery 6  │ 15.77 ms │                        16.04 ms │ no change │
│ QQuery 7  │ 42.54 ms │                        44.02 ms │ no change │
│ QQuery 8  │ 41.70 ms │                        41.34 ms │ no change │
│ QQuery 9  │ 48.64 ms │                        48.69 ms │ no change │
│ QQuery 10 │ 41.74 ms │                        41.69 ms │ no change │
│ QQuery 11 │ 13.17 ms │                        12.90 ms │ no change │
│ QQuery 12 │ 23.71 ms │                        23.95 ms │ no change │
│ QQuery 13 │ 31.59 ms │                        31.62 ms │ no change │
│ QQuery 14 │ 22.76 ms │                        22.86 ms │ no change │
│ QQuery 15 │ 30.16 ms │                        30.15 ms │ no change │
│ QQuery 16 │ 13.43 ms │                        13.51 ms │ no change │
│ QQuery 17 │ 68.53 ms │                        68.87 ms │ no change │
│ QQuery 18 │ 58.38 ms │                        57.86 ms │ no change │
│ QQuery 19 │ 32.01 ms │                        32.20 ms │ no change │
│ QQuery 20 │ 31.02 ms │                        30.85 ms │ no change │
│ QQuery 21 │ 54.81 ms │                        53.32 ms │ no change │
│ QQuery 22 │ 13.78 ms │                        13.64 ms │ no change │
└───────────┴──────────┴─────────────────────────────────┴───────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                              ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 725.53ms │
│ Total Time (feat_sort-merge-join-projection)   │ 725.19ms │
│ Average Time (HEAD)                            │  32.98ms │
│ Average Time (feat_sort-merge-join-projection) │  32.96ms │
│ Queries Faster                                 │        0 │
│ Queries Slower                                 │        0 │
│ Queries with No Change                         │       22 │
│ Queries with Failure                           │        0 │
└────────────────────────────────────────────────┴──────────┘

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

Comparing HEAD and feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Query     ┃                           HEAD ┃ feat_sort-merge-join-projection ┃    Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ QQuery 1  │ 37.95 / 38.46 ±0.79 / 40.03 ms │  37.78 / 38.75 ±1.52 / 41.76 ms │ no change │
│ QQuery 2  │ 19.23 / 19.79 ±0.91 / 21.59 ms │  18.84 / 19.46 ±0.70 / 20.69 ms │ no change │
│ QQuery 3  │ 30.06 / 32.09 ±1.55 / 33.79 ms │  30.85 / 32.69 ±1.56 / 35.07 ms │ no change │
│ QQuery 4  │ 17.22 / 17.86 ±0.56 / 18.76 ms │  17.25 / 17.91 ±0.64 / 19.04 ms │ no change │
│ QQuery 5  │ 37.32 / 39.39 ±1.05 / 40.21 ms │  36.96 / 39.37 ±1.21 / 40.20 ms │ no change │
│ QQuery 6  │ 15.77 / 16.84 ±1.94 / 20.71 ms │  16.04 / 16.74 ±0.86 / 18.35 ms │ no change │
│ QQuery 7  │ 42.54 / 45.31 ±1.85 / 47.25 ms │  44.02 / 45.49 ±0.83 / 46.35 ms │ no change │
│ QQuery 8  │ 41.70 / 42.51 ±0.71 / 43.47 ms │  41.34 / 41.80 ±0.26 / 42.16 ms │ no change │
│ QQuery 9  │ 48.64 / 50.08 ±1.09 / 52.00 ms │  48.69 / 50.42 ±1.26 / 52.60 ms │ no change │
│ QQuery 10 │ 41.74 / 41.92 ±0.24 / 42.39 ms │  41.69 / 41.89 ±0.13 / 42.01 ms │ no change │
│ QQuery 11 │ 13.17 / 13.30 ±0.11 / 13.49 ms │  12.90 / 13.24 ±0.29 / 13.72 ms │ no change │
│ QQuery 12 │ 23.71 / 25.50 ±1.72 / 28.15 ms │  23.95 / 24.29 ±0.26 / 24.72 ms │ no change │
│ QQuery 13 │ 31.59 / 32.32 ±0.52 / 33.00 ms │  31.62 / 32.92 ±1.31 / 35.11 ms │ no change │
│ QQuery 14 │ 22.76 / 23.05 ±0.22 / 23.41 ms │  22.86 / 22.95 ±0.11 / 23.15 ms │ no change │
│ QQuery 15 │ 30.16 / 31.37 ±1.17 / 33.05 ms │  30.15 / 30.38 ±0.12 / 30.49 ms │ no change │
│ QQuery 16 │ 13.43 / 13.70 ±0.20 / 13.94 ms │  13.51 / 13.66 ±0.17 / 13.96 ms │ no change │
│ QQuery 17 │ 68.53 / 69.88 ±0.97 / 71.56 ms │  68.87 / 70.06 ±1.59 / 73.18 ms │ no change │
│ QQuery 18 │ 58.38 / 59.59 ±1.19 / 61.85 ms │  57.86 / 60.16 ±1.33 / 61.71 ms │ no change │
│ QQuery 19 │ 32.01 / 32.46 ±0.32 / 32.82 ms │  32.20 / 32.42 ±0.18 / 32.75 ms │ no change │
│ QQuery 20 │ 31.02 / 31.62 ±0.41 / 32.17 ms │  30.85 / 31.22 ±0.28 / 31.73 ms │ no change │
│ QQuery 21 │ 54.81 / 57.06 ±2.20 / 60.23 ms │  53.32 / 55.77 ±1.75 / 58.42 ms │ no change │
│ QQuery 22 │ 13.78 / 13.93 ±0.14 / 14.18 ms │  13.64 / 13.78 ±0.15 / 14.02 ms │ no change │
└───────────┴────────────────────────────────┴─────────────────────────────────┴───────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┓
┃ Benchmark Summary                              ┃          ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 748.04ms │
│ Total Time (feat_sort-merge-join-projection)   │ 745.38ms │
│ Average Time (HEAD)                            │  34.00ms │
│ Average Time (feat_sort-merge-join-projection) │  33.88ms │
│ 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	5.0s
Peak memory	1.2 GiB
Avg memory	527.6 MiB
CPU user	21.2s
CPU sys	1.6s
Peak spill	0 B

