You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@datafusion.apache.org by "Rachelint (via GitHub)" <gi...@apache.org> on 2026/10/07 10:23:23 UTC
[PR] perf: reserve group-key storage across partial flushes [datafusion]
Rachelint opened a new pull request, #26112:
URL: https://github.com/apache/datafusion/pull/26112
## Which issue does this PR close?
None. Experimental follow-up to #26109; do not merge yet.
## Rationale for this change
Partial aggregation can flush a small working set repeatedly. #26109 retains the single StringView hash table and reserves its payload, but other group-key buffers still grow again after each flush. This experiment extends allocation retention and replacement-buffer reservation to other key types.
## What changes are included in this PR?
Based on #26109 at `522c7492f0ee2596a8d244c46bda805196f786e2` (and the byte/row threshold experiment in #26100):
- Reserve single StringView views as well as payload; retain hash scratch storage.
- Retain primitive and byte-key hash tables, and reserve replacement key buffers using the previous emitted lengths.
- Keep multi-column builders and scratch Vec capacities across threshold flushes. Reserve primitive, byte, view, fixed-size-binary, list and dictionary storage; retain row-backed storage where output is independent.
- Extend contiguous StringView payload storage to multi-column view keys.
- Preserve aggregate batch-group-index capacity on threshold flushes.
- Count retained capacities in the existing working-set threshold and memory reservation. Memory-pressure emission keeps releasing allocations.
Output batches own emitted buffers. Replacement buffers are reserved for the next cycle; live output buffers are never overwritten.
## What is the testing strategy for this PR?
Passed:
- `cargo fmt --all` and `git diff --check`.
- `cargo clippy --all-targets --all-features -- -D warnings`.
- 141 group-values tests, including three successive flush/refill cycles across nullable primitive, strings/views, mixed, list, dictionary, fixed-size-binary and row-backed keys while retaining earlier outputs.
- 39 binary-map/view-map tests.
- `aggregate_partial_flush.slt`, checking integers, strings, views, mixed keys, nulls, distinct and aggregate states.
- Generated config documentation and `doc_prettier_check.sh --write --allow-dirty`.
ECS full ClickBench comparison is running: all 43 queries; main (`7b1e55d4`), #26109 (`522c7492`) and this branch. Both candidates use 2 MiB / zero row threshold, 12 partitions, batch size 8192 and CPUs 0-11. Two reversed rounds, four iterations per cell, discarding the first iteration. Results will be added after completion.
## Are there any user-facing changes?
Only when the experimental Partial flush thresholds are enabled (both default to zero). More group-key allocation capacity survives threshold flushes. Final aggregation and its bucket behavior are unchanged.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] perf: reserve group-key storage across partial flushes [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #26112:
URL: https://github.com/apache/datafusion/pull/26112#issuecomment-6036958798
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/26112#issuecomment-6036917922)
**Instance:** `c4a-highcpu-48` (12 vCPU / 65 GiB) | `Linux bench-c6036917922-3119-lsx6s 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux`
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 48
On-line CPU(s) list: 0-47
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 48
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 3 MiB (48 instances)
L1i cache: 3 MiB (48 instances)
L2 cache: 96 MiB (48 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-47
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
Comparing codex/partial-reserve-all (5008df083795701eac24112064ac6711bb75791a) to 522c7492f0ee2596a8d244c46bda805196f786e2 [diff](https://github.com/apache/datafusion/compare/522c7492f0ee2596a8d244c46bda805196f786e2..5008df083795701eac24112064ac6711bb75791a)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_BYTES: "2097152"
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_ROWS: "0"
baseline:
ref: "522c7492f0ee2596a8d244c46bda805196f786e2"
```
</details>
Results will be posted here when complete
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] perf: reserve group-key storage across partial flushes [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #26112:
URL: https://github.com/apache/datafusion/pull/26112#issuecomment-6036678632
π€ Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/26112#issuecomment-6036558620)
**Instance:** `c4a-highcpu-48` (12 vCPU / 65 GiB) | `Linux bench-c6036558620-3118-k8tr7 6.12.94+ #1 SMP Fri Aug 21 08:00:16 UTC 2026 aarch64 GNU/Linux`
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 48
On-line CPU(s) list: 0-47
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 48
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 3 MiB (48 instances)
L1i cache: 3 MiB (48 instances)
L2 cache: 96 MiB (48 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-47
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
Comparing codex/partial-reserve-all (5008df083795701eac24112064ac6711bb75791a) to 522c7492f0ee2596a8d244c46bda805196f786e2 [diff](https://github.com/apache/datafusion/compare/522c7492f0ee2596a8d244c46bda805196f786e2..5008df083795701eac24112064ac6711bb75791a)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_BYTES: "2097152"
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_ROWS: "0"
baseline:
ref: "522c7492f0ee2596a8d244c46bda805196f786e2"
```
</details>
Results will be posted here when complete
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] perf: reserve group-key storage across partial flushes [datafusion]
Posted by "Rachelint (via GitHub)" <gi...@apache.org>.
