Skip to content

Prefetch morsels across files in FileStream (bounded at 20)#21682

Closed
Dandandan wants to merge 4 commits intoapache:mainfrom
Dandandan:prefetch-morsels-on-21351
Closed

Prefetch morsels across files in FileStream (bounded at 20)#21682
Dandandan wants to merge 4 commits intoapache:mainfrom
Dandandan:prefetch-morsels-on-21351

Conversation

@Dandandan
Copy link
Copy Markdown
Contributor

Which issue does this PR close?

Stacks on top of #21351.

Rationale for this change

PR #21351 enables dynamic work scheduling in FileStream but keeps the same single-outstanding-I/O-per-partition property as main. This PR implements the follow-on item @alamb listed:

  1. Trying to issue multiple IOs by the same partition (aka to interleave IO and CPU work more)

It lets each partition prefetch upcoming files while the active reader decodes the current file, so planner I/O is no longer serialized within a partition.

What changes are included in this PR?

  1. New FileStreamState::Prefetch variant and PrefetchState that drives multiple PendingMorselPlanner I/Os concurrently and issues planner I/O for upcoming files while the active reader is blocked.
  2. Prefetching is bounded at MAX_PREFETCH_MORSELS = 20 in-flight morsel-producing work items (pending I/O + ready planners + ready morsels + active reader) to cap buffering.
  3. Enabled by default via FileStreamBuilder; the legacy single-I/O ScanState path is preserved and opt-in-able via FileStreamBuilder::with_prefetch(false).
  4. Two new snapshot tests:
    • morsel_prefetch_overlaps_io_across_files — verifies file2's planner I/O is issued while file1's I/O is still pending.
    • morsel_no_prefetch_keeps_files_sequential — verifies with_prefetch(false) preserves the legacy single-I/O behavior.

The reader takes priority over prefetching (step order: poll pending I/O → poll reader → plan → promote morsel → morselize next file), so user-visible latency is not delayed by opening new files, and all existing snapshot tests pass unchanged.

Are these changes tested?

Yes — 27 file_stream tests pass, including the two new prefetch-specific tests. Full datafusion-datasource and datafusion crate test suites pass locally. Clippy is clean on the affected crates.

Are there any user-facing changes?

Yes — prefetching is on by default, so multi-file scans may now have multiple planner I/Os in flight per partition. Users can opt out via FileStreamBuilder::with_prefetch(false).

🤖 Generated with Claude Code

alamb and others added 4 commits April 14, 2026 11:49
Co-authored-by: Oleks V <comphead@users.noreply.github.com>
Adds a `FileStreamState::Prefetch` variant that drives multiple planner
I/O operations concurrently, so I/O for upcoming files overlaps with
CPU decoding of the current file. In-flight morsel-producing work
(pending I/O + ready planners + ready morsels + active reader) is
capped at 20 to bound buffering.

Enabled by default via `FileStreamBuilder` (legacy single-I/O
`ScanState` remains available with `with_prefetch(false)`). Stacks on
top of apache#21351.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@Dandandan
Copy link
Copy Markdown
Contributor Author

run benchmarks

@github-actions github-actions bot added the datasource Changes to the datasource crate label Apr 16, 2026
@adriangbot
Copy link
Copy Markdown

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4262256588-1390-n9xxl 6.12.55+ #1 SMP Sun Feb 1 08:59:41 UTC 2026 aarch64 GNU/Linux

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

Comparing prefetch-morsels-on-21351 (2165d79) to 961c5fc (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete


File an issue against this benchmark runner

@adriangbot
Copy link
Copy Markdown

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4262256588-1392-xlxfx 6.12.55+ #1 SMP Sun Feb 1 08:59:41 UTC 2026 aarch64 GNU/Linux

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

Comparing prefetch-morsels-on-21351 (2165d79) to 961c5fc (merge-base) diff using: tpch
Results will be posted here when complete


File an issue against this benchmark runner

@adriangbot
Copy link
Copy Markdown

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4262256588-1391-k8mwr 6.12.55+ #1 SMP Sun Feb 1 08:59:41 UTC 2026 aarch64 GNU/Linux

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

Comparing prefetch-morsels-on-21351 (2165d79) to 961c5fc (merge-base) diff using: tpcds
Results will be posted here when complete


File an issue against this benchmark runner

@Dandandan
Copy link
Copy Markdown
Contributor Author

run benchmarks

@adriangbot
Copy link
Copy Markdown

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4262276962-1393-4qdjk 6.12.55+ #1 SMP Sun Feb 1 08:59:41 UTC 2026 aarch64 GNU/Linux

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

Comparing prefetch-morsels-on-21351 (2165d79) to 961c5fc (merge-base) diff using: clickbench_partitioned
Results will be posted here when complete


File an issue against this benchmark runner

@adriangbot
Copy link
Copy Markdown

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4262276962-1395-t7gr5 6.12.55+ #1 SMP Sun Feb 1 08:59:41 UTC 2026 aarch64 GNU/Linux

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

Comparing prefetch-morsels-on-21351 (2165d79) to 961c5fc (merge-base) diff using: tpch
Results will be posted here when complete


File an issue against this benchmark runner

@adriangbot
Copy link
Copy Markdown

🤖 Benchmark running (GKE) | trigger
Instance: c4a-highmem-16 (12 vCPU / 65 GiB) | Linux bench-c4262276962-1394-hr64f 6.12.55+ #1 SMP Sun Feb 1 08:59:41 UTC 2026 aarch64 GNU/Linux

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

Comparing prefetch-morsels-on-21351 (2165d79) to 961c5fc (merge-base) diff using: tpcds
Results will be posted here when complete


File an issue against this benchmark runner

@adriangbot
Copy link
Copy Markdown

Benchmark for this request failed.

Last 20 lines of output:

Click to expand
BENCHMARK: tpch
QUERY: All
DATAFUSION_DIR: /workspace/datafusion-base
BRANCH_NAME: HEAD
DATA_DIR: /workspace/datafusion-bench/benchmarks/data
RESULTS_DIR: /workspace/datafusion-bench/benchmarks/results/HEAD
CARGO_COMMAND: cargo run --release
PREFER_HASH_JOIN: true
SIMULATE_LATENCY: false
***************************
RESULTS_FILE: /workspace/datafusion-bench/benchmarks/results/HEAD/tpch_sf1.json
Running tpch benchmark...
+ cargo run --release --bin dfbench -- tpch --iterations 5 --path /workspace/datafusion-bench/benchmarks/data/tpch_sf1 --scale-factor 1 --prefer_hash_join true --format parquet -o /workspace/datafusion-bench/benchmarks/results/HEAD/tpch_sf1.json
    Finished `release` profile [optimized] target(s) in 0.22s
     Running `/workspace/datafusion-base/target/release/dfbench tpch --iterations 5 --path /workspace/datafusion-bench/benchmarks/data/tpch_sf1 --scale-factor 1 --prefer_hash_join true --format parquet -o /workspace/datafusion-bench/benchmarks/results/HEAD/tpch_sf1.json`
error: unexpected argument '--scale-factor' found

Usage: dfbench tpch --path <PATH> --iterations <ITERATIONS>

For more information, try '--help'.

File an issue against this benchmark runner

@adriangbot
Copy link
Copy Markdown

Benchmark for this request failed.