tpch — branch

Metric	Value
Wall time	5.0s
Peak memory	1.2 GiB
Avg memory	522.5 MiB
CPU user	21.2s
CPU sys	1.6s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-08-20T13:54:41Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing feat/sort-merge-join-projection (dc0e106) to 40c208e (merge-base) diff

Run configuration

run benchmark tpch10
env:
  PREFER_HASH_JOIN: "false"

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 feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf10.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃      HEAD ┃ feat_sort-merge-join-projection ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │ 306.81 ms │                       306.58 ms │     no change │
│ QQuery 2  │ 102.90 ms │                       101.90 ms │     no change │
│ QQuery 3  │ 229.37 ms │                       224.35 ms │     no change │
│ QQuery 4  │ 117.04 ms │                       111.35 ms │     no change │
│ QQuery 5  │ 352.56 ms │                       355.57 ms │     no change │
│ QQuery 6  │ 122.80 ms │                       132.57 ms │  1.08x slower │
│ QQuery 7  │ 482.89 ms │                       524.23 ms │  1.09x slower │
│ QQuery 8  │ 399.33 ms │                       375.35 ms │ +1.06x faster │
│ QQuery 9  │ 549.54 ms │                       550.42 ms │     no change │
│ QQuery 10 │ 292.42 ms │                       292.18 ms │     no change │
│ QQuery 11 │  84.73 ms │                        85.76 ms │     no change │
│ QQuery 12 │ 173.60 ms │                       175.57 ms │     no change │
│ QQuery 13 │ 295.13 ms │                       292.90 ms │     no change │
│ QQuery 14 │ 167.23 ms │                       165.45 ms │     no change │
│ QQuery 15 │ 293.68 ms │                       324.42 ms │  1.10x slower │
│ QQuery 16 │  65.34 ms │                        70.44 ms │  1.08x slower │
│ QQuery 17 │ 610.13 ms │                       608.98 ms │     no change │
│ QQuery 18 │ 704.00 ms │                       670.80 ms │     no change │
│ QQuery 19 │ 247.70 ms │                       237.11 ms │     no change │
│ QQuery 20 │ 271.84 ms │                       272.02 ms │     no change │
│ QQuery 21 │ 641.19 ms │                       650.37 ms │     no change │
│ QQuery 22 │  59.14 ms │                        64.67 ms │  1.09x slower │
└───────────┴───────────┴─────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 6569.38ms │
│ Total Time (feat_sort-merge-join-projection)   │ 6592.98ms │
│ Average Time (HEAD)                            │  298.61ms │
│ Average Time (feat_sort-merge-join-projection) │  299.68ms │
│ Queries Faster                                 │         1 │
│ Queries Slower                                 │         5 │
│ Queries with No Change                         │        16 │
│ Queries with Failure                           │         0 │
└────────────────────────────────────────────────┴───────────┘