Rachelint commented on PR #26112:
URL: https://github.com/apache/datafusion/pull/26112#issuecomment-6036558620
run benchmark clickbench_partitioned
baseline:
ref: 522c7492f0ee2596a8d244c46bda805196f786e2
env:
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_BYTES: "2097152"
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_ROWS: "0"
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] perf: reserve group-key storage across partial flushes [datafusion]
Posted by "github-actions[bot] (via GitHub)" <gi...@apache.org>.
github-actions[bot] commented on PR #26112:
URL: https://github.com/apache/datafusion/pull/26112#issuecomment-6036045744
<!-- semver-check-comment -->
Thank you for opening this pull request!
Reviewer note: [cargo-semver-checks](https://github.com/obi1kenobi/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>
<summary>Details</summary>
```
Cloning apache/main
Building datafusion-common v55.1.0 (current)
Built [ 29.370s] (current)
Parsing datafusion-common v55.1.0 (current)
Parsed [ 0.054s] (current)
Building datafusion-common v55.1.0 (baseline)
Built [ 28.359s] (baseline)
Parsing datafusion-common v55.1.0 (baseline)
Parsed [ 0.054s] (baseline)
Checking datafusion-common v55.1.0 -> v55.1.0 (no change; assume patch)
Checked [ 0.866s] 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 ExecutionOptions.partial_aggregation_flush_bytes in /home/runner/work/datafusion/datafusion/datafusion/common/src/config.rs:897
field ExecutionOptions.partial_aggregation_flush_rows in /home/runner/work/datafusion/datafusion/datafusion/common/src/config.rs:897
Summary semver requires new major version: 1 major and 0 minor checks failed
Finished [ 61.022s] datafusion-common
Building datafusion-physical-expr-common v55.1.0 (current)
Built [ 20.137s] (current)
Parsing datafusion-physical-expr-common v55.1.0 (current)
Parsed [ 0.018s] (current)
Building datafusion-physical-expr-common v55.1.0 (baseline)
Built [ 19.935s] (baseline)
Parsing datafusion-physical-expr-common v55.1.0 (baseline)
Parsed [ 0.020s] (baseline)
Checking datafusion-physical-expr-common v55.1.0 -> v55.1.0 (no change; assume patch)
Checked [ 0.242s] 223 checks: 223 pass, 31 skip
Summary no semver update required
Finished [ 41.117s] datafusion-physical-expr-common
Building datafusion-physical-plan v55.1.0 (current)
Built [ 31.101s] (current)
Parsing datafusion-physical-plan v55.1.0 (current)
Parsed [ 0.148s] (current)
Building datafusion-physical-plan v55.1.0 (baseline)
Built [ 31.543s] (baseline)
Parsing datafusion-physical-plan v55.1.0 (baseline)
Parsed [ 0.153s] (baseline)
Checking datafusion-physical-plan v55.1.0 -> v55.1.0 (no change; assume patch)
Checked [ 0.769s] 223 checks: 223 pass, 31 skip
Summary no semver update required
Finished [ 65.228s] datafusion-physical-plan
Building datafusion-sqllogictest v55.1.0 (current)
Built [ 81.186s] (current)
Parsing datafusion-sqllogictest v55.1.0 (current)
Parsed [ 0.014s] (current)
Building datafusion-sqllogictest v55.1.0 (baseline)
Built [ 79.804s] (baseline)
Parsing datafusion-sqllogictest v55.1.0 (baseline)
Parsed [ 0.014s] (baseline)
Checking datafusion-sqllogictest v55.1.0 -> v55.1.0 (no change; assume patch)
Checked [ 0.112s] 223 checks: 223 pass, 31 skip
Summary no semver update required
Finished [ 163.494s] datafusion-sqllogictest
```
</details>
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] perf: reserve group-key storage across partial flushes [datafusion]
Posted by "Rachelint (via GitHub)" <gi...@apache.org>.
Rachelint commented on PR #26112:
URL: https://github.com/apache/datafusion/pull/26112#issuecomment-6036917922
run benchmark clickbench_partitioned
baseline:
ref: 522c7492f0ee2596a8d244c46bda805196f786e2
env:
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_BYTES: "2097152"
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_ROWS: "0"
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] perf: reserve group-key storage across partial flushes [datafusion]
Posted by "Rachelint (via GitHub)" <gi...@apache.org>.