Last 20 lines of output:

Click to expand
BENCHMARK: tpch
QUERY: All
DATAFUSION_DIR: /workspace/datafusion-base
BRANCH_NAME: HEAD
DATA_DIR: /workspace/datafusion-bench/benchmarks/data
RESULTS_DIR: /workspace/datafusion-bench/benchmarks/results/HEAD
CARGO_COMMAND: cargo run --release
PREFER_HASH_JOIN: true
SIMULATE_LATENCY: false
***************************
RESULTS_FILE: /workspace/datafusion-bench/benchmarks/results/HEAD/tpch_sf1.json
Running tpch benchmark...
+ cargo run --release --bin dfbench -- tpch --iterations 5 --path /workspace/datafusion-bench/benchmarks/data/tpch_sf1 --scale-factor 1 --prefer_hash_join true --format parquet -o /workspace/datafusion-bench/benchmarks/results/HEAD/tpch_sf1.json
    Finished `release` profile [optimized] target(s) in 0.22s
     Running `/workspace/datafusion-base/target/release/dfbench tpch --iterations 5 --path /workspace/datafusion-bench/benchmarks/data/tpch_sf1 --scale-factor 1 --prefer_hash_join true --format parquet -o /workspace/datafusion-bench/benchmarks/results/HEAD/tpch_sf1.json`
error: unexpected argument '--scale-factor' found

Usage: dfbench tpch --path <PATH> --iterations <ITERATIONS>

For more information, try '--help'.

File an issue against this benchmark runner

@adriangbot
Copy link
Copy Markdown

🤖 Benchmark completed (GKE) | trigger

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

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

Comparing HEAD and prefetch-morsels-on-21351
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃             prefetch-morsels-on-21351 ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.21 / 4.39 ±6.29 / 16.96 ms │          1.27 / 4.65 ±6.56 / 17.78 ms │  1.06x slower │
│ QQuery 1  │        14.41 / 14.59 ±0.12 / 14.73 ms │        12.74 / 12.96 ±0.19 / 13.18 ms │ +1.13x faster │
│ QQuery 2  │        43.66 / 43.98 ±0.24 / 44.29 ms │        36.24 / 37.17 ±0.52 / 37.86 ms │ +1.18x faster │
│ QQuery 3  │        42.25 / 45.21 ±2.13 / 47.33 ms │        33.77 / 35.60 ±1.40 / 37.40 ms │ +1.27x faster │
│ QQuery 4  │     286.07 / 290.56 ±3.48 / 295.15 ms │     271.10 / 282.06 ±7.46 / 292.83 ms │     no change │
│ QQuery 5  │     337.63 / 341.59 ±4.37 / 347.55 ms │     313.01 / 328.27 ±8.66 / 335.46 ms │     no change │
│ QQuery 6  │           5.62 / 6.21 ±0.32 / 6.55 ms │          7.09 / 8.77 ±2.02 / 12.71 ms │  1.41x slower │
│ QQuery 7  │        16.91 / 17.41 ±0.61 / 18.60 ms │        13.73 / 14.01 ±0.18 / 14.24 ms │ +1.24x faster │
│ QQuery 8  │     401.42 / 415.94 ±7.80 / 424.79 ms │    396.62 / 421.56 ±14.65 / 440.35 ms │     no change │
│ QQuery 9  │    625.25 / 642.85 ±11.95 / 654.29 ms │    544.94 / 574.08 ±16.12 / 594.70 ms │ +1.12x faster │
│ QQuery 10 │        90.72 / 92.59 ±2.40 / 97.26 ms │        79.18 / 80.55 ±1.21 / 82.71 ms │ +1.15x faster │
│ QQuery 11 │     103.81 / 104.42 ±0.48 / 105.17 ms │        86.81 / 89.13 ±1.67 / 90.87 ms │ +1.17x faster │
│ QQuery 12 │     336.91 / 343.15 ±4.51 / 349.22 ms │     355.18 / 365.22 ±8.25 / 376.56 ms │  1.06x slower │
│ QQuery 13 │     448.07 / 464.98 ±9.71 / 474.51 ms │    448.54 / 514.21 ±42.44 / 581.09 ms │  1.11x slower │
│ QQuery 14 │     335.10 / 338.88 ±3.23 / 344.44 ms │     356.85 / 363.13 ±3.92 / 368.90 ms │  1.07x slower │
│ QQuery 15 │    342.21 / 362.23 ±15.59 / 389.58 ms │     402.91 / 411.03 ±6.24 / 417.05 ms │  1.13x slower │
│ QQuery 16 │    699.85 / 714.26 ±10.53 / 731.39 ms │    793.24 / 805.85 ±12.91 / 829.30 ms │  1.13x slower │
│ QQuery 17 │     702.87 / 704.61 ±1.76 / 707.28 ms │    766.03 / 784.20 ±11.43 / 796.22 ms │  1.11x slower │
│ QQuery 18 │ 1406.04 / 1466.04 ±30.16 / 1486.55 ms │ 1466.58 / 1555.87 ±49.22 / 1616.92 ms │  1.06x slower │
│ QQuery 19 │        35.38 / 36.53 ±1.72 / 39.92 ms │        30.28 / 31.97 ±1.55 / 34.63 ms │ +1.14x faster │
│ QQuery 20 │    717.93 / 735.10 ±14.47 / 753.39 ms │    520.08 / 531.60 ±10.64 / 550.67 ms │ +1.38x faster │
│ QQuery 21 │     763.65 / 771.31 ±7.67 / 781.18 ms │     595.97 / 601.95 ±3.65 / 605.97 ms │ +1.28x faster │
│ QQuery 22 │  1126.15 / 1134.00 ±6.24 / 1143.26 ms │  1057.35 / 1059.75 ±2.03 / 1062.60 ms │ +1.07x faster │
│ QQuery 23 │ 3049.37 / 3082.77 ±20.07 / 3106.74 ms │ 3353.45 / 3446.12 ±61.22 / 3510.60 ms │  1.12x slower │
│ QQuery 24 │      98.77 / 100.43 ±0.93 / 101.32 ms │        44.96 / 48.43 ±3.07 / 52.36 ms │ +2.07x faster │
│ QQuery 25 │     138.81 / 139.96 ±1.18 / 142.14 ms │     115.71 / 117.61 ±1.03 / 118.53 ms │ +1.19x faster │
│ QQuery 26 │      99.77 / 100.26 ±0.51 / 101.09 ms │        45.37 / 45.92 ±0.45 / 46.70 ms │ +2.18x faster │
│ QQuery 27 │    849.75 / 863.82 ±13.50 / 883.67 ms │     678.38 / 688.84 ±6.95 / 699.63 ms │ +1.25x faster │
│ QQuery 28 │ 3257.85 / 3313.96 ±38.45 / 3371.24 ms │ 3050.19 / 3118.74 ±44.01 / 3171.93 ms │ +1.06x faster │
│ QQuery 29 │        52.49 / 56.55 ±3.54 / 62.68 ms │        44.19 / 50.03 ±4.97 / 57.08 ms │ +1.13x faster │
│ QQuery 30 │     366.33 / 378.54 ±9.81 / 390.51 ms │     351.49 / 355.83 ±5.94 / 367.46 ms │ +1.06x faster │
│ QQuery 31 │    366.58 / 390.41 ±14.62 / 409.22 ms │     341.22 / 350.06 ±7.71 / 364.34 ms │ +1.12x faster │
│ QQuery 32 │ 1269.35 / 1355.87 ±54.79 / 1412.36 ms │ 1221.24 / 1325.95 ±80.38 / 1458.99 ms │     no change │
│ QQuery 33 │ 1652.19 / 1721.24 ±65.33 / 1820.53 ms │ 1588.07 / 1626.96 ±22.83 / 1657.40 ms │ +1.06x faster │
│ QQuery 34 │ 1696.23 / 1717.59 ±19.12 / 1753.37 ms │ 1615.30 / 1640.47 ±25.53 / 1681.11 ms │     no change │
│ QQuery 35 │    446.99 / 465.57 ±14.99 / 486.73 ms │    368.57 / 384.67 ±14.93 / 407.83 ms │ +1.21x faster │
│ QQuery 36 │     125.82 / 128.08 ±1.35 / 129.63 ms │        64.07 / 71.65 ±4.94 / 78.67 ms │ +1.79x faster │
│ QQuery 37 │        48.19 / 51.18 ±2.33 / 53.28 ms │        37.84 / 40.92 ±1.62 / 42.61 ms │ +1.25x faster │
│ QQuery 38 │        76.42 / 80.16 ±2.95 / 84.83 ms │        42.94 / 44.97 ±1.38 / 47.01 ms │ +1.78x faster │
│ QQuery 39 │     207.71 / 220.52 ±8.80 / 231.26 ms │     141.88 / 148.13 ±3.49 / 151.24 ms │ +1.49x faster │
│ QQuery 40 │        24.98 / 27.04 ±1.24 / 28.87 ms │        18.94 / 20.96 ±1.95 / 23.30 ms │ +1.29x faster │
│ QQuery 41 │        22.13 / 23.73 ±1.87 / 27.17 ms │        14.40 / 15.72 ±0.77 / 16.76 ms │ +1.51x faster │
│ QQuery 42 │        21.18 / 22.19 ±0.99 / 23.95 ms │        13.91 / 14.38 ±0.38 / 14.89 ms │ +1.54x faster │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                        ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                        │ 23330.69ms │
│ Total Time (prefetch-morsels-on-21351)   │ 22469.98ms │
│ Average Time (HEAD)                      │   542.57ms │
│ Average Time (prefetch-morsels-on-21351) │   522.56ms │
│ Queries Faster                           │         28 │
│ Queries Slower                           │         10 │
│ Queries with No Change                   │          5 │
│ Queries with Failure                     │          0 │
└──────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric Value
Wall time 117.8s
Peak memory 38.3 GiB
Avg memory 28.7 GiB
CPU user 1099.6s
CPU sys 100.8s
Peak spill 0 B