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

Comparing HEAD and feat_sort-merge-join-projection
--------------------
Benchmark tpch_sf10.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                               HEAD ┃    feat_sort-merge-join-projection ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │  306.81 / 307.40 ±0.40 / 308.00 ms │  306.58 / 307.53 ±0.74 / 308.59 ms │     no change │
│ QQuery 2  │  102.90 / 105.12 ±1.61 / 106.88 ms │  101.90 / 104.85 ±2.73 / 108.08 ms │     no change │
│ QQuery 3  │  229.37 / 238.34 ±6.15 / 246.09 ms │  224.35 / 228.44 ±4.32 / 235.46 ms │     no change │
│ QQuery 4  │  117.04 / 118.75 ±0.95 / 119.96 ms │  111.35 / 112.82 ±1.07 / 114.02 ms │     no change │
│ QQuery 5  │  352.56 / 362.59 ±6.67 / 369.52 ms │ 355.57 / 370.68 ±16.50 / 392.90 ms │     no change │
│ QQuery 6  │  122.80 / 124.40 ±2.25 / 128.80 ms │  132.57 / 136.13 ±5.08 / 146.19 ms │  1.09x slower │
│ QQuery 7  │ 482.89 / 530.02 ±41.34 / 601.65 ms │ 524.23 / 550.15 ±18.07 / 572.16 ms │     no change │
│ QQuery 8  │ 399.33 / 414.53 ±11.46 / 433.88 ms │  375.35 / 383.90 ±7.67 / 395.62 ms │ +1.08x faster │
│ QQuery 9  │ 549.54 / 560.63 ±11.20 / 581.38 ms │  550.42 / 560.85 ±8.72 / 576.56 ms │     no change │
│ QQuery 10 │  292.42 / 299.04 ±5.54 / 306.84 ms │  292.18 / 302.52 ±8.27 / 313.08 ms │     no change │
│ QQuery 11 │     84.73 / 89.68 ±4.65 / 98.16 ms │     85.76 / 89.44 ±3.96 / 96.33 ms │     no change │
│ QQuery 12 │  173.60 / 180.47 ±7.91 / 195.94 ms │  175.57 / 182.40 ±8.40 / 198.78 ms │     no change │
│ QQuery 13 │  295.13 / 301.87 ±6.99 / 314.50 ms │  292.90 / 300.89 ±6.24 / 309.02 ms │     no change │
│ QQuery 14 │  167.23 / 172.87 ±5.91 / 181.88 ms │  165.45 / 180.68 ±8.53 / 190.09 ms │     no change │
│ QQuery 15 │  293.68 / 295.92 ±1.93 / 298.50 ms │  324.42 / 328.86 ±2.53 / 332.08 ms │  1.11x slower │
│ QQuery 16 │     65.34 / 66.51 ±1.20 / 68.57 ms │     70.44 / 73.65 ±2.32 / 77.05 ms │  1.11x slower │
│ QQuery 17 │ 610.13 / 639.19 ±42.36 / 722.16 ms │ 608.98 / 648.22 ±27.33 / 679.04 ms │     no change │
│ QQuery 18 │ 704.00 / 796.45 ±52.67 / 838.83 ms │ 670.80 / 691.21 ±11.58 / 706.57 ms │ +1.15x faster │
│ QQuery 19 │ 247.70 / 262.78 ±15.79 / 285.67 ms │ 237.11 / 255.56 ±20.09 / 294.28 ms │     no change │
│ QQuery 20 │  271.84 / 285.90 ±9.70 / 299.31 ms │ 272.02 / 286.78 ±12.07 / 307.28 ms │     no change │
│ QQuery 21 │  641.19 / 652.13 ±7.63 / 662.70 ms │ 650.37 / 687.56 ±42.28 / 762.41 ms │  1.05x slower │
│ QQuery 22 │     59.14 / 60.52 ±1.07 / 62.01 ms │     64.67 / 70.62 ±6.32 / 82.87 ms │  1.17x slower │
└───────────┴────────────────────────────────────┴────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃           ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 6865.11ms │
│ Total Time (feat_sort-merge-join-projection)   │ 6853.72ms │
│ Average Time (HEAD)                            │  312.05ms │
│ Average Time (feat_sort-merge-join-projection) │  311.53ms │
│ Queries Faster                                 │         2 │
│ Queries Slower                                 │         5 │
│ Queries with No Change                         │        15 │
│ Queries with Failure                           │         0 │
└────────────────────────────────────────────────┴───────────┘

Resource Usage

tpch10 — base (merge-base)

Metric	Value
Wall time	35.0s
Peak memory	4.8 GiB
Avg memory	1.5 GiB
CPU user	346.9s
CPU sys	19.4s
Peak spill	0 B

tpch10 — branch

Metric	Value
Wall time	35.0s
Peak memory	5.0 GiB
Avg memory	1.6 GiB
CPU user	344.3s
CPU sys	20.3s
Peak spill	0 B

File an issue against this benchmark runner

adriangbot · 2026-08-20T13:56:38Z

🤖 Benchmark completed (GKE) | trigger

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

Comparing feat/sort-merge-join-projection (dc0e106) to 40c208e (merge-base) diff

Run configuration

run benchmark tpcds
env:
  PREFER_HASH_JOIN: "false"