Rachelint closed pull request #26112: perf: reserve group-key storage across partial flushes
URL: https://github.com/apache/datafusion/pull/26112
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] perf: reserve group-key storage across partial flushes [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #26112:
URL: https://github.com/apache/datafusion/pull/26112#issuecomment-6036880350
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/26112#issuecomment-6036558620)
**Instance:** `c4a-highcpu-48` (12 vCPU / 65 GiB)
Comparing codex/partial-reserve-all (5008df083795701eac24112064ac6711bb75791a) to 522c7492f0ee2596a8d244c46bda805196f786e2 [diff](https://github.com/apache/datafusion/compare/522c7492f0ee2596a8d244c46bda805196f786e2..5008df083795701eac24112064ac6711bb75791a)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_BYTES: "2097152"
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_ROWS: "0"
baseline:
ref: "522c7492f0ee2596a8d244c46bda805196f786e2"
```
</details>
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 48
On-line CPU(s) list: 0-47
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 48
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 3 MiB (48 instances)
L1i cache: 3 MiB (48 instances)
L2 cache: 96 MiB (48 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-47
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and codex_partial-reserve-all
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β codex_partial-reserve-all β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.20 ms β 1.20 ms β no change β
β QQuery 1 β 11.62 ms β 11.84 ms β no change β
β QQuery 2 β 36.44 ms β 36.61 ms β no change β
β QQuery 3 β 30.88 ms β 30.93 ms β no change β
β QQuery 4 β 221.23 ms β 185.04 ms β +1.20x faster β
β QQuery 5 β 262.82 ms β 253.19 ms β no change β
β QQuery 6 β 1.23 ms β 1.25 ms β no change β
β QQuery 7 β 12.94 ms β 12.81 ms β no change β
β QQuery 8 β 320.45 ms β 297.48 ms β +1.08x faster β
β QQuery 9 β 469.35 ms β 476.42 ms β no change β
β QQuery 10 β 65.31 ms β 63.91 ms β no change β
β QQuery 11 β 75.35 ms β 74.88 ms β no change β
β QQuery 12 β 242.30 ms β 246.80 ms β no change β
β QQuery 13 β 358.47 ms β 350.00 ms β no change β
β QQuery 14 β 273.32 ms β 256.70 ms β +1.06x faster β
β QQuery 15 β 253.17 ms β 228.55 ms β +1.11x faster β
β QQuery 16 β 576.40 ms β 525.36 ms β +1.10x faster β
β QQuery 17 β 568.47 ms β 529.93 ms β +1.07x faster β
β QQuery 18 β 1076.59 ms β 1012.08 ms β +1.06x faster β
β QQuery 19 β 27.10 ms β 26.97 ms β no change β
β QQuery 20 β 519.56 ms β 514.33 ms β no change β
β QQuery 21 β 508.93 ms β 506.30 ms β no change β
β QQuery 22 β 1056.23 ms β 991.69 ms β +1.07x faster β
β QQuery 23 β 3178.31 ms β 3073.35 ms β no change β
β QQuery 24 β 39.95 ms β 40.31 ms β no change β
β QQuery 25 β 106.12 ms β 106.56 ms β no change β
β QQuery 26 β 40.79 ms β 40.54 ms β no change β
β QQuery 27 β 518.77 ms β 514.20 ms β no change β
β QQuery 28 β 2913.72 ms β 2906.21 ms β no change β
β QQuery 29 β 38.80 ms β 39.08 ms β no change β
β QQuery 30 β 288.75 ms β 274.36 ms β no change β
β QQuery 31 β 285.52 ms β 274.03 ms β no change β
β QQuery 32 β 1069.78 ms β 1114.09 ms β no change β
β QQuery 33 β 1305.69 ms β 1341.16 ms β no change β
β QQuery 34 β 1339.09 ms β 1343.96 ms β no change β
β QQuery 35 β 250.98 ms β 218.83 ms β +1.15x faster β