clickbench_partitioned — branch

Metric Value
Wall time 113.6s
Peak memory 41.4 GiB
Avg memory 29.8 GiB
CPU user 1161.8s
CPU sys 102.4s
Peak spill 0 B

File an issue against this benchmark runner

@adriangbot
Copy link
Copy Markdown

🤖 Benchmark completed (GKE) | trigger

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

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

Comparing HEAD and prefetch-morsels-on-21351
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃                prefetch-morsels-on-21351 ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │              6.74 / 7.22 ±0.81 / 8.84 ms │              6.71 / 7.22 ±0.83 / 8.87 ms │     no change │
│ QQuery 2  │        144.37 / 146.07 ±0.92 / 146.92 ms │        144.30 / 145.12 ±0.59 / 145.97 ms │     no change │
│ QQuery 3  │        114.25 / 114.69 ±0.43 / 115.42 ms │        113.38 / 113.87 ±0.27 / 114.09 ms │     no change │
│ QQuery 4  │    1371.84 / 1411.06 ±25.12 / 1450.40 ms │    1365.89 / 1411.49 ±26.06 / 1442.46 ms │     no change │
│ QQuery 5  │        171.72 / 173.97 ±1.27 / 175.51 ms │        171.98 / 173.45 ±0.97 / 174.97 ms │     no change │
│ QQuery 6  │       847.38 / 884.68 ±19.81 / 904.64 ms │       837.60 / 877.43 ±44.45 / 962.32 ms │     no change │
│ QQuery 7  │        342.91 / 346.12 ±1.75 / 347.73 ms │        342.65 / 345.63 ±1.75 / 347.16 ms │     no change │
│ QQuery 8  │        115.94 / 116.58 ±0.76 / 118.05 ms │        116.83 / 118.58 ±1.19 / 120.32 ms │     no change │
│ QQuery 9  │        101.27 / 108.40 ±6.79 / 117.08 ms │        100.56 / 108.38 ±6.51 / 116.67 ms │     no change │
│ QQuery 10 │        106.66 / 107.91 ±1.03 / 109.33 ms │        106.01 / 107.18 ±0.79 / 108.31 ms │     no change │
│ QQuery 11 │      990.01 / 999.79 ±10.76 / 1019.39 ms │     994.02 / 1013.30 ±14.23 / 1029.27 ms │     no change │
│ QQuery 12 │           46.27 / 47.67 ±0.89 / 48.77 ms │           45.39 / 47.56 ±1.44 / 48.95 ms │     no change │
│ QQuery 13 │        405.88 / 407.65 ±1.48 / 409.60 ms │        408.80 / 410.96 ±1.50 / 412.67 ms │     no change │
│ QQuery 14 │     1011.78 / 1022.53 ±7.49 / 1032.46 ms │     1004.56 / 1017.98 ±8.05 / 1026.53 ms │     no change │
│ QQuery 15 │           16.17 / 16.86 ±0.76 / 18.33 ms │           15.97 / 17.15 ±0.95 / 18.37 ms │     no change │
│ QQuery 16 │              7.63 / 8.29 ±0.86 / 9.92 ms │              7.32 / 7.67 ±0.36 / 8.26 ms │ +1.08x faster │
│ QQuery 17 │        231.70 / 233.26 ±1.16 / 234.90 ms │        230.17 / 232.44 ±1.84 / 235.46 ms │     no change │
│ QQuery 18 │        129.69 / 130.77 ±0.82 / 131.78 ms │        127.83 / 129.02 ±1.21 / 130.59 ms │     no change │
│ QQuery 19 │        155.63 / 158.22 ±1.53 / 160.43 ms │        155.31 / 157.17 ±1.37 / 159.46 ms │     no change │
│ QQuery 20 │           14.49 / 14.96 ±0.38 / 15.55 ms │           14.44 / 14.87 ±0.49 / 15.54 ms │     no change │
│ QQuery 21 │           19.99 / 20.50 ±0.46 / 21.24 ms │           19.72 / 20.14 ±0.26 / 20.46 ms │     no change │
│ QQuery 22 │       491.10 / 504.32 ±11.38 / 525.24 ms │        498.59 / 510.59 ±9.73 / 528.32 ms │     no change │
│ QQuery 23 │        891.54 / 905.62 ±9.99 / 921.68 ms │        893.99 / 902.24 ±6.49 / 908.41 ms │     no change │
│ QQuery 24 │        387.79 / 390.46 ±2.56 / 394.88 ms │        388.39 / 391.48 ±3.36 / 396.78 ms │     no change │
│ QQuery 25 │        344.48 / 347.11 ±2.74 / 351.08 ms │        342.34 / 345.40 ±2.34 / 349.02 ms │     no change │
│ QQuery 26 │           83.62 / 84.77 ±0.83 / 86.09 ms │           79.93 / 82.27 ±1.43 / 84.13 ms │     no change │
│ QQuery 27 │              7.20 / 7.60 ±0.53 / 8.64 ms │              6.94 / 7.08 ±0.13 / 7.31 ms │ +1.07x faster │
│ QQuery 28 │        148.92 / 150.64 ±1.39 / 153.08 ms │        149.56 / 150.57 ±0.88 / 151.66 ms │     no change │
│ QQuery 29 │        282.50 / 285.61 ±1.91 / 287.69 ms │        281.55 / 284.69 ±1.92 / 286.88 ms │     no change │
│ QQuery 30 │           43.44 / 44.34 ±0.85 / 45.90 ms │           42.54 / 45.26 ±1.61 / 47.08 ms │     no change │
│ QQuery 31 │        173.27 / 174.30 ±1.31 / 176.81 ms │        171.10 / 172.36 ±1.08 / 173.95 ms │     no change │
│ QQuery 32 │           56.96 / 57.72 ±0.60 / 58.59 ms │           56.64 / 57.98 ±1.44 / 60.66 ms │     no change │
│ QQuery 33 │        140.85 / 143.63 ±2.16 / 147.18 ms │        140.78 / 142.82 ±1.67 / 145.79 ms │     no change │
│ QQuery 34 │              7.41 / 7.57 ±0.16 / 7.84 ms │              7.31 / 7.79 ±0.49 / 8.69 ms │     no change │
│ QQuery 35 │        107.24 / 109.52 ±1.42 / 111.56 ms │        105.51 / 108.53 ±1.59 / 109.97 ms │     no change │
│ QQuery 36 │              7.04 / 7.11 ±0.06 / 7.21 ms │              6.64 / 6.89 ±0.19 / 7.10 ms │     no change │
│ QQuery 37 │              8.59 / 9.13 ±0.42 / 9.66 ms │              8.55 / 9.10 ±0.40 / 9.53 ms │     no change │
│ QQuery 38 │           83.09 / 89.94 ±5.42 / 99.29 ms │           82.84 / 87.56 ±3.93 / 94.78 ms │     no change │
│ QQuery 39 │        127.89 / 131.83 ±3.02 / 136.55 ms │        125.52 / 128.69 ±2.42 / 132.83 ms │     no change │
│ QQuery 40 │        112.88 / 117.70 ±6.73 / 130.94 ms │       108.44 / 115.12 ±10.37 / 135.50 ms │     no change │
│ QQuery 41 │           14.81 / 15.61 ±0.72 / 16.60 ms │           14.17 / 14.81 ±0.60 / 15.85 ms │ +1.05x faster │
│ QQuery 42 │        109.24 / 110.27 ±0.67 / 111.26 ms │        106.35 / 107.74 ±1.46 / 110.01 ms │     no change │
│ QQuery 43 │              6.21 / 6.64 ±0.60 / 7.78 ms │              5.92 / 6.27 ±0.31 / 6.77 ms │ +1.06x faster │
│ QQuery 44 │           12.08 / 12.58 ±0.80 / 14.16 ms │           11.67 / 11.90 ±0.17 / 12.12 ms │ +1.06x faster │
│ QQuery 45 │           51.20 / 51.72 ±0.52 / 52.56 ms │           50.43 / 51.47 ±0.81 / 52.58 ms │     no change │
│ QQuery 46 │              8.87 / 9.32 ±0.37 / 9.83 ms │              8.48 / 8.77 ±0.27 / 9.27 ms │ +1.06x faster │
│ QQuery 47 │        749.06 / 760.83 ±9.66 / 770.95 ms │       719.60 / 733.62 ±10.19 / 750.35 ms │     no change │
│ QQuery 48 │        290.73 / 294.43 ±2.29 / 296.89 ms │        292.00 / 294.48 ±2.15 / 297.21 ms │     no change │
│ QQuery 49 │        254.04 / 254.99 ±0.70 / 256.10 ms │        250.71 / 253.32 ±2.29 / 256.81 ms │     no change │
│ QQuery 50 │        225.30 / 231.29 ±3.69 / 235.18 ms │        231.21 / 235.63 ±3.93 / 242.06 ms │     no change │