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 feat_sort-merge-join-projection
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃       HEAD ┃ feat_sort-merge-join-projection ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │   23.76 ms │                        23.67 ms │     no change │
│ QQuery 2  │   79.47 ms │                        78.62 ms │     no change │
│ QQuery 3  │   50.73 ms │                        50.67 ms │     no change │
│ QQuery 4  │  797.06 ms │                       799.51 ms │     no change │
│ QQuery 5  │  104.69 ms │                       105.71 ms │     no change │
│ QQuery 6  │   70.06 ms │                        69.41 ms │     no change │
│ QQuery 7  │  121.61 ms │                       120.35 ms │     no change │
│ QQuery 8  │   66.10 ms │                        66.09 ms │     no change │
│ QQuery 9  │   51.28 ms │                        50.66 ms │     no change │
│ QQuery 10 │  134.47 ms │                       133.35 ms │     no change │
│ QQuery 11 │  418.78 ms │                       414.30 ms │     no change │
│ QQuery 12 │   32.07 ms │                        31.63 ms │     no change │
│ QQuery 13 │  123.51 ms │                       122.59 ms │     no change │
│ QQuery 14 │  813.26 ms │                       810.58 ms │     no change │
│ QQuery 15 │   54.44 ms │                        54.06 ms │     no change │
│ QQuery 16 │   77.66 ms │                        78.06 ms │     no change │
│ QQuery 17 │  167.09 ms │                       164.99 ms │     no change │
│ QQuery 18 │  148.71 ms │                       148.19 ms │     no change │
│ QQuery 19 │   74.97 ms │                        74.33 ms │     no change │
│ QQuery 20 │   45.05 ms │                        44.59 ms │     no change │
│ QQuery 21 │  382.29 ms │                       388.55 ms │     no change │
│ QQuery 22 │   99.10 ms │                        99.84 ms │     no change │
│ QQuery 23 │  552.35 ms │                       555.56 ms │     no change │
│ QQuery 24 │  344.49 ms │                       342.10 ms │     no change │
│ QQuery 25 │  187.93 ms │                       187.64 ms │     no change │
│ QQuery 26 │   99.53 ms │                        98.85 ms │     no change │
│ QQuery 27 │  123.22 ms │                       120.77 ms │     no change │
│ QQuery 28 │   58.49 ms │                        60.37 ms │     no change │
│ QQuery 29 │  167.23 ms │                       166.53 ms │     no change │
│ QQuery 30 │   46.70 ms │                        47.21 ms │     no change │
│ QQuery 31 │  235.93 ms │                       237.47 ms │     no change │
│ QQuery 32 │   67.31 ms │                        67.34 ms │     no change │
│ QQuery 33 │   70.51 ms │                        71.51 ms │     no change │
│ QQuery 34 │   69.29 ms │                        69.24 ms │     no change │
│ QQuery 35 │  123.65 ms │                       122.25 ms │     no change │
│ QQuery 36 │   85.14 ms │                        84.73 ms │     no change │
│ QQuery 37 │   32.38 ms │                        32.30 ms │     no change │
│ QQuery 38 │   97.04 ms │                        96.69 ms │     no change │
│ QQuery 39 │ 1611.77 ms │                      1608.70 ms │     no change │
│ QQuery 40 │  102.58 ms │                       102.37 ms │     no change │
│ QQuery 41 │   11.32 ms │                        11.37 ms │     no change │
│ QQuery 42 │   49.33 ms │                        48.86 ms │     no change │
│ QQuery 43 │   56.82 ms │                        56.67 ms │     no change │
│ QQuery 44 │   12.35 ms │                        12.39 ms │     no change │
│ QQuery 45 │   46.67 ms │                        46.15 ms │     no change │
│ QQuery 46 │  119.13 ms │                       119.22 ms │     no change │
│ QQuery 47 │  355.71 ms │                       361.32 ms │     no change │
│ QQuery 48 │  120.20 ms │                       123.21 ms │     no change │
│ QQuery 49 │   79.40 ms │                        79.70 ms │     no change │
│ QQuery 50 │   87.27 ms │                        87.76 ms │     no change │
│ QQuery 51 │  121.36 ms │                       121.74 ms │     no change │
│ QQuery 52 │   49.76 ms │                        50.35 ms │     no change │
│ QQuery 53 │   53.64 ms │                        54.97 ms │     no change │
│ QQuery 54 │   96.60 ms │                        98.13 ms │     no change │
│ QQuery 55 │   47.28 ms │                        47.41 ms │     no change │
│ QQuery 56 │   69.75 ms │                        71.80 ms │     no change │
│ QQuery 57 │  217.25 ms │                       222.14 ms │     no change │
│ QQuery 58 │  185.69 ms │                       187.54 ms │     no change │
│ QQuery 59 │  121.12 ms │                       121.56 ms │     no change │
│ QQuery 60 │   71.61 ms │                        71.84 ms │     no change │
│ QQuery 61 │  101.93 ms │                       100.29 ms │     no change │
│ QQuery 62 │   65.24 ms │                        65.80 ms │     no change │
│ QQuery 63 │   54.06 ms │                        54.47 ms │     no change │
│ QQuery 64 │  658.54 ms │                       664.94 ms │     no change │
│ QQuery 65 │  111.39 ms │                       110.47 ms │     no change │
│ QQuery 66 │  167.41 ms │                       165.82 ms │     no change │
│ QQuery 67 │  135.26 ms │                       134.62 ms │     no change │
│ QQuery 68 │  126.05 ms │                       126.13 ms │     no change │
│ QQuery 69 │  126.74 ms │                       125.50 ms │     no change │
│ QQuery 70 │  203.93 ms │                       202.38 ms │     no change │
│ QQuery 71 │   66.45 ms │                        66.74 ms │     no change │
│ QQuery 72 │ 6496.62 ms │                      6454.29 ms │     no change │
│ QQuery 73 │   67.07 ms │                        67.92 ms │     no change │
│ QQuery 74 │  274.88 ms │                       276.48 ms │     no change │
│ QQuery 75 │  210.66 ms │                       208.49 ms │     no change │
│ QQuery 76 │   38.54 ms │                        38.22 ms │     no change │
│ QQuery 77 │  106.13 ms │                       105.82 ms │     no change │
│ QQuery 78 │  283.53 ms │                       282.13 ms │     no change │
│ QQuery 79 │  122.25 ms │                       118.65 ms │     no change │
│ QQuery 80 │  234.11 ms │                       226.22 ms │     no change │
│ QQuery 81 │   46.79 ms │                        45.35 ms │     no change │
│ QQuery 82 │   45.61 ms │                        44.60 ms │     no change │
│ QQuery 83 │   54.14 ms │                        54.11 ms │     no change │
│ QQuery 84 │   46.93 ms │                        44.89 ms │     no change │
│ QQuery 85 │  137.14 ms │                       130.00 ms │ +1.05x faster │
│ QQuery 86 │   27.27 ms │                        27.32 ms │     no change │
│ QQuery 87 │   99.94 ms │                        99.93 ms │     no change │
│ QQuery 88 │  164.87 ms │                       163.02 ms │     no change │
│ QQuery 89 │   59.91 ms │                        59.45 ms │     no change │
│ QQuery 90 │   26.47 ms │                        26.25 ms │     no change │
│ QQuery 91 │   52.64 ms │                        51.59 ms │     no change │
│ QQuery 92 │   50.79 ms │                        47.68 ms │ +1.07x faster │
│ QQuery 93 │  219.03 ms │                       211.52 ms │     no change │
│ QQuery 94 │   63.45 ms │                        58.33 ms │ +1.09x faster │
│ QQuery 95 │  164.60 ms │                       155.37 ms │ +1.06x faster │
│ QQuery 96 │   45.13 ms │                        45.84 ms │     no change │
│ QQuery 97 │   80.30 ms │                        79.66 ms │     no change │
│ QQuery 98 │   65.81 ms │                        64.48 ms │     no change │
│ QQuery 99 │  139.58 ms │                       138.22 ms │     no change │
└───────────┴────────────┴─────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 21225.19ms │
│ Total Time (feat_sort-merge-join-projection)   │ 21138.54ms │
│ Average Time (HEAD)                            │   214.40ms │
│ Average Time (feat_sort-merge-join-projection) │   213.52ms │
│ Queries Faster                                 │          4 │
│ Queries Slower                                 │          0 │
│ Queries with No Change                         │         95 │
│ Queries with Failure                           │          0 │
└────────────────────────────────────────────────┴────────────┘