β QQuery 36 β 59.32 ms β 60.77 ms β no change β
β QQuery 37 β 34.53 ms β 35.15 ms β no change β
β QQuery 38 β 43.37 ms β 39.86 ms β +1.09x faster β
β QQuery 39 β 139.79 ms β 132.85 ms β no change β
β QQuery 40 β 14.03 ms β 14.05 ms β no change β
β QQuery 41 β 13.68 ms β 13.75 ms β no change β
β QQuery 42 β 13.21 ms β 13.16 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 18663.58ms β
β Total Time (codex_partial-reserve-all) β 18230.53ms β
β Average Time (HEAD) β 434.04ms β
β Average Time (codex_partial-reserve-all) β 423.97ms β
β Queries Faster β 10 β
β Queries Slower β 0 β
β Queries with No Change β 33 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and codex_partial-reserve-all
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³βββββββββββββββββββββββββββββββββββββββββ³βββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β codex_partial-reserve-all β Change β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.20 / 4.03 Β±5.51 / 15.05 ms β 1.20 / 4.06 Β±5.56 / 15.18 ms β no change β
β QQuery 1 β 11.62 / 11.82 Β±0.11 / 11.92 ms β 11.84 / 11.96 Β±0.11 / 12.17 ms β no change β
β QQuery 2 β 36.44 / 36.60 Β±0.13 / 36.78 ms β 36.61 / 36.85 Β±0.17 / 37.13 ms β no change β
β QQuery 3 β 30.88 / 31.31 Β±0.59 / 32.42 ms β 30.93 / 31.17 Β±0.21 / 31.56 ms β no change β
β QQuery 4 β 221.23 / 224.24 Β±1.64 / 225.93 ms β 185.04 / 186.77 Β±1.18 / 188.33 ms β +1.20x faster β
β QQuery 5 β 262.82 / 265.31 Β±1.89 / 268.37 ms β 253.19 / 261.70 Β±6.51 / 271.20 ms β no change β
β QQuery 6 β 1.23 / 1.37 Β±0.22 / 1.79 ms β 1.25 / 1.38 Β±0.21 / 1.81 ms β no change β
β QQuery 7 β 12.94 / 13.02 Β±0.06 / 13.09 ms β 12.81 / 12.93 Β±0.06 / 13.01 ms β no change β
β QQuery 8 β 320.45 / 325.04 Β±2.80 / 328.14 ms β 297.48 / 301.09 Β±2.97 / 306.17 ms β +1.08x faster β
β QQuery 9 β 469.35 / 479.01 Β±5.29 / 483.57 ms β 476.42 / 486.32 Β±7.53 / 498.27 ms β no change β
β QQuery 10 β 65.31 / 66.14 Β±0.66 / 67.04 ms β 63.91 / 65.59 Β±1.74 / 68.94 ms β no change β
β QQuery 11 β 75.35 / 76.97 Β±1.49 / 79.69 ms β 74.88 / 76.55 Β±2.29 / 81.05 ms β no change β
β QQuery 12 β 242.30 / 249.16 Β±3.75 / 253.06 ms β 246.80 / 253.94 Β±5.25 / 262.46 ms β no change β
β QQuery 13 β 358.47 / 361.92 Β±3.00 / 366.96 ms β 350.00 / 357.19 Β±5.06 / 363.67 ms β no change β
β QQuery 14 β 273.32 / 278.15 Β±4.18 / 285.82 ms β 256.70 / 266.48 Β±11.96 / 290.02 ms β no change β
β QQuery 15 β 253.17 / 258.68 Β±4.62 / 266.11 ms β 228.55 / 236.89 Β±7.58 / 250.23 ms β +1.09x faster β
β QQuery 16 β 576.40 / 582.08 Β±4.27 / 589.10 ms β 525.36 / 539.02 Β±8.66 / 549.29 ms β +1.08x faster β
β QQuery 17 β 568.47 / 582.32 Β±11.44 / 602.82 ms β 529.93 / 540.58 Β±9.40 / 558.11 ms β +1.08x faster β
β QQuery 18 β 1076.59 / 1085.53 Β±8.86 / 1101.16 ms β 1012.08 / 1037.33 Β±27.63 / 1087.30 ms β no change β
β QQuery 19 β 27.10 / 27.38 Β±0.22 / 27.71 ms β 26.97 / 27.23 Β±0.22 / 27.62 ms β no change β
β QQuery 20 β 519.56 / 525.68 Β±6.27 / 536.88 ms β 514.33 / 525.09 Β±9.37 / 541.61 ms β no change β
β QQuery 21 β 508.93 / 519.94 Β±5.84 / 525.93 ms β 506.30 / 517.17 Β±6.66 / 525.16 ms β no change β
β QQuery 22 β 1056.23 / 1105.82 Β±28.28 / 1132.16 ms β 991.69 / 1003.38 Β±10.47 / 1021.42 ms β +1.10x faster β
β QQuery 23 β 3178.31 / 3296.81 Β±103.54 / 3475.53 ms β 3073.35 / 3258.64 Β±129.14 / 3396.73 ms β no change β