│ QQuery 51 │        181.07 / 185.82 ±3.36 / 189.16 ms │        183.79 / 185.36 ±1.67 / 187.96 ms │     no change │
│ QQuery 52 │        108.22 / 109.28 ±0.89 / 110.63 ms │        106.63 / 107.56 ±0.75 / 108.87 ms │     no change │
│ QQuery 53 │        102.45 / 103.36 ±0.60 / 104.24 ms │        102.22 / 103.83 ±1.18 / 104.99 ms │     no change │
│ QQuery 54 │        146.51 / 148.54 ±1.44 / 150.81 ms │        148.35 / 149.36 ±0.90 / 150.84 ms │     no change │
│ QQuery 55 │        108.94 / 109.88 ±0.79 / 111.24 ms │        106.41 / 108.30 ±1.13 / 109.80 ms │     no change │
│ QQuery 56 │        141.68 / 143.30 ±1.77 / 146.04 ms │        141.82 / 143.03 ±1.55 / 146.01 ms │     no change │
│ QQuery 57 │        177.38 / 179.23 ±1.63 / 182.16 ms │        176.20 / 178.25 ±1.91 / 181.71 ms │     no change │
│ QQuery 58 │        296.38 / 303.43 ±8.11 / 319.30 ms │       296.69 / 310.93 ±10.26 / 328.45 ms │     no change │
│ QQuery 59 │        201.14 / 202.54 ±1.26 / 204.63 ms │        197.96 / 201.52 ±2.99 / 205.48 ms │     no change │
│ QQuery 60 │        141.51 / 145.40 ±2.80 / 149.88 ms │        143.55 / 144.72 ±0.73 / 145.68 ms │     no change │
│ QQuery 61 │           13.84 / 14.32 ±0.47 / 15.15 ms │           13.17 / 13.49 ±0.31 / 14.08 ms │ +1.06x faster │
│ QQuery 62 │       903.75 / 937.93 ±26.64 / 977.79 ms │       900.19 / 925.77 ±20.11 / 960.86 ms │     no change │
│ QQuery 63 │        102.08 / 105.58 ±1.99 / 108.22 ms │        104.26 / 106.08 ±2.44 / 110.57 ms │     no change │
│ QQuery 64 │        691.17 / 698.69 ±6.83 / 711.36 ms │        693.19 / 697.90 ±4.30 / 705.40 ms │     no change │
│ QQuery 65 │        260.36 / 264.18 ±3.06 / 268.53 ms │        258.66 / 262.96 ±3.48 / 268.82 ms │     no change │
│ QQuery 66 │        244.32 / 260.22 ±8.30 / 266.98 ms │        251.20 / 255.17 ±4.08 / 262.26 ms │     no change │
│ QQuery 67 │        320.31 / 328.56 ±6.93 / 340.09 ms │        313.72 / 321.90 ±7.95 / 336.64 ms │     no change │
│ QQuery 68 │            9.79 / 11.55 ±1.20 / 13.40 ms │            9.42 / 10.29 ±0.46 / 10.71 ms │ +1.12x faster │
│ QQuery 69 │        104.37 / 107.58 ±4.11 / 115.38 ms │        103.85 / 105.86 ±1.95 / 109.14 ms │     no change │
│ QQuery 70 │       342.31 / 356.12 ±12.90 / 379.86 ms │       342.01 / 354.27 ±12.24 / 376.93 ms │     no change │
│ QQuery 71 │        132.85 / 136.13 ±2.59 / 139.04 ms │        133.21 / 137.06 ±2.81 / 141.27 ms │     no change │
│ QQuery 72 │        634.14 / 636.95 ±1.76 / 638.87 ms │        623.85 / 636.38 ±7.73 / 646.54 ms │     no change │
│ QQuery 73 │             6.84 / 8.22 ±1.38 / 10.43 ms │              7.21 / 8.82 ±1.02 / 9.89 ms │  1.07x slower │
│ QQuery 74 │        620.96 / 629.02 ±5.31 / 637.41 ms │       610.81 / 625.86 ±12.86 / 648.69 ms │     no change │
│ QQuery 75 │        277.04 / 281.08 ±2.30 / 283.79 ms │        278.81 / 280.40 ±1.25 / 281.83 ms │     no change │
│ QQuery 76 │        131.70 / 132.84 ±0.68 / 133.56 ms │        131.65 / 132.83 ±1.08 / 134.25 ms │     no change │
│ QQuery 77 │        188.75 / 190.06 ±0.82 / 191.31 ms │        186.77 / 189.71 ±1.87 / 192.34 ms │     no change │
│ QQuery 78 │        345.73 / 349.77 ±4.09 / 355.81 ms │        341.90 / 346.31 ±4.75 / 354.76 ms │     no change │
│ QQuery 79 │        232.95 / 238.37 ±3.02 / 240.98 ms │        236.48 / 238.02 ±0.85 / 239.06 ms │     no change │
│ QQuery 80 │        322.68 / 326.70 ±2.34 / 329.02 ms │        322.12 / 325.53 ±2.91 / 330.72 ms │     no change │
│ QQuery 81 │           26.48 / 27.48 ±0.70 / 28.64 ms │           27.32 / 28.81 ±2.26 / 33.30 ms │     no change │
│ QQuery 82 │        198.96 / 200.26 ±1.02 / 201.75 ms │        199.36 / 203.67 ±2.43 / 206.11 ms │     no change │
│ QQuery 83 │           39.09 / 40.50 ±1.04 / 42.08 ms │           39.17 / 40.05 ±0.73 / 41.12 ms │     no change │
│ QQuery 84 │           49.01 / 50.93 ±1.98 / 54.62 ms │           48.28 / 49.34 ±0.60 / 50.04 ms │     no change │
│ QQuery 85 │        148.83 / 150.40 ±1.91 / 154.11 ms │        147.87 / 150.34 ±1.80 / 153.01 ms │     no change │
│ QQuery 86 │           39.68 / 40.50 ±0.48 / 40.89 ms │           39.36 / 40.41 ±0.86 / 41.88 ms │     no change │
│ QQuery 87 │           85.18 / 90.72 ±4.07 / 96.79 ms │           86.58 / 88.63 ±2.44 / 93.42 ms │     no change │
│ QQuery 88 │        101.17 / 102.24 ±0.97 / 103.80 ms │        100.54 / 100.98 ±0.46 / 101.86 ms │     no change │
│ QQuery 89 │        118.04 / 120.14 ±1.39 / 121.50 ms │        119.48 / 121.63 ±1.68 / 124.42 ms │     no change │
│ QQuery 90 │           23.48 / 24.08 ±0.35 / 24.50 ms │           23.49 / 24.39 ±0.94 / 25.98 ms │     no change │
│ QQuery 91 │           62.20 / 64.96 ±1.93 / 67.56 ms │           63.43 / 64.20 ±0.64 / 65.03 ms │     no change │
│ QQuery 92 │           57.39 / 59.05 ±1.24 / 60.87 ms │           57.18 / 58.39 ±0.83 / 59.42 ms │     no change │
│ QQuery 93 │        189.86 / 191.45 ±1.02 / 193.02 ms │        185.08 / 189.16 ±2.53 / 192.44 ms │     no change │
│ QQuery 94 │           62.04 / 62.69 ±0.70 / 63.63 ms │           62.10 / 63.42 ±1.24 / 65.55 ms │     no change │
│ QQuery 95 │        129.35 / 131.28 ±1.97 / 134.78 ms │        128.92 / 130.19 ±1.27 / 132.44 ms │     no change │
│ QQuery 96 │           71.09 / 73.86 ±1.72 / 76.19 ms │           70.37 / 73.86 ±1.80 / 75.61 ms │     no change │
│ QQuery 97 │        127.05 / 128.07 ±1.30 / 130.61 ms │        125.62 / 128.24 ±1.43 / 129.92 ms │     no change │
│ QQuery 98 │        150.41 / 154.50 ±3.83 / 161.01 ms │        154.83 / 156.99 ±1.47 / 159.21 ms │     no change │
│ QQuery 99 │ 10772.77 / 10838.87 ±45.40 / 10913.54 ms │ 10774.68 / 10810.26 ±44.94 / 10894.73 ms │     no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                        ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                        │ 31990.45ms │
│ Total Time (prefetch-morsels-on-21351)   │ 31881.47ms │
│ Average Time (HEAD)                      │   323.14ms │
│ Average Time (prefetch-morsels-on-21351) │   322.04ms │
│ Queries Faster                           │          8 │
│ Queries Slower                           │          1 │
│ Queries with No Change                   │         90 │
│ Queries with Failure                     │          0 │
└──────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric Value
Wall time 160.3s
Peak memory 5.5 GiB
Avg memory 4.5 GiB
CPU user 265.3s
CPU sys 17.0s
Peak spill 0 B