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

Comparing HEAD and feat_sort-merge-join-projection
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                   HEAD ┃        feat_sort-merge-join-projection ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │         23.76 / 24.48 ±1.00 / 26.45 ms │         23.67 / 24.33 ±1.15 / 26.63 ms │     no change │
│ QQuery 2  │         79.47 / 79.84 ±0.37 / 80.49 ms │         78.62 / 79.01 ±0.33 / 79.53 ms │     no change │
│ QQuery 3  │         50.73 / 51.17 ±0.26 / 51.49 ms │         50.67 / 50.75 ±0.05 / 50.82 ms │     no change │
│ QQuery 4  │      797.06 / 808.03 ±8.87 / 819.24 ms │     799.51 / 808.88 ±11.45 / 829.89 ms │     no change │
│ QQuery 5  │      104.69 / 111.34 ±8.30 / 125.21 ms │      105.71 / 109.45 ±4.67 / 118.63 ms │     no change │
│ QQuery 6  │         70.06 / 70.42 ±0.30 / 70.88 ms │         69.41 / 70.08 ±0.72 / 71.49 ms │     no change │
│ QQuery 7  │      121.61 / 125.36 ±3.22 / 130.24 ms │      120.35 / 124.21 ±2.81 / 129.13 ms │     no change │
│ QQuery 8  │         66.10 / 66.36 ±0.20 / 66.64 ms │         66.09 / 66.19 ±0.09 / 66.35 ms │     no change │
│ QQuery 9  │         51.28 / 53.42 ±1.25 / 54.70 ms │         50.66 / 52.25 ±1.34 / 54.15 ms │     no change │
│ QQuery 10 │     134.47 / 144.64 ±10.00 / 162.96 ms │     133.35 / 143.23 ±13.16 / 168.11 ms │     no change │
│ QQuery 11 │      418.78 / 428.75 ±8.59 / 443.95 ms │     414.30 / 426.79 ±14.36 / 454.68 ms │     no change │
│ QQuery 12 │         32.07 / 32.26 ±0.13 / 32.42 ms │         31.63 / 32.16 ±0.31 / 32.55 ms │     no change │
│ QQuery 13 │      123.51 / 129.03 ±9.26 / 147.41 ms │      122.59 / 127.58 ±7.15 / 141.79 ms │     no change │
│ QQuery 14 │      813.26 / 826.46 ±9.89 / 839.12 ms │     810.58 / 828.22 ±13.23 / 847.00 ms │     no change │
│ QQuery 15 │         54.44 / 56.33 ±2.59 / 61.46 ms │         54.06 / 54.59 ±0.53 / 55.50 ms │     no change │
│ QQuery 16 │         77.66 / 80.74 ±3.29 / 85.03 ms │         78.06 / 83.61 ±7.47 / 98.39 ms │     no change │
│ QQuery 17 │      167.09 / 169.76 ±2.19 / 173.18 ms │      164.99 / 170.58 ±4.18 / 176.32 ms │     no change │
│ QQuery 18 │      148.71 / 153.75 ±3.36 / 157.67 ms │      148.19 / 151.43 ±3.76 / 158.62 ms │     no change │
│ QQuery 19 │         74.97 / 78.25 ±3.56 / 84.66 ms │         74.33 / 78.46 ±3.80 / 85.37 ms │     no change │
│ QQuery 20 │         45.05 / 45.48 ±0.35 / 45.95 ms │         44.59 / 45.05 ±0.33 / 45.40 ms │     no change │
│ QQuery 21 │      382.29 / 386.92 ±3.07 / 390.40 ms │      388.55 / 392.19 ±2.22 / 395.05 ms │     no change │
│ QQuery 22 │       99.10 / 103.05 ±7.40 / 117.85 ms │       99.84 / 101.44 ±2.83 / 107.08 ms │     no change │
│ QQuery 23 │     552.35 / 568.60 ±11.48 / 582.70 ms │      555.56 / 571.22 ±8.72 / 580.06 ms │     no change │
│ QQuery 24 │      344.49 / 351.15 ±6.52 / 362.84 ms │      342.10 / 349.73 ±6.19 / 360.96 ms │     no change │
│ QQuery 25 │      187.93 / 191.94 ±4.34 / 197.89 ms │      187.64 / 192.04 ±4.52 / 198.88 ms │     no change │
│ QQuery 26 │       99.53 / 104.60 ±3.92 / 110.40 ms │       98.85 / 106.98 ±8.51 / 119.89 ms │     no change │
│ QQuery 27 │      123.22 / 126.27 ±3.17 / 130.70 ms │      120.77 / 124.29 ±2.53 / 128.55 ms │     no change │
│ QQuery 28 │         58.49 / 62.43 ±2.99 / 67.63 ms │         60.37 / 65.40 ±8.26 / 81.78 ms │     no change │
│ QQuery 29 │      167.23 / 169.51 ±2.47 / 174.11 ms │      166.53 / 169.03 ±2.64 / 172.66 ms │     no change │
│ QQuery 30 │         46.70 / 50.32 ±2.72 / 54.12 ms │         47.21 / 51.52 ±6.28 / 63.90 ms │     no change │
│ QQuery 31 │      235.93 / 241.11 ±5.78 / 252.08 ms │      237.47 / 243.19 ±5.79 / 253.92 ms │     no change │
│ QQuery 32 │         67.31 / 68.13 ±0.71 / 69.12 ms │         67.34 / 68.56 ±0.86 / 69.64 ms │     no change │
│ QQuery 33 │         70.51 / 73.83 ±4.88 / 83.48 ms │         71.51 / 74.81 ±4.02 / 82.53 ms │     no change │
│ QQuery 34 │         69.29 / 69.91 ±0.50 / 70.81 ms │         69.24 / 70.21 ±0.70 / 71.24 ms │     no change │