β QQuery 24 β 39.95 / 151.03 Β±93.15 / 251.89 ms β 40.31 / 52.48 Β±22.59 / 97.65 ms β +2.88x faster β
β QQuery 25 β 106.12 / 107.71 Β±1.58 / 110.60 ms β 106.56 / 195.79 Β±175.22 / 546.22 ms β 1.82x slower β
β QQuery 26 β 40.79 / 41.76 Β±1.26 / 44.25 ms β 40.54 / 41.68 Β±1.39 / 44.37 ms β no change β
β QQuery 27 β 518.77 / 524.46 Β±5.65 / 533.91 ms β 514.20 / 524.58 Β±10.90 / 545.55 ms β no change β
β QQuery 28 β 2913.72 / 2926.61 Β±10.87 / 2945.34 ms β 2906.21 / 2925.58 Β±22.97 / 2970.57 ms β no change β
β QQuery 29 β 38.80 / 46.80 Β±10.15 / 65.53 ms β 39.08 / 39.43 Β±0.35 / 40.08 ms β +1.19x faster β
β QQuery 30 β 288.75 / 296.93 Β±8.69 / 312.44 ms β 274.36 / 287.90 Β±10.36 / 305.45 ms β no change β
β QQuery 31 β 285.52 / 293.05 Β±5.01 / 300.55 ms β 274.03 / 290.58 Β±13.30 / 311.94 ms β no change β
β QQuery 32 β 1069.78 / 1178.55 Β±65.35 / 1258.81 ms β 1114.09 / 1261.26 Β±148.73 / 1470.80 ms β 1.07x slower β
β QQuery 33 β 1305.69 / 1356.63 Β±56.93 / 1454.99 ms β 1341.16 / 1369.55 Β±24.02 / 1406.15 ms β no change β
β QQuery 34 β 1339.09 / 1361.65 Β±23.40 / 1400.11 ms β 1343.96 / 1408.63 Β±50.44 / 1492.69 ms β no change β
β QQuery 35 β 250.98 / 277.95 Β±21.18 / 298.32 ms β 218.83 / 252.01 Β±29.35 / 291.83 ms β +1.10x faster β
β QQuery 36 β 59.32 / 65.52 Β±3.61 / 70.51 ms β 60.77 / 62.46 Β±1.49 / 64.96 ms β no change β
β QQuery 37 β 34.53 / 38.04 Β±3.25 / 43.68 ms β 35.15 / 36.91 Β±2.76 / 42.40 ms β no change β
β QQuery 38 β 43.37 / 46.18 Β±4.50 / 55.14 ms β 39.86 / 41.26 Β±1.21 / 42.88 ms β +1.12x faster β
β QQuery 39 β 139.79 / 144.10 Β±4.15 / 151.00 ms β 132.85 / 141.27 Β±5.50 / 146.28 ms β no change β
β QQuery 40 β 14.03 / 14.34 Β±0.18 / 14.58 ms β 14.05 / 14.66 Β±0.59 / 15.73 ms β no change β
β QQuery 41 β 13.68 / 14.06 Β±0.66 / 15.37 ms β 13.75 / 14.10 Β±0.41 / 14.91 ms β no change β
β QQuery 42 β 13.21 / 13.47 Β±0.13 / 13.57 ms β 13.16 / 13.35 Β±0.12 / 13.53 ms β no change β
βββββββββββββ΄βββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 19307.17ms β
β Total Time (codex_partial-reserve-all) β 19012.77ms β
β Average Time (HEAD) β 449.00ms β
β Average Time (codex_partial-reserve-all) β 442.16ms β
β Queries Faster β 10 β
β Queries Slower β 2 β
β Queries with No Change β 31 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 100.0s |
| Peak memory | 18.1 GiB |
| Avg memory | 5.2 GiB |
| CPU user | 951.6s |
| CPU sys | 85.3s |
| Peak spill | 0 B |
**clickbench_partitioned β branch**
| Metric | Value |
|--------|-------|
| Wall time | 100.0s |
| Peak memory | 18.5 GiB |
| Avg memory | 5.5 GiB |
| CPU user | 934.7s |
| CPU sys | 86.8s |
| Peak spill | 0 B |
</details>
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org
Re: [PR] perf: reserve group-key storage across partial flushes [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #26112:
URL: https://github.com/apache/datafusion/pull/26112#issuecomment-6037088468
π€ Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/26112#issuecomment-6036917922)
**Instance:** `c4a-highcpu-48` (12 vCPU / 65 GiB)
Comparing codex/partial-reserve-all (5008df083795701eac24112064ac6711bb75791a) to 522c7492f0ee2596a8d244c46bda805196f786e2 [diff](https://github.com/apache/datafusion/compare/522c7492f0ee2596a8d244c46bda805196f786e2..5008df083795701eac24112064ac6711bb75791a)
<details><summary>Run configuration</summary>
```yaml