tpcds — branch

Metric Value
Wall time 159.7s
Peak memory 5.3 GiB
Avg memory 4.5 GiB
CPU user 264.9s
CPU sys 17.3s
Peak spill 0 B

File an issue against this benchmark runner

@adriangbot
Copy link
Copy Markdown

🤖 Benchmark completed (GKE) | trigger

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

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

Comparing HEAD and prefetch-morsels-on-21351
--------------------
Benchmark clickbench_partitioned.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                  HEAD ┃             prefetch-morsels-on-21351 ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 0  │          1.22 / 4.50 ±6.42 / 17.34 ms │          1.20 / 4.48 ±6.41 / 17.29 ms │     no change │
│ QQuery 1  │        14.23 / 14.53 ±0.17 / 14.72 ms │        11.84 / 12.35 ±0.29 / 12.71 ms │ +1.18x faster │
│ QQuery 2  │        43.77 / 44.42 ±0.37 / 44.90 ms │        36.01 / 36.52 ±0.38 / 36.98 ms │ +1.22x faster │
│ QQuery 3  │        44.02 / 45.13 ±0.87 / 46.43 ms │        33.70 / 35.10 ±1.67 / 38.35 ms │ +1.29x faster │
│ QQuery 4  │     295.13 / 298.69 ±3.16 / 302.90 ms │     274.09 / 280.94 ±3.56 / 284.34 ms │ +1.06x faster │
│ QQuery 5  │     347.58 / 351.23 ±4.95 / 361.00 ms │     319.57 / 329.72 ±5.88 / 338.00 ms │ +1.07x faster │
│ QQuery 6  │           5.87 / 6.67 ±0.78 / 7.99 ms │         7.85 / 10.54 ±3.88 / 18.25 ms │  1.58x slower │
│ QQuery 7  │        16.61 / 17.14 ±0.57 / 18.02 ms │        14.45 / 15.03 ±0.64 / 16.20 ms │ +1.14x faster │
│ QQuery 8  │    422.99 / 436.96 ±10.44 / 447.84 ms │    399.50 / 419.92 ±12.88 / 434.63 ms │     no change │
│ QQuery 9  │    701.39 / 731.53 ±21.63 / 765.73 ms │    559.86 / 573.96 ±11.87 / 595.63 ms │ +1.27x faster │
│ QQuery 10 │       92.47 / 96.85 ±3.20 / 101.08 ms │        81.14 / 83.28 ±2.58 / 88.35 ms │ +1.16x faster │
│ QQuery 11 │     108.01 / 111.23 ±1.78 / 113.09 ms │        87.98 / 90.86 ±2.11 / 94.19 ms │ +1.22x faster │
│ QQuery 12 │    374.82 / 386.02 ±11.48 / 407.95 ms │     327.07 / 342.39 ±8.51 / 351.90 ms │ +1.13x faster │
│ QQuery 13 │     515.75 / 522.20 ±8.33 / 537.70 ms │    456.76 / 475.77 ±14.95 / 494.27 ms │ +1.10x faster │
│ QQuery 14 │     368.72 / 376.41 ±6.46 / 387.51 ms │     327.38 / 334.21 ±6.62 / 343.58 ms │ +1.13x faster │
│ QQuery 15 │    380.87 / 420.09 ±30.94 / 464.34 ms │    363.14 / 381.23 ±16.99 / 410.80 ms │ +1.10x faster │
│ QQuery 16 │    753.27 / 770.03 ±13.97 / 795.62 ms │     731.40 / 740.88 ±6.31 / 750.70 ms │     no change │
│ QQuery 17 │    779.57 / 814.83 ±33.16 / 872.91 ms │     722.71 / 732.48 ±7.85 / 746.55 ms │ +1.11x faster │
│ QQuery 18 │ 1517.69 / 1580.82 ±52.18 / 1648.89 ms │ 1448.84 / 1522.24 ±54.90 / 1578.98 ms │     no change │
│ QQuery 19 │        35.08 / 40.02 ±7.56 / 55.00 ms │       29.42 / 36.59 ±14.15 / 64.89 ms │ +1.09x faster │
│ QQuery 20 │    725.86 / 746.41 ±24.31 / 779.12 ms │     519.02 / 527.37 ±8.62 / 542.87 ms │ +1.42x faster │
│ QQuery 21 │    774.77 / 798.15 ±13.56 / 814.84 ms │     592.28 / 595.54 ±2.97 / 599.83 ms │ +1.34x faster │
│ QQuery 22 │  1163.13 / 1174.93 ±9.58 / 1187.80 ms │  1055.82 / 1060.58 ±3.79 / 1066.28 ms │ +1.11x faster │
│ QQuery 23 │ 3289.68 / 3300.76 ±11.18 / 3315.44 ms │ 3313.83 / 3341.92 ±36.25 / 3411.56 ms │     no change │
│ QQuery 24 │     103.76 / 108.55 ±4.08 / 115.90 ms │        42.59 / 44.28 ±1.09 / 45.98 ms │ +2.45x faster │
│ QQuery 25 │     139.12 / 146.71 ±4.94 / 153.53 ms │     112.46 / 113.98 ±0.83 / 114.82 ms │ +1.29x faster │
│ QQuery 26 │     101.29 / 104.79 ±2.23 / 108.06 ms │        42.93 / 44.43 ±1.53 / 46.75 ms │ +2.36x faster │
│ QQuery 27 │     879.76 / 888.13 ±6.10 / 898.78 ms │     662.86 / 666.65 ±2.73 / 670.56 ms │ +1.33x faster │
│ QQuery 28 │ 3381.48 / 3409.17 ±27.84 / 3456.35 ms │ 3012.20 / 3029.77 ±14.21 / 3043.23 ms │ +1.13x faster │
│ QQuery 29 │        52.39 / 55.85 ±2.84 / 60.83 ms │       43.23 / 50.58 ±11.48 / 73.41 ms │ +1.10x faster │
│ QQuery 30 │    387.07 / 412.01 ±14.36 / 425.92 ms │     321.67 / 328.67 ±6.57 / 339.86 ms │ +1.25x faster │
│ QQuery 31 │    405.73 / 424.26 ±10.14 / 432.33 ms │     324.98 / 328.89 ±2.06 / 330.78 ms │ +1.29x faster │
│ QQuery 32 │ 1161.61 / 1229.42 ±39.42 / 1280.42 ms │ 1058.47 / 1202.01 ±83.09 / 1295.76 ms │     no change │
│ QQuery 33 │ 1569.46 / 1625.18 ±58.88 / 1729.75 ms │ 1423.79 / 1439.28 ±14.39 / 1463.05 ms │ +1.13x faster │
│ QQuery 34 │  1520.95 / 1529.38 ±9.60 / 1546.98 ms │ 1403.39 / 1425.69 ±14.36 / 1444.50 ms │ +1.07x faster │
│ QQuery 35 │     405.65 / 412.93 ±5.20 / 420.20 ms │     310.19 / 316.76 ±9.70 / 336.06 ms │ +1.30x faster │
│ QQuery 36 │     121.61 / 123.74 ±1.10 / 124.65 ms │        63.45 / 68.01 ±3.24 / 72.46 ms │ +1.82x faster │
│ QQuery 37 │        50.65 / 52.19 ±1.33 / 53.86 ms │        37.42 / 39.11 ±1.05 / 40.52 ms │ +1.33x faster │
│ QQuery 38 │        76.02 / 77.98 ±1.90 / 81.37 ms │        41.41 / 42.39 ±0.71 / 43.56 ms │ +1.84x faster │
│ QQuery 39 │     220.04 / 224.93 ±7.40 / 239.43 ms │     125.83 / 130.67 ±3.55 / 135.62 ms │ +1.72x faster │
│ QQuery 40 │        24.58 / 27.33 ±1.85 / 30.35 ms │        14.78 / 17.13 ±1.39 / 18.97 ms │ +1.60x faster │
│ QQuery 41 │        20.69 / 21.95 ±1.29 / 24.24 ms │        13.64 / 15.18 ±1.25 / 17.45 ms │ +1.45x faster │
│ QQuery 42 │        19.22 / 20.16 ±0.60 / 20.68 ms │        12.98 / 13.24 ±0.22 / 13.55 ms │ +1.52x faster │
└───────────┴───────────────────────────────────────┴───────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                        ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                        │ 23984.21ms │
│ Total Time (prefetch-morsels-on-21351)   │ 21300.63ms │
│ Average Time (HEAD)                      │   557.77ms │
│ Average Time (prefetch-morsels-on-21351) │   495.36ms │
│ Queries Faster                           │         36 │
│ Queries Slower                           │          1 │
│ Queries with No Change                   │          6 │
│ Queries with Failure                     │          0 │
└──────────────────────────────────────────┴────────────┘

Resource Usage

clickbench_partitioned — base (merge-base)

Metric Value
Wall time 121.1s
Peak memory 39.7 GiB
Avg memory 29.7 GiB
CPU user 1137.9s
CPU sys 97.0s
Peak spill 0 B

clickbench_partitioned — branch

Metric Value
Wall time 107.7s
Peak memory 42.0 GiB
Avg memory 29.7 GiB
CPU user 1103.3s
CPU sys 93.8s
Peak spill 0 B