│ QQuery 35 │      123.65 / 128.62 ±7.64 / 143.82 ms │      122.25 / 128.41 ±7.99 / 144.21 ms │     no change │
│ QQuery 36 │         85.14 / 88.23 ±3.08 / 92.62 ms │         84.73 / 87.73 ±2.47 / 90.76 ms │     no change │
│ QQuery 37 │         32.38 / 32.84 ±0.37 / 33.43 ms │         32.30 / 32.75 ±0.43 / 33.56 ms │     no change │
│ QQuery 38 │       97.04 / 101.93 ±7.41 / 116.38 ms │       96.69 / 101.65 ±7.41 / 116.34 ms │     no change │
│ QQuery 39 │   1611.77 / 1617.55 ±7.04 / 1631.01 ms │  1608.70 / 1623.24 ±10.89 / 1639.37 ms │     no change │
│ QQuery 40 │      102.58 / 103.13 ±0.78 / 104.65 ms │      102.37 / 103.26 ±0.66 / 104.24 ms │     no change │
│ QQuery 41 │         11.32 / 11.50 ±0.28 / 12.05 ms │         11.37 / 11.58 ±0.29 / 12.13 ms │     no change │
│ QQuery 42 │         49.33 / 49.42 ±0.09 / 49.58 ms │         48.86 / 53.24 ±5.87 / 64.40 ms │  1.08x slower │
│ QQuery 43 │         56.82 / 59.28 ±1.95 / 61.34 ms │         56.67 / 57.08 ±0.40 / 57.82 ms │     no change │
│ QQuery 44 │         12.35 / 12.58 ±0.30 / 13.18 ms │         12.39 / 12.60 ±0.34 / 13.28 ms │     no change │
│ QQuery 45 │         46.67 / 47.05 ±0.28 / 47.53 ms │         46.15 / 46.51 ±0.23 / 46.74 ms │     no change │
│ QQuery 46 │      119.13 / 123.03 ±2.90 / 125.86 ms │      119.22 / 124.42 ±3.63 / 129.79 ms │     no change │
│ QQuery 47 │      355.71 / 361.72 ±4.34 / 368.00 ms │     361.32 / 376.25 ±10.74 / 389.38 ms │     no change │
│ QQuery 48 │      120.20 / 129.16 ±7.77 / 142.47 ms │      123.21 / 126.10 ±3.17 / 130.20 ms │     no change │
│ QQuery 49 │         79.40 / 80.52 ±0.87 / 81.62 ms │         79.70 / 82.56 ±2.14 / 86.23 ms │     no change │
│ QQuery 50 │         87.27 / 90.81 ±4.17 / 98.69 ms │         87.76 / 88.50 ±0.73 / 89.85 ms │     no change │
│ QQuery 51 │      121.36 / 124.49 ±4.57 / 133.55 ms │      121.74 / 127.28 ±6.39 / 138.83 ms │     no change │
│ QQuery 52 │         49.76 / 53.78 ±5.21 / 64.03 ms │         50.35 / 51.35 ±1.01 / 53.31 ms │     no change │
│ QQuery 53 │         53.64 / 54.41 ±0.66 / 55.48 ms │         54.97 / 55.11 ±0.17 / 55.45 ms │     no change │
│ QQuery 54 │        96.60 / 98.31 ±2.69 / 103.67 ms │       98.13 / 103.94 ±5.65 / 110.98 ms │  1.06x slower │
│ QQuery 55 │         47.28 / 47.59 ±0.23 / 47.98 ms │         47.41 / 48.60 ±0.74 / 49.58 ms │     no change │
│ QQuery 56 │         69.75 / 71.54 ±0.98 / 72.56 ms │         71.80 / 73.70 ±1.36 / 75.49 ms │     no change │
│ QQuery 57 │      217.25 / 220.61 ±3.09 / 224.73 ms │      222.14 / 229.98 ±8.63 / 242.49 ms │     no change │
│ QQuery 58 │      185.69 / 191.82 ±6.85 / 204.79 ms │      187.54 / 192.96 ±4.10 / 199.20 ms │     no change │
│ QQuery 59 │      121.12 / 124.89 ±4.04 / 131.73 ms │      121.56 / 125.33 ±5.83 / 136.90 ms │     no change │
│ QQuery 60 │         71.61 / 73.34 ±1.13 / 74.89 ms │         71.84 / 73.43 ±1.53 / 76.31 ms │     no change │
│ QQuery 61 │     101.93 / 110.67 ±11.65 / 133.25 ms │      100.29 / 107.21 ±5.57 / 114.32 ms │     no change │
│ QQuery 62 │         65.24 / 66.35 ±0.88 / 67.89 ms │         65.80 / 66.41 ±0.57 / 67.49 ms │     no change │
│ QQuery 63 │         54.06 / 54.40 ±0.24 / 54.74 ms │         54.47 / 60.15 ±7.10 / 73.47 ms │  1.11x slower │
│ QQuery 64 │      658.54 / 662.19 ±3.56 / 668.50 ms │      664.94 / 673.83 ±6.48 / 683.06 ms │     no change │
│ QQuery 65 │      111.39 / 111.95 ±0.72 / 113.35 ms │      110.47 / 114.15 ±6.86 / 127.86 ms │     no change │
│ QQuery 66 │      167.41 / 169.68 ±2.52 / 173.31 ms │      165.82 / 172.30 ±8.12 / 188.14 ms │     no change │
│ QQuery 67 │      135.26 / 136.48 ±1.92 / 140.28 ms │      134.62 / 136.10 ±1.67 / 138.48 ms │     no change │
│ QQuery 68 │      126.05 / 127.84 ±1.71 / 129.90 ms │      126.13 / 132.41 ±9.54 / 151.08 ms │     no change │
│ QQuery 69 │      126.74 / 134.25 ±9.64 / 152.24 ms │      125.50 / 126.32 ±0.69 / 127.55 ms │ +1.06x faster │
│ QQuery 70 │      203.93 / 208.36 ±6.36 / 220.85 ms │      202.38 / 206.82 ±4.40 / 215.01 ms │     no change │