run benchmark clickbench_partitioned
env:
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_BYTES: "2097152"
DATAFUSION_EXECUTION_PARTIAL_AGGREGATION_FLUSH_ROWS: "0"
baseline:
ref: "522c7492f0ee2596a8d244c46bda805196f786e2"
```
</details>
<details><summary>CPU Details (lscpu)</summary>
```
Architecture: aarch64
CPU op-mode(s): 64-bit
Byte Order: Little Endian
CPU(s): 48
On-line CPU(s) list: 0-47
Vendor ID: ARM
Model name: Neoverse-V2
Model: 1
Thread(s) per core: 1
Core(s) per cluster: 48
Socket(s): -
Cluster(s): 1
Stepping: r0p1
BogoMIPS: 2000.00
Flags: fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm svesha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bf16 dgh rng bti
L1d cache: 3 MiB (48 instances)
L1i cache: 3 MiB (48 instances)
L2 cache: 96 MiB (48 instances)
L3 cache: 80 MiB (1 instance)
NUMA node(s): 1
NUMA node0 CPU(s): 0-47
Vulnerability Gather data sampling: Not affected
Vulnerability Indirect target selection: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Mitigation; CSV2, BHB
Vulnerability Srbds: Not affected
Vulnerability Tsa: Not affected
Vulnerability Tsx async abort: Not affected
Vulnerability Vmscape: Not affected
```
</details>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and codex_partial-reserve-all
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³βββββββββββββ³ββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β codex_partial-reserve-all β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.22 ms β 1.20 ms β no change β
β QQuery 1 β 11.42 ms β 11.67 ms β no change β
β QQuery 2 β 36.54 ms β 36.57 ms β no change β
β QQuery 3 β 30.67 ms β 30.69 ms β no change β
β QQuery 4 β 219.97 ms β 184.57 ms β +1.19x faster β
β QQuery 5 β 252.25 ms β 254.80 ms β no change β
β QQuery 6 β 1.24 ms β 1.24 ms β no change β
β QQuery 7 β 12.38 ms β 12.78 ms β no change β
β QQuery 8 β 315.06 ms β 295.21 ms β +1.07x faster β
β QQuery 9 β 458.67 ms β 471.76 ms β no change β
β QQuery 10 β 64.55 ms β 63.38 ms β no change β
β QQuery 11 β 74.56 ms β 73.96 ms β no change β
β QQuery 12 β 235.26 ms β 238.39 ms β no change β
β QQuery 13 β 348.24 ms β 356.99 ms β no change β
β QQuery 14 β 267.82 ms β 261.26 ms β no change β
β QQuery 15 β 250.66 ms β 221.56 ms β +1.13x faster β
β QQuery 16 β 564.48 ms β 536.71 ms β no change β
β QQuery 17 β 566.72 ms β 526.01 ms β +1.08x faster β
β QQuery 18 β 1086.85 ms β 1005.27 ms β +1.08x faster β
β QQuery 19 β 26.67 ms β 26.81 ms β no change β
β QQuery 20 β 516.52 ms β 514.45 ms β no change β
β QQuery 21 β 515.30 ms β 508.24 ms β no change β
β QQuery 22 β 998.82 ms β 987.67 ms β no change β
β QQuery 23 β 3147.02 ms β 3064.37 ms β no change β
β QQuery 24 β 40.00 ms β 40.06 ms β no change β
β QQuery 25 β 104.09 ms β 105.05 ms β no change β
β QQuery 26 β 40.45 ms β 40.71 ms β no change β
β QQuery 27 β 507.51 ms β 509.91 ms β no change β
β QQuery 28 β 2884.71 ms β 2859.11 ms β no change β
β QQuery 29 β 38.79 ms β 38.78 ms β no change β
β QQuery 30 β 287.63 ms β 280.36 ms β no change β
β QQuery 31 β 278.69 ms β 276.30 ms β no change β
β QQuery 32 β 1061.31 ms β 1100.30 ms β no change β
β QQuery 33 β 1347.39 ms β 1307.35 ms β no change β
β QQuery 34 β 1326.42 ms β 1324.34 ms β no change β
β QQuery 35 β 251.69 ms β 217.07 ms β +1.16x faster β
β QQuery 36 β 59.00 ms β 60.50 ms β no change β
β QQuery 37 β 34.35 ms β 34.25 ms β no change β