File an issue against this benchmark runner

@adriangbot
Copy link
Copy Markdown

🤖 Benchmark completed (GKE) | trigger

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

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

Comparing HEAD and prefetch-morsels-on-21351
--------------------
Benchmark tpcds_sf1.json
--------------------
┏━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓
┃ Query     ┃                                     HEAD ┃                prefetch-morsels-on-21351 ┃        Change ┃
┡━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩
│ QQuery 1  │              6.67 / 7.21 ±0.84 / 8.88 ms │              6.49 / 6.93 ±0.79 / 8.52 ms │     no change │
│ QQuery 2  │        145.82 / 146.67 ±0.62 / 147.69 ms │        143.83 / 145.13 ±1.12 / 146.70 ms │     no change │
│ QQuery 3  │        114.45 / 115.33 ±0.63 / 116.24 ms │        112.98 / 114.14 ±1.22 / 115.68 ms │     no change │
│ QQuery 4  │    1311.76 / 1359.14 ±46.19 / 1438.81 ms │    1268.53 / 1298.94 ±17.18 / 1319.48 ms │     no change │
│ QQuery 5  │        173.38 / 175.72 ±1.23 / 176.97 ms │        171.98 / 173.10 ±0.73 / 173.96 ms │     no change │
│ QQuery 6  │       826.90 / 860.13 ±22.22 / 892.70 ms │       830.82 / 849.97 ±15.47 / 870.81 ms │     no change │
│ QQuery 7  │        342.91 / 347.05 ±3.99 / 352.35 ms │        341.08 / 343.13 ±2.62 / 348.09 ms │     no change │
│ QQuery 8  │        115.66 / 117.98 ±1.39 / 119.25 ms │        113.91 / 116.63 ±1.66 / 118.98 ms │     no change │
│ QQuery 9  │        101.38 / 109.87 ±9.36 / 127.63 ms │        101.06 / 105.69 ±3.15 / 109.17 ms │     no change │
│ QQuery 10 │        105.51 / 107.62 ±1.28 / 109.49 ms │        105.79 / 106.57 ±0.91 / 108.28 ms │     no change │
│ QQuery 11 │      959.84 / 975.47 ±19.05 / 1008.89 ms │       870.22 / 885.39 ±10.01 / 896.50 ms │ +1.10x faster │
│ QQuery 12 │           45.86 / 46.80 ±0.81 / 48.30 ms │           45.15 / 46.31 ±1.47 / 49.14 ms │     no change │
│ QQuery 13 │        400.88 / 405.72 ±3.00 / 409.32 ms │        395.94 / 399.64 ±2.44 / 402.77 ms │     no change │
│ QQuery 14 │      999.98 / 1007.11 ±6.30 / 1016.38 ms │       990.97 / 997.19 ±4.32 / 1003.89 ms │     no change │
│ QQuery 15 │           15.11 / 16.18 ±0.83 / 17.38 ms │           14.86 / 15.47 ±0.80 / 16.97 ms │     no change │
│ QQuery 16 │              7.11 / 7.65 ±0.53 / 8.64 ms │              7.17 / 8.36 ±0.85 / 9.27 ms │  1.09x slower │
│ QQuery 17 │        226.53 / 229.17 ±2.53 / 233.86 ms │        227.45 / 232.60 ±3.70 / 237.20 ms │     no change │
│ QQuery 18 │        127.17 / 129.50 ±1.42 / 130.80 ms │        125.05 / 129.66 ±3.03 / 133.05 ms │     no change │
│ QQuery 19 │        156.09 / 158.21 ±1.63 / 160.39 ms │        153.67 / 155.83 ±1.76 / 157.98 ms │     no change │
│ QQuery 20 │           13.61 / 14.77 ±0.86 / 15.98 ms │           13.01 / 13.41 ±0.37 / 13.99 ms │ +1.10x faster │
│ QQuery 21 │           18.98 / 19.77 ±0.61 / 20.78 ms │           18.76 / 19.27 ±0.43 / 20.00 ms │     no change │
│ QQuery 22 │        486.97 / 488.77 ±1.30 / 489.97 ms │        486.12 / 490.47 ±2.94 / 494.19 ms │     no change │
│ QQuery 23 │       866.45 / 885.02 ±13.48 / 902.34 ms │       851.82 / 878.15 ±18.88 / 904.65 ms │     no change │
│ QQuery 24 │        382.35 / 387.03 ±4.06 / 393.74 ms │        380.63 / 382.83 ±2.52 / 387.73 ms │     no change │
│ QQuery 25 │        340.51 / 344.18 ±2.30 / 347.23 ms │        339.61 / 340.28 ±0.61 / 341.14 ms │     no change │
│ QQuery 26 │           82.18 / 84.29 ±1.78 / 86.62 ms │           78.83 / 81.62 ±2.47 / 84.96 ms │     no change │
│ QQuery 27 │              6.97 / 7.63 ±0.58 / 8.68 ms │              6.87 / 7.33 ±0.53 / 8.34 ms │     no change │
│ QQuery 28 │        148.43 / 150.34 ±2.34 / 154.84 ms │        150.37 / 151.95 ±0.96 / 153.00 ms │     no change │
│ QQuery 29 │        282.26 / 284.17 ±2.08 / 288.20 ms │        283.63 / 285.81 ±1.91 / 288.77 ms │     no change │
│ QQuery 30 │           42.06 / 44.08 ±1.31 / 45.96 ms │           43.18 / 44.88 ±0.94 / 45.96 ms │     no change │
│ QQuery 31 │        169.47 / 172.81 ±2.95 / 177.96 ms │        168.58 / 171.72 ±2.16 / 175.07 ms │     no change │
│ QQuery 32 │           57.30 / 58.84 ±1.38 / 61.32 ms │           55.91 / 56.99 ±0.73 / 57.85 ms │     no change │
│ QQuery 33 │        142.06 / 143.62 ±0.85 / 144.67 ms │        139.12 / 140.90 ±0.93 / 141.79 ms │     no change │
│ QQuery 34 │              6.85 / 7.09 ±0.23 / 7.52 ms │              6.79 / 7.09 ±0.22 / 7.37 ms │     no change │