│ QQuery 71 │         66.45 / 66.91 ±0.30 / 67.24 ms │         66.74 / 67.74 ±1.31 / 70.28 ms │     no change │
│ QQuery 72 │ 6496.62 / 6738.37 ±176.17 / 6937.59 ms │ 6454.29 / 6728.45 ±165.50 / 6905.29 ms │     no change │
│ QQuery 73 │      67.07 / 151.80 ±96.50 / 320.72 ms │      67.92 / 123.71 ±81.68 / 279.43 ms │ +1.23x faster │
│ QQuery 74 │     274.88 / 335.78 ±97.50 / 529.83 ms │    276.48 / 358.20 ±138.09 / 633.60 ms │  1.07x slower │
│ QQuery 75 │      210.66 / 219.13 ±8.54 / 233.64 ms │      208.49 / 219.93 ±6.46 / 227.71 ms │     no change │
│ QQuery 76 │         38.54 / 41.47 ±3.53 / 47.87 ms │        38.22 / 44.10 ±10.98 / 66.04 ms │  1.06x slower │
│ QQuery 77 │      106.13 / 110.88 ±3.75 / 115.51 ms │      105.82 / 109.42 ±3.25 / 114.99 ms │     no change │
│ QQuery 78 │     283.53 / 304.78 ±12.45 / 321.52 ms │     282.13 / 301.64 ±19.30 / 326.54 ms │     no change │
│ QQuery 79 │      122.25 / 128.67 ±5.98 / 137.51 ms │      118.65 / 123.98 ±2.96 / 127.83 ms │     no change │
│ QQuery 80 │      234.11 / 245.07 ±8.42 / 255.92 ms │     226.22 / 243.14 ±17.44 / 276.38 ms │     no change │
│ QQuery 81 │         46.79 / 50.65 ±4.07 / 57.37 ms │        45.35 / 56.35 ±12.58 / 77.39 ms │  1.11x slower │
│ QQuery 82 │         45.61 / 50.15 ±5.43 / 59.87 ms │         44.60 / 46.12 ±1.29 / 48.37 ms │ +1.09x faster │
│ QQuery 83 │         54.14 / 55.06 ±0.71 / 56.08 ms │         54.11 / 56.07 ±2.01 / 59.26 ms │     no change │
│ QQuery 84 │         46.93 / 49.52 ±3.49 / 56.36 ms │         44.89 / 45.82 ±1.00 / 47.64 ms │ +1.08x faster │
│ QQuery 85 │      137.14 / 143.12 ±4.40 / 150.66 ms │     130.00 / 139.08 ±10.17 / 158.23 ms │     no change │
│ QQuery 86 │         27.27 / 30.04 ±3.34 / 36.05 ms │         27.32 / 29.32 ±1.70 / 31.37 ms │     no change │
│ QQuery 87 │       99.94 / 103.79 ±2.96 / 107.73 ms │       99.93 / 105.29 ±8.02 / 121.12 ms │     no change │
│ QQuery 88 │      164.87 / 171.37 ±4.51 / 176.66 ms │      163.02 / 167.58 ±5.65 / 176.98 ms │     no change │
│ QQuery 89 │         59.91 / 62.29 ±2.04 / 65.67 ms │         59.45 / 61.18 ±1.35 / 62.62 ms │     no change │
│ QQuery 90 │         26.47 / 26.73 ±0.29 / 27.30 ms │         26.25 / 26.57 ±0.43 / 27.41 ms │     no change │
│ QQuery 91 │         52.64 / 54.03 ±1.41 / 56.40 ms │         51.59 / 52.34 ±0.85 / 53.89 ms │     no change │
│ QQuery 92 │         50.79 / 54.72 ±3.05 / 59.35 ms │        47.68 / 55.32 ±13.23 / 81.73 ms │     no change │
│ QQuery 93 │     219.03 / 231.44 ±11.21 / 252.44 ms │     211.52 / 224.19 ±12.84 / 246.44 ms │     no change │
│ QQuery 94 │         63.45 / 66.13 ±1.57 / 68.00 ms │         58.33 / 62.82 ±2.68 / 66.16 ms │ +1.05x faster │
│ QQuery 95 │      164.60 / 172.70 ±8.21 / 187.61 ms │     155.37 / 171.15 ±14.25 / 197.66 ms │     no change │
│ QQuery 96 │         45.13 / 46.19 ±0.64 / 47.14 ms │         45.84 / 46.85 ±1.08 / 48.68 ms │     no change │
│ QQuery 97 │         80.30 / 83.20 ±2.84 / 87.44 ms │         79.66 / 81.44 ±1.57 / 84.20 ms │     no change │
│ QQuery 98 │         65.81 / 68.85 ±2.22 / 71.84 ms │         64.48 / 67.67 ±2.76 / 71.26 ms │     no change │
│ QQuery 99 │      139.58 / 148.92 ±6.78 / 157.72 ms │      138.22 / 139.43 ±1.14 / 141.43 ms │ +1.07x faster │
└───────────┴────────────────────────────────────────┴────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                              ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                              │ 21995.71ms │
│ Total Time (feat_sort-merge-join-projection)   │ 21995.64ms │
│ Average Time (HEAD)                            │   222.18ms │
│ Average Time (feat_sort-merge-join-projection) │   222.18ms │
│ Queries Faster                                 │          6 │
│ Queries Slower                                 │          6 │
│ Queries with No Change                         │         87 │
│ Queries with Failure                           │          0 │
└────────────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric	Value
Wall time	115.0s
Peak memory	26.1 GiB
Avg memory	3.8 GiB
CPU user	615.0s
CPU sys	29.9s
Peak spill	0 B