β QQuery 38 β 40.03 ms β 42.12 ms β 1.05x slower β
β QQuery 39 β 129.05 ms β 130.78 ms β no change β
β QQuery 40 β 13.55 ms β 13.58 ms β no change β
β QQuery 41 β 13.37 ms β 13.02 ms β no change β
β QQuery 42 β 12.65 ms β 12.75 ms β no change β
βββββββββββββ΄βββββββββββββ΄ββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 18473.56ms β
β Total Time (codex_partial-reserve-all) β 18091.91ms β
β Average Time (HEAD) β 429.62ms β
β Average Time (codex_partial-reserve-all) β 420.74ms β
β Queries Faster β 6 β
β Queries Slower β 1 β
β Queries with No Change β 36 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
Distribution per query (min / mean Β±stddev / max):
Comparing HEAD and codex_partial-reserve-all
--------------------
Benchmark clickbench_partitioned.json
--------------------
βββββββββββββ³ββββββββββββββββββββββββββββββββββββββββ³βββββββββββββββββββββββββββββββββββββββββ³ββββββββββββββββ
β Query β HEAD β codex_partial-reserve-all β Change β
β‘βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β QQuery 0 β 1.22 / 3.95 Β±5.39 / 14.74 ms β 1.20 / 4.01 Β±5.49 / 14.99 ms β no change β
β QQuery 1 β 11.42 / 11.67 Β±0.19 / 12.02 ms β 11.67 / 11.82 Β±0.15 / 12.03 ms β no change β
β QQuery 2 β 36.54 / 36.88 Β±0.28 / 37.23 ms β 36.57 / 36.97 Β±0.25 / 37.27 ms β no change β
β QQuery 3 β 30.67 / 31.11 Β±0.63 / 32.36 ms β 30.69 / 30.76 Β±0.05 / 30.84 ms β no change β
β QQuery 4 β 219.97 / 221.09 Β±1.31 / 223.53 ms β 184.57 / 189.42 Β±3.97 / 195.69 ms β +1.17x faster β
β QQuery 5 β 252.25 / 257.56 Β±3.10 / 261.65 ms β 254.80 / 257.73 Β±2.39 / 260.50 ms β no change β
β QQuery 6 β 1.24 / 1.38 Β±0.21 / 1.81 ms β 1.24 / 1.37 Β±0.20 / 1.76 ms β no change β
β QQuery 7 β 12.38 / 12.56 Β±0.11 / 12.71 ms β 12.78 / 12.84 Β±0.04 / 12.91 ms β no change β
β QQuery 8 β 315.06 / 316.95 Β±2.72 / 322.35 ms β 295.21 / 298.21 Β±2.53 / 302.46 ms β +1.06x faster β
β QQuery 9 β 458.67 / 463.01 Β±3.12 / 466.83 ms β 471.76 / 481.79 Β±11.17 / 503.37 ms β no change β
β QQuery 10 β 64.55 / 66.07 Β±1.64 / 68.61 ms β 63.38 / 64.13 Β±0.67 / 65.13 ms β no change β
β QQuery 11 β 74.56 / 75.94 Β±1.09 / 77.77 ms β 73.96 / 79.53 Β±4.63 / 87.30 ms β no change β
β QQuery 12 β 235.26 / 239.67 Β±2.95 / 243.70 ms β 238.39 / 245.64 Β±5.40 / 251.90 ms β no change β
β QQuery 13 β 348.24 / 352.41 Β±3.03 / 357.19 ms β 356.99 / 370.29 Β±18.83 / 407.37 ms β 1.05x slower β
β QQuery 14 β 267.82 / 279.62 Β±6.48 / 285.91 ms β 261.26 / 267.23 Β±5.48 / 277.03 ms β no change β
β QQuery 15 β 250.66 / 253.71 Β±3.21 / 258.23 ms β 221.56 / 231.32 Β±6.27 / 241.14 ms β +1.10x faster β
β QQuery 16 β 564.48 / 576.99 Β±13.30 / 602.68 ms β 536.71 / 549.22 Β±15.62 / 574.79 ms β no change β
β QQuery 17 β 566.72 / 579.78 Β±9.62 / 592.36 ms β 526.01 / 535.68 Β±5.83 / 542.73 ms β +1.08x faster β
β QQuery 18 β 1086.85 / 1097.96 Β±11.61 / 1119.06 ms β 1005.27 / 1029.53 Β±29.49 / 1086.38 ms β +1.07x faster β
β QQuery 19 β 26.67 / 26.89 Β±0.15 / 27.08 ms β 26.81 / 26.90 Β±0.06 / 26.96 ms β no change β
β QQuery 20 β 516.52 / 523.45 Β±7.67 / 535.02 ms β 514.45 / 525.28 Β±8.56 / 538.23 ms β no change β
β QQuery 21 β 515.30 / 525.37 Β±9.32 / 536.86 ms β 508.24 / 513.87 Β±3.64 / 518.81 ms β no change β
β QQuery 22 β 998.82 / 1019.64 Β±10.55 / 1027.61 ms β 987.67 / 998.60 Β±11.07 / 1018.65 ms β no change β