│ QQuery 35 │        107.73 / 110.35 ±1.42 / 111.94 ms │        107.29 / 108.29 ±1.04 / 110.27 ms │     no change │
│ QQuery 36 │              6.48 / 6.66 ±0.17 / 6.89 ms │              6.60 / 6.98 ±0.25 / 7.34 ms │     no change │
│ QQuery 37 │              8.34 / 8.63 ±0.25 / 9.08 ms │              8.03 / 8.77 ±0.49 / 9.42 ms │     no change │
│ QQuery 38 │           86.42 / 88.73 ±2.55 / 93.37 ms │           82.83 / 86.00 ±5.05 / 96.07 ms │     no change │
│ QQuery 39 │        125.87 / 128.94 ±3.38 / 135.04 ms │        121.79 / 124.67 ±1.46 / 125.86 ms │     no change │
│ QQuery 40 │        109.21 / 115.97 ±6.64 / 128.49 ms │        109.43 / 114.74 ±8.58 / 131.83 ms │     no change │
│ QQuery 41 │           14.43 / 15.61 ±1.32 / 18.13 ms │           13.88 / 14.99 ±0.89 / 16.12 ms │     no change │
│ QQuery 42 │        109.73 / 110.57 ±0.70 / 111.36 ms │        106.31 / 107.77 ±0.94 / 109.23 ms │     no change │
│ QQuery 43 │              6.11 / 6.28 ±0.16 / 6.59 ms │              5.95 / 6.06 ±0.16 / 6.38 ms │     no change │
│ QQuery 44 │           12.00 / 12.50 ±0.50 / 13.41 ms │           11.52 / 11.93 ±0.30 / 12.30 ms │     no change │
│ QQuery 45 │           51.19 / 51.55 ±0.26 / 52.00 ms │           50.08 / 50.56 ±0.30 / 50.88 ms │     no change │
│ QQuery 46 │              8.58 / 8.84 ±0.30 / 9.32 ms │             8.25 / 8.88 ±0.70 / 10.24 ms │     no change │
│ QQuery 47 │        689.50 / 695.90 ±6.04 / 704.14 ms │        680.92 / 693.46 ±8.96 / 708.58 ms │     no change │
│ QQuery 48 │        284.50 / 291.62 ±4.63 / 296.12 ms │        281.49 / 286.68 ±4.36 / 294.06 ms │     no change │
│ QQuery 49 │        251.30 / 254.55 ±3.66 / 260.84 ms │        248.87 / 252.85 ±3.87 / 259.78 ms │     no change │
│ QQuery 50 │        218.83 / 228.33 ±5.40 / 233.52 ms │        213.13 / 222.73 ±7.07 / 234.72 ms │     no change │
│ QQuery 51 │        180.18 / 183.55 ±3.27 / 188.53 ms │        179.66 / 181.44 ±1.77 / 184.77 ms │     no change │
│ QQuery 52 │        107.49 / 108.97 ±1.24 / 111.03 ms │        107.40 / 108.46 ±1.39 / 111.20 ms │     no change │
│ QQuery 53 │        103.77 / 104.67 ±0.69 / 105.74 ms │        102.48 / 103.60 ±0.71 / 104.48 ms │     no change │
│ QQuery 54 │        145.50 / 146.77 ±1.17 / 148.39 ms │        144.14 / 146.53 ±1.36 / 147.76 ms │     no change │
│ QQuery 55 │        105.87 / 107.82 ±1.71 / 110.96 ms │        105.60 / 107.15 ±1.05 / 108.62 ms │     no change │
│ QQuery 56 │        141.84 / 143.29 ±1.17 / 144.76 ms │        139.09 / 141.46 ±1.29 / 142.57 ms │     no change │
│ QQuery 57 │        170.33 / 175.09 ±2.58 / 177.58 ms │        172.65 / 173.58 ±0.77 / 174.59 ms │     no change │
│ QQuery 58 │        288.24 / 295.15 ±4.41 / 300.00 ms │        291.30 / 299.40 ±8.10 / 314.63 ms │     no change │
│ QQuery 59 │        197.64 / 199.19 ±1.51 / 202.06 ms │        195.64 / 199.18 ±2.61 / 202.65 ms │     no change │
│ QQuery 60 │        141.60 / 142.78 ±0.94 / 144.33 ms │        141.82 / 144.01 ±1.96 / 147.40 ms │     no change │
│ QQuery 61 │           12.83 / 13.17 ±0.29 / 13.55 ms │           12.90 / 13.18 ±0.26 / 13.57 ms │     no change │
│ QQuery 62 │       879.72 / 923.52 ±38.44 / 971.97 ms │       883.70 / 915.48 ±25.11 / 956.80 ms │     no change │
│ QQuery 63 │        102.66 / 105.31 ±1.51 / 106.61 ms │        104.38 / 105.70 ±1.27 / 108.06 ms │     no change │
│ QQuery 64 │        681.32 / 686.89 ±3.13 / 690.68 ms │        676.10 / 684.07 ±5.90 / 693.70 ms │     no change │
│ QQuery 65 │        249.60 / 252.30 ±1.89 / 254.71 ms │        245.98 / 253.59 ±5.49 / 259.41 ms │     no change │
│ QQuery 66 │       251.35 / 267.01 ±15.04 / 288.78 ms │       238.82 / 254.81 ±12.26 / 276.60 ms │     no change │
│ QQuery 67 │        313.53 / 317.69 ±3.96 / 323.20 ms │       302.59 / 314.09 ±10.08 / 332.18 ms │     no change │
│ QQuery 68 │            8.57 / 10.25 ±0.96 / 11.55 ms │            9.97 / 11.40 ±0.90 / 12.40 ms │  1.11x slower │
│ QQuery 69 │        101.84 / 104.34 ±1.73 / 107.09 ms │        101.08 / 102.33 ±1.34 / 104.27 ms │     no change │
│ QQuery 70 │        341.23 / 354.43 ±9.91 / 367.87 ms │        344.18 / 355.47 ±9.65 / 371.83 ms │     no change │
│ QQuery 71 │        134.59 / 136.27 ±1.35 / 137.52 ms │        134.62 / 138.73 ±6.13 / 150.92 ms │     no change │
│ QQuery 72 │       592.68 / 614.88 ±12.90 / 632.12 ms │        611.51 / 619.32 ±9.56 / 636.21 ms │     no change │