tpcds — branch

Metric	Value
Wall time	115.0s
Peak memory	24.2 GiB
Avg memory	3.8 GiB
CPU user	616.1s
CPU sys	30.7s
Peak spill	0 B

File an issue against this benchmark runner

github-actions Bot added core Core DataFusion crate sqllogictest SQL Logic Tests (.slt) proto Related to proto crate physical-plan Changes to the physical-plan crate labels Aug 20, 2026

Dandandan force-pushed the feat/sort-merge-join-projection branch from 65415ac to 2d55b49 Compare August 20, 2026 11:35

github-actions Bot added the auto detected api change Auto detected API change label Aug 20, 2026

Dandandan marked this pull request as ready for review August 20, 2026 12:11

jayzhan211 reviewed Aug 20, 2026

View reviewed changes

Comment thread datafusion/physical-plan/src/joins/sort_merge_join/exec.rs

Comment thread datafusion/physical-plan/src/joins/sort_merge_join/exec.rs

Comment thread datafusion/physical-plan/src/joins/sort_merge_join/exec.rs Outdated

Read the projection field directly instead of through a public helper

dd3a41e

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012YiAABcW4WSqij31zz2P6c

Dandandan mentioned this pull request Aug 21, 2026

feat: cost-based join order enumeration #24456

Draft

Uh oh!

Conversation

Dandandan commented Aug 20, 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!

Dandandan commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

github-actions Bot commented Aug 20, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

Dandandan commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

codecov-commenter commented Aug 20, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

jayzhan211 left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Dandandan commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

adriangbot commented Aug 20, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Dandandan commented Aug 20, 2026 •

edited

Loading

github-actions Bot commented Aug 20, 2026 •

edited

Loading

codecov-commenter commented Aug 20, 2026 •

edited

Loading