β QQuery 23 β 3147.02 / 3224.59 Β±85.20 / 3389.99 ms β 3064.37 / 3268.40 Β±126.10 / 3425.98 ms β no change β
β QQuery 24 β 40.00 / 116.98 Β±124.38 / 361.73 ms β 40.06 / 61.27 Β±41.27 / 143.81 ms β +1.91x faster β
β QQuery 25 β 104.09 / 152.51 Β±92.22 / 336.85 ms β 105.05 / 187.23 Β±158.95 / 505.11 ms β 1.23x slower β
β QQuery 26 β 40.45 / 41.79 Β±1.94 / 45.58 ms β 40.71 / 41.42 Β±0.89 / 43.01 ms β no change β
β QQuery 27 β 507.51 / 516.23 Β±6.30 / 526.58 ms β 509.91 / 519.01 Β±6.86 / 528.38 ms β no change β
β QQuery 28 β 2884.71 / 2897.17 Β±10.84 / 2911.49 ms β 2859.11 / 2893.76 Β±20.70 / 2922.16 ms β no change β
β QQuery 29 β 38.79 / 43.22 Β±7.94 / 59.07 ms β 38.78 / 39.22 Β±0.41 / 39.95 ms β +1.10x faster β
β QQuery 30 β 287.63 / 290.23 Β±2.47 / 293.38 ms β 280.36 / 288.71 Β±11.42 / 311.31 ms β no change β
β QQuery 31 β 278.69 / 301.65 Β±22.39 / 343.53 ms β 276.30 / 287.16 Β±6.88 / 295.39 ms β no change β
β QQuery 32 β 1061.31 / 1103.43 Β±25.92 / 1136.65 ms β 1100.30 / 1204.96 Β±97.04 / 1371.76 ms β 1.09x slower β
β QQuery 33 β 1347.39 / 1386.98 Β±30.72 / 1425.67 ms β 1307.35 / 1376.13 Β±49.49 / 1459.33 ms β no change β
β QQuery 34 β 1326.42 / 1352.94 Β±22.61 / 1383.15 ms β 1324.34 / 1356.60 Β±29.11 / 1392.13 ms β no change β
β QQuery 35 β 251.69 / 278.94 Β±44.86 / 368.23 ms β 217.07 / 234.43 Β±23.39 / 280.45 ms β +1.19x faster β
β QQuery 36 β 59.00 / 63.19 Β±5.22 / 73.44 ms β 60.50 / 65.11 Β±5.14 / 74.66 ms β no change β
β QQuery 37 β 34.35 / 39.09 Β±5.54 / 49.58 ms β 34.25 / 37.09 Β±3.14 / 42.72 ms β +1.05x faster β
β QQuery 38 β 40.03 / 44.55 Β±4.29 / 52.24 ms β 42.12 / 44.01 Β±1.57 / 45.89 ms β no change β
β QQuery 39 β 129.05 / 140.41 Β±8.08 / 153.73 ms β 130.78 / 134.99 Β±4.94 / 144.11 ms β no change β
β QQuery 40 β 13.55 / 14.63 Β±1.95 / 18.52 ms β 13.58 / 13.87 Β±0.27 / 14.24 ms β +1.06x faster β
β QQuery 41 β 13.37 / 15.78 Β±2.47 / 19.24 ms β 13.02 / 15.10 Β±3.96 / 23.02 ms β no change β
β QQuery 42 β 12.65 / 12.89 Β±0.17 / 13.10 ms β 12.75 / 12.85 Β±0.09 / 12.96 ms β no change β
βββββββββββββ΄ββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββββββββββββββββββββββββββββββ΄ββββββββββββββββ
ββββββββββββββββββββββββββββββββββββββββββββ³βββββββββββββ
β Benchmark Summary β β
β‘ββββββββββββββββββββββββββββββββββββββββββββββββββββββββ©
β Total Time (HEAD) β 19010.90ms β
β Total Time (codex_partial-reserve-all) β 18843.46ms β
β Average Time (HEAD) β 442.11ms β
β Average Time (codex_partial-reserve-all) β 438.22ms β
β Queries Faster β 10 β
β Queries Slower β 3 β
β Queries with No Change β 30 β
β Queries with Failure β 0 β
ββββββββββββββββββββββββββββββββββββββββββββ΄βββββββββββββ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned β base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 100.0s |
| Peak memory | 18.3 GiB |
| Avg memory | 5.3 GiB |
| CPU user | 948.8s |
| CPU sys | 81.4s |
| Peak spill | 0 B |
**clickbench_partitioned β branch**
| Metric | Value |
|--------|-------|
| Wall time | 95.0s |
| Peak memory | 18.1 GiB |
| Avg memory | 5.6 GiB |
| CPU user | 932.4s |
| CPU sys | 82.7s |
| Peak spill | 0 B |
</details>
---
[File an issue](https://github.com/adriangb/datafusion-benchmarking/issues) against this benchmark runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscribe@datafusion.apache.org
For additional commands, e-mail: github-help@datafusion.apache.org