│ QQuery 73 │              6.51 / 7.21 ±0.43 / 7.76 ms │              7.22 / 8.21 ±0.71 / 9.31 ms │  1.14x slower │
│ QQuery 74 │       559.43 / 577.20 ±20.18 / 608.14 ms │        567.70 / 579.65 ±7.07 / 586.96 ms │     no change │
│ QQuery 75 │        277.84 / 279.42 ±1.10 / 280.59 ms │        274.82 / 278.45 ±2.57 / 282.39 ms │     no change │
│ QQuery 76 │        134.35 / 135.73 ±0.95 / 136.71 ms │        133.26 / 134.22 ±0.83 / 135.26 ms │     no change │
│ QQuery 77 │        190.69 / 192.08 ±0.90 / 193.01 ms │        186.80 / 189.27 ±1.88 / 191.68 ms │     no change │
│ QQuery 78 │        341.98 / 349.42 ±6.76 / 360.47 ms │        338.77 / 341.44 ±1.57 / 343.05 ms │     no change │
│ QQuery 79 │        232.11 / 236.48 ±3.34 / 241.28 ms │        228.20 / 232.44 ±2.77 / 236.56 ms │     no change │
│ QQuery 80 │        322.59 / 325.61 ±2.82 / 330.30 ms │        319.84 / 321.68 ±1.96 / 325.20 ms │     no change │
│ QQuery 81 │           26.92 / 28.18 ±1.00 / 29.97 ms │           26.43 / 27.89 ±1.17 / 29.38 ms │     no change │
│ QQuery 82 │        200.73 / 204.35 ±2.26 / 207.48 ms │        202.35 / 203.27 ±0.75 / 204.44 ms │     no change │
│ QQuery 83 │           38.31 / 39.63 ±0.83 / 40.92 ms │           38.67 / 39.67 ±0.55 / 40.29 ms │     no change │
│ QQuery 84 │           49.45 / 50.88 ±0.80 / 51.66 ms │           47.67 / 48.77 ±1.26 / 51.15 ms │     no change │
│ QQuery 85 │        145.71 / 149.66 ±2.54 / 153.55 ms │        146.12 / 147.33 ±0.89 / 148.54 ms │     no change │
│ QQuery 86 │           39.53 / 40.43 ±0.65 / 41.10 ms │           39.14 / 39.88 ±0.87 / 41.53 ms │     no change │
│ QQuery 87 │           85.59 / 88.37 ±2.67 / 92.37 ms │           84.93 / 88.16 ±3.49 / 94.07 ms │     no change │
│ QQuery 88 │        100.23 / 100.52 ±0.25 / 100.90 ms │         99.43 / 100.56 ±0.99 / 101.89 ms │     no change │
│ QQuery 89 │        119.08 / 119.78 ±0.64 / 120.67 ms │        118.30 / 121.06 ±1.58 / 123.05 ms │     no change │
│ QQuery 90 │           23.49 / 24.39 ±0.68 / 25.52 ms │           23.03 / 24.29 ±1.15 / 26.33 ms │     no change │
│ QQuery 91 │           62.29 / 65.33 ±2.23 / 68.68 ms │           61.31 / 64.37 ±1.79 / 66.50 ms │     no change │
│ QQuery 92 │           57.57 / 58.69 ±0.87 / 59.67 ms │           56.43 / 57.38 ±0.95 / 59.16 ms │     no change │
│ QQuery 93 │        187.38 / 188.32 ±1.08 / 190.04 ms │        185.67 / 187.00 ±0.84 / 188.26 ms │     no change │
│ QQuery 94 │           61.13 / 61.69 ±0.36 / 62.21 ms │           60.58 / 62.05 ±1.05 / 63.74 ms │     no change │
│ QQuery 95 │        127.48 / 129.14 ±1.38 / 130.85 ms │        127.61 / 129.06 ±1.36 / 131.23 ms │     no change │
│ QQuery 96 │           71.58 / 73.45 ±1.13 / 74.88 ms │           73.27 / 74.64 ±0.84 / 75.47 ms │     no change │
│ QQuery 97 │        125.42 / 126.75 ±1.17 / 128.79 ms │        125.31 / 127.08 ±1.05 / 128.39 ms │     no change │
│ QQuery 98 │        151.19 / 155.33 ±2.16 / 157.36 ms │        154.72 / 156.36 ±1.02 / 157.64 ms │     no change │
│ QQuery 99 │ 10751.95 / 10782.14 ±31.58 / 10836.85 ms │ 10797.36 / 10829.90 ±30.41 / 10867.36 ms │     no change │
└───────────┴──────────────────────────────────────────┴──────────────────────────────────────────┴───────────────┘
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┓
┃ Benchmark Summary                        ┃            ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━┩
│ Total Time (HEAD)                        │ 31533.47ms │
│ Total Time (prefetch-morsels-on-21351)   │ 31295.82ms │
│ Average Time (HEAD)                      │   318.52ms │
│ Average Time (prefetch-morsels-on-21351) │   316.12ms │
│ Queries Faster                           │          2 │
│ Queries Slower                           │          3 │
│ Queries with No Change                   │         94 │
│ Queries with Failure                     │          0 │
└──────────────────────────────────────────┴────────────┘

Resource Usage

tpcds — base (merge-base)

Metric Value
Wall time 158.0s
Peak memory 5.3 GiB
Avg memory 4.4 GiB
CPU user 260.1s
CPU sys 17.3s
Peak spill 0 B

tpcds — branch

Metric Value
Wall time 156.8s
Peak memory 5.6 GiB
Avg memory 4.6 GiB
CPU user 258.6s
CPU sys 17.0s
Peak spill 0 B

File an issue against this benchmark runner

@Dandandan Dandandan closed this Apr 16, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

datasource Changes to the datasource crate

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants