You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@datafusion.apache.org by "Dandandan (via GitHub)" <gi...@apache.org> on 2026/04/02 19:23:32 UTC
[PR] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Dandandan opened a new pull request, #21326:
URL: https://github.com/apache/datafusion/pull/21326
## Which issue does this PR close?
N/A - performance improvement
## Rationale for this change
`CoalescePartitionsExec::execute()` eagerly spawns tasks for all input partitions immediately, even before any batch is polled from the output stream. This wastes resources when streams are created but never consumed (e.g., query cancelled before first poll), and creates an unnecessary burst of concurrent tasks when many `CoalescePartitionsExec` nodes exist in a plan.
## What changes are included in this PR?
Introduces a `CoalescePartitionsStream` that defers spawning input partition tasks until the first `poll_next()` call, rather than in `execute()`. The single-partition fast path is unchanged.
## How are these changes tested?
All existing tests pass (8/8), covering:
- Multi-partition merge
- Single/multi partition with fetch limits
- Drop cancellation
- Panic propagation
๐ค Generated with [Claude Code](https://claude.com/claude-code)
--
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180077148
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21326#issuecomment-4180001312)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<details><summary>CPU Details (lscpu)</summary>
```
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>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and lazy-coalesce-partitions
--------------------
Benchmark tpch_sf1.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโ
โ Query โ HEAD โ lazy-coalesce-partitions โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 45.29 / 46.03 ยฑ0.76 / 47.32 ms โ 45.36 / 46.22 ยฑ0.93 / 47.49 ms โ no change โ
โ QQuery 2 โ 20.84 / 21.39 ยฑ0.50 / 22.20 ms โ 24.70 / 24.89 ยฑ0.15 / 25.15 ms โ 1.16x slower โ
โ QQuery 3 โ 31.80 / 32.02 ยฑ0.21 / 32.41 ms โ 32.18 / 33.34 ยฑ1.29 / 35.21 ms โ no change โ
โ QQuery 4 โ 20.56 / 21.34 ยฑ0.82 / 22.63 ms โ 20.19 / 21.33 ยฑ0.79 / 22.36 ms โ no change โ
โ QQuery 5 โ 47.96 / 50.02 ยฑ1.61 / 51.73 ms โ 49.80 / 51.08 ยฑ0.98 / 52.22 ms โ no change โ
โ QQuery 6 โ 17.11 / 17.35 ยฑ0.18 / 17.56 ms โ 17.16 / 17.76 ยฑ0.94 / 19.62 ms โ no change โ
โ QQuery 7 โ 56.28 / 57.26 ยฑ0.87 / 58.35 ms โ 55.37 / 56.42 ยฑ1.53 / 59.43 ms โ no change โ
โ QQuery 8 โ 47.64 / 48.54 ยฑ0.73 / 49.82 ms โ 52.01 / 52.26 ยฑ0.23 / 52.63 ms โ 1.08x slower โ
โ QQuery 9 โ 54.41 / 55.36 ยฑ1.00 / 57.21 ms โ 65.51 / 67.19 ยฑ1.21 / 69.03 ms โ 1.21x slower โ
โ QQuery 10 โ 70.54 / 72.48 ยฑ1.42 / 74.78 ms โ 71.71 / 74.30 ยฑ2.11 / 77.56 ms โ no change โ
โ QQuery 11 โ 13.95 / 14.25 ยฑ0.31 / 14.80 ms โ 14.69 / 14.88 ยฑ0.23 / 15.32 ms โ no change โ
โ QQuery 12 โ 28.02 / 28.36 ยฑ0.37 / 29.04 ms โ 27.50 / 27.88 ยฑ0.21 / 28.09 ms โ no change โ
โ QQuery 13 โ 39.00 / 40.46 ยฑ1.71 / 43.78 ms โ 38.68 / 40.01 ยฑ1.03 / 41.62 ms โ no change โ
โ QQuery 14 โ 28.41 / 28.78 ยฑ0.44 / 29.63 ms โ 28.43 / 28.60 ยฑ0.12 / 28.76 ms โ no change โ
โ QQuery 15 โ 33.56 / 34.26 ยฑ0.61 / 35.31 ms โ 33.63 / 35.27 ยฑ0.90 / 36.25 ms โ no change โ
โ QQuery 16 โ 15.93 / 16.25 ยฑ0.19 / 16.46 ms โ 17.93 / 18.36 ยฑ0.36 / 18.97 ms โ 1.13x slower โ
โ QQuery 17 โ 72.74 / 74.09 ยฑ1.02 / 75.19 ms โ 71.94 / 73.83 ยฑ1.49 / 76.43 ms โ no change โ
โ QQuery 18 โ 76.29 / 79.28 ยฑ1.83 / 81.32 ms โ 76.89 / 79.02 ยฑ1.27 / 80.40 ms โ no change โ
โ QQuery 19 โ 37.39 / 37.75 ยฑ0.37 / 38.38 ms โ 37.09 / 37.52 ยฑ0.48 / 38.39 ms โ no change โ
โ QQuery 20 โ 39.63 / 40.48 ยฑ0.61 / 41.24 ms โ 41.86 / 42.89 ยฑ0.88 / 44.19 ms โ 1.06x slower โ
โ QQuery 21 โ 63.92 / 65.25 ยฑ1.04 / 66.85 ms โ 62.84 / 64.52 ยฑ1.24 / 66.02 ms โ no change โ
โ QQuery 22 โ 17.94 / 18.17 ยฑ0.15 / 18.35 ms โ 25.42 / 26.01 ยฑ0.38 / 26.59 ms โ 1.43x slower โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 899.18ms โ
โ Total Time (lazy-coalesce-partitions) โ 933.59ms โ
โ Average Time (HEAD) โ 40.87ms โ
โ Average Time (lazy-coalesce-partitions) โ 42.44ms โ
โ Queries Faster โ 0 โ
โ Queries Slower โ 6 โ
โ Queries with No Change โ 16 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpch โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 4.8s |
| Peak memory | 4.1 GiB |
| Avg memory | 3.6 GiB |
| CPU user | 33.4s |
| CPU sys | 3.0s |
| Disk read | 0 B |
| Disk write | 140.0 KiB |
**tpch โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 4.9s |
| Peak memory | 4.1 GiB |
| Avg memory | 3.6 GiB |
| CPU user | 33.3s |
| CPU sys | 3.0s |
| Disk read | 0 B |
| Disk write | 80.0 KiB |
</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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180224727
Benchmark for [this request](https://github.com/apache/datafusion/pull/21326#issuecomment-4180220240) failed.
Last 20 lines of output:
<details><summary>Click to expand</summary>
```
stats_merge
strpos
struct_query_sql
substr
substr_index
substring
sum
to_char
to_hex
to_timestamp
topk_aggregate
topk_repartition
translate
trim
trunc
unhex
upper
uuid
window_query_sql
with_hashes
```
</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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan closed pull request #21326: Defer task spawning in CoalescePartitionsExec to first poll
URL: https://github.com/apache/datafusion/pull/21326
--
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180013998
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21326#issuecomment-4180001312)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4180001312-666-txkzj 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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): 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 lazy-coalesce-partitions (e2339fb28f11ecb28d17232f6d11984422123281) to 4084a18 (merge-base) [diff](https://github.com/apache/datafusion/compare/4084a18bc709d65f9901fa0d51932867ce9f6342..e2339fb28f11ecb28d17232f6d11984422123281) using: clickbench_partitioned
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180188343
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21326#issuecomment-4180178576)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4180178576-676-srgtw 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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): 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 lazy-coalesce-partitions (e2339fb28f11ecb28d17232f6d11984422123281) to 4084a18 (merge-base) [diff](https://github.com/apache/datafusion/compare/4084a18bc709d65f9901fa0d51932867ce9f6342..e2339fb28f11ecb28d17232f6d11984422123281) using: tpch10
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180224346
๐ค Criterion benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21326#issuecomment-4180220240)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4180220240-677-hlmvg 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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): 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 lazy-coalesce-partitions (e2339fb28f11ecb28d17232f6d11984422123281) to 4084a18 (merge-base) [diff](https://github.com/apache/datafusion/compare/4084a18bc709d65f9901fa0d51932867ce9f6342..e2339fb28f11ecb28d17232f6d11984422123281)
BENCH_NAME=tpcds10
BENCH_COMMAND=cargo bench --features=parquet --bench tpcds10
BENCH_FILTER=
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180105805
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21326#issuecomment-4180001312)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<details><summary>CPU Details (lscpu)</summary>
```
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>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and lazy-coalesce-partitions
--------------------
Benchmark tpcds_sf1.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ lazy-coalesce-partitions โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 43.65 / 44.31 ยฑ0.88 / 45.99 ms โ 44.56 / 45.57 ยฑ0.70 / 46.72 ms โ no change โ
โ QQuery 2 โ 145.06 / 146.47 ยฑ0.74 / 147.00 ms โ 150.14 / 151.73 ยฑ1.03 / 153.13 ms โ no change โ
โ QQuery 3 โ 115.23 / 116.21 ยฑ0.99 / 117.93 ms โ 116.95 / 118.25 ยฑ0.96 / 119.15 ms โ no change โ
โ QQuery 4 โ 1386.84 / 1422.69 ยฑ23.95 / 1454.01 ms โ 1718.64 / 1763.56 ยฑ27.71 / 1797.66 ms โ 1.24x slower โ
โ QQuery 5 โ 176.05 / 178.21 ยฑ1.66 / 181.08 ms โ 175.49 / 177.08 ยฑ1.20 / 178.68 ms โ no change โ
โ QQuery 6 โ 1047.37 / 1093.65 ยฑ30.09 / 1128.21 ms โ 1032.39 / 1064.29 ยฑ35.51 / 1110.85 ms โ no change โ
โ QQuery 7 โ 361.65 / 366.25 ยฑ3.32 / 371.53 ms โ 355.82 / 361.09 ยฑ5.47 / 371.63 ms โ no change โ
โ QQuery 8 โ 118.88 / 119.45 ยฑ0.43 / 120.19 ms โ 120.38 / 122.10 ยฑ1.04 / 123.37 ms โ no change โ
โ QQuery 9 โ 102.37 / 106.58 ยฑ3.23 / 110.43 ms โ 812.67 / 816.71 ยฑ3.08 / 821.84 ms โ 7.66x slower โ
โ QQuery 10 โ 108.48 / 110.88 ยฑ2.05 / 113.50 ms โ 130.80 / 132.55 ยฑ1.10 / 134.11 ms โ 1.20x slower โ
โ QQuery 11 โ 1009.78 / 1023.99 ยฑ15.63 / 1053.16 ms โ 1150.50 / 1160.09 ยฑ7.30 / 1172.72 ms โ 1.13x slower โ
โ QQuery 12 โ 45.78 / 47.77 ยฑ1.17 / 48.98 ms โ 47.51 / 48.95 ยฑ1.60 / 51.97 ms โ no change โ
โ QQuery 13 โ 407.40 / 409.72 ยฑ1.63 / 411.84 ms โ 406.38 / 409.82 ยฑ2.80 / 414.79 ms โ no change โ
โ QQuery 14 โ 1025.40 / 1036.56 ยฑ12.20 / 1059.35 ms โ 1741.56 / 1752.12 ยฑ8.84 / 1765.37 ms โ 1.69x slower โ
โ QQuery 15 โ 16.26 / 17.23 ยฑ0.77 / 18.06 ms โ 16.51 / 17.77 ยฑ1.15 / 19.25 ms โ no change โ
โ QQuery 16 โ 41.41 / 42.52 ยฑ0.75 / 43.63 ms โ 42.23 / 43.44 ยฑ0.70 / 44.31 ms โ no change โ
โ QQuery 17 โ 244.04 / 245.74 ยฑ1.11 / 247.16 ms โ 242.55 / 245.71 ยฑ3.11 / 251.68 ms โ no change โ
โ QQuery 18 โ 130.71 / 132.46 ยฑ1.51 / 135.12 ms โ 132.30 / 134.25 ยฑ1.56 / 136.23 ms โ no change โ
โ QQuery 19 โ 158.08 / 159.66 ยฑ1.09 / 161.38 ms โ 156.37 / 157.99 ยฑ1.28 / 159.67 ms โ no change โ
โ QQuery 20 โ 14.36 / 15.04 ยฑ0.77 / 16.24 ms โ 15.90 / 16.64 ยฑ0.44 / 17.06 ms โ 1.11x slower โ
โ QQuery 21 โ 20.05 / 20.71 ยฑ0.61 / 21.80 ms โ 21.10 / 21.66 ยฑ0.65 / 22.91 ms โ no change โ
โ QQuery 22 โ 491.13 / 493.92 ยฑ1.52 / 495.70 ms โ 496.03 / 498.69 ยฑ2.70 / 501.99 ms โ no change โ
โ QQuery 23 โ 919.44 / 932.73 ยฑ8.45 / 942.20 ms โ 1230.52 / 1260.30 ยฑ20.77 / 1293.27 ms โ 1.35x slower โ
โ QQuery 24 โ 422.32 / 424.25 ยฑ2.58 / 429.29 ms โ 816.31 / 820.61 ยฑ4.35 / 828.97 ms โ 1.93x slower โ
โ QQuery 25 โ 360.72 / 363.27 ยฑ1.97 / 365.63 ms โ 359.14 / 362.76 ยฑ2.62 / 366.35 ms โ no change โ
โ QQuery 26 โ 82.06 / 84.46 ยฑ1.39 / 85.86 ms โ 82.95 / 84.53 ยฑ1.40 / 86.81 ms โ no change โ
โ QQuery 27 โ 353.82 / 357.94 ยฑ3.32 / 362.41 ms โ 351.04 / 352.95 ยฑ2.04 / 356.68 ms โ no change โ
โ QQuery 28 โ 151.54 / 152.79 ยฑ0.98 / 154.23 ms โ 844.73 / 847.12 ยฑ2.16 / 849.90 ms โ 5.54x slower โ
โ QQuery 29 โ 302.99 / 307.74 ยฑ3.15 / 312.32 ms โ 303.55 / 304.99 ยฑ1.26 / 306.96 ms โ no change โ
โ QQuery 30 โ 42.60 / 44.73 ยฑ1.75 / 47.88 ms โ 51.65 / 52.80 ยฑ0.90 / 53.90 ms โ 1.18x slower โ
โ QQuery 31 โ 172.08 / 174.81 ยฑ1.94 / 178.02 ms โ 426.34 / 428.69 ยฑ2.41 / 432.98 ms โ 2.45x slower โ
โ QQuery 32 โ 57.73 / 59.83 ยฑ1.35 / 61.51 ms โ 58.28 / 59.29 ยฑ0.77 / 60.56 ms โ no change โ
โ QQuery 33 โ 143.46 / 145.63 ยฑ1.28 / 147.13 ms โ 143.70 / 145.81 ยฑ1.15 / 146.96 ms โ no change โ
โ QQuery 34 โ 108.72 / 110.20 ยฑ1.21 / 112.10 ms โ 109.44 / 110.65 ยฑ0.99 / 112.28 ms โ no change โ
โ QQuery 35 โ 111.13 / 112.14 ยฑ1.31 / 114.63 ms โ 136.96 / 138.14 ยฑ1.03 / 139.99 ms โ 1.23x slower โ
โ QQuery 36 โ 219.48 / 223.16 ยฑ2.48 / 226.22 ms โ 214.68 / 220.36 ยฑ3.69 / 226.19 ms โ no change โ
โ QQuery 37 โ 181.11 / 183.16 ยฑ1.13 / 184.20 ms โ 178.89 / 181.96 ยฑ1.99 / 184.79 ms โ no change โ
โ QQuery 38 โ 86.84 / 90.04 ยฑ3.05 / 93.83 ms โ 118.76 / 121.43 ยฑ1.68 / 123.86 ms โ 1.35x slower โ
โ QQuery 39 โ 127.04 / 129.82 ยฑ2.49 / 134.33 ms โ 127.98 / 134.06 ยฑ3.87 / 139.63 ms โ no change โ
โ QQuery 40 โ 111.89 / 118.91 ยฑ6.43 / 130.60 ms โ 110.28 / 119.63 ยฑ9.39 / 137.25 ms โ no change โ
โ QQuery 41 โ 14.67 / 15.72 ยฑ0.99 / 17.33 ms โ 15.56 / 16.66 ยฑ0.91 / 18.15 ms โ 1.06x slower โ
โ QQuery 42 โ 108.66 / 110.24 ยฑ1.15 / 111.39 ms โ 109.60 / 111.22 ยฑ1.98 / 114.47 ms โ no change โ
โ QQuery 43 โ 84.34 / 85.84 ยฑ1.41 / 87.63 ms โ 85.40 / 86.02 ยฑ1.06 / 88.13 ms โ no change โ
โ QQuery 44 โ 12.06 / 12.46 ยฑ0.42 / 13.25 ms โ 11.38 / 11.60 ยฑ0.12 / 11.71 ms โ +1.07x faster โ
โ QQuery 45 โ 53.46 / 54.34 ยฑ0.98 / 55.76 ms โ 53.57 / 54.53 ยฑ0.77 / 55.48 ms โ no change โ
โ QQuery 46 โ 235.65 / 238.81 ยฑ2.71 / 242.76 ms โ 240.34 / 242.92 ยฑ2.20 / 246.67 ms โ no change โ
โ QQuery 47 โ 734.05 / 740.68 ยฑ3.91 / 745.77 ms โ 754.30 / 759.29 ยฑ5.91 / 770.78 ms โ no change โ
โ QQuery 48 โ 294.73 / 298.10 ยฑ2.56 / 301.14 ms โ 281.89 / 292.82 ยฑ6.04 / 299.64 ms โ no change โ
โ QQuery 49 โ 256.75 / 259.44 ยฑ2.08 / 262.57 ms โ 256.47 / 258.57 ยฑ1.68 / 261.30 ms โ no change โ
โ QQuery 50 โ 237.07 / 240.53 ยฑ3.20 / 245.52 ms โ 234.42 / 238.39 ยฑ3.67 / 244.69 ms โ no change โ
โ QQuery 51 โ 184.91 / 186.44 ยฑ1.07 / 188.11 ms โ 186.15 / 188.97 ยฑ1.69 / 191.19 ms โ no change โ
โ QQuery 52 โ 109.19 / 110.03 ยฑ0.82 / 111.05 ms โ 109.61 / 110.77 ยฑ1.47 / 113.62 ms โ no change โ
โ QQuery 53 โ 104.35 / 105.37 ยฑ0.94 / 106.73 ms โ 106.05 / 106.70 ยฑ0.71 / 107.79 ms โ no change โ
โ QQuery 54 โ 150.47 / 151.62 ยฑ0.91 / 153.08 ms โ 151.38 / 154.20 ยฑ1.57 / 155.59 ms โ no change โ
โ QQuery 55 โ 108.53 / 109.16 ยฑ0.54 / 109.90 ms โ 109.46 / 110.50 ยฑ1.50 / 113.45 ms โ no change โ
โ QQuery 56 โ 142.34 / 143.70 ยฑ1.56 / 146.62 ms โ 142.64 / 145.56 ยฑ1.74 / 147.82 ms โ no change โ
โ QQuery 57 โ 174.79 / 175.62 ยฑ0.42 / 175.89 ms โ 176.96 / 180.63 ยฑ2.96 / 185.66 ms โ no change โ
โ QQuery 58 โ 299.64 / 310.90 ยฑ7.98 / 320.51 ms โ 387.61 / 395.90 ยฑ8.53 / 410.80 ms โ 1.27x slower โ
โ QQuery 59 โ 200.04 / 204.63 ยฑ5.70 / 215.63 ms โ 381.08 / 385.06 ยฑ3.04 / 389.47 ms โ 1.88x slower โ
โ QQuery 60 โ 145.13 / 146.29 ยฑ1.01 / 147.92 ms โ 147.54 / 149.30 ยฑ1.35 / 151.71 ms โ no change โ
โ QQuery 61 โ 173.41 / 174.84 ยฑ1.24 / 176.60 ms โ 322.70 / 326.26 ยฑ2.30 / 328.61 ms โ 1.87x slower โ
โ QQuery 62 โ 925.94 / 982.11 ยฑ45.10 / 1031.40 ms โ 946.29 / 970.27 ยฑ18.32 / 989.77 ms โ no change โ
โ QQuery 63 โ 106.22 / 107.45 ยฑ1.43 / 110.18 ms โ 107.73 / 110.83 ยฑ2.73 / 114.17 ms โ no change โ
โ QQuery 64 โ 710.74 / 714.84 ยฑ3.70 / 720.59 ms โ 1378.78 / 1390.37 ยฑ12.78 / 1413.96 ms โ 1.95x slower โ
โ QQuery 65 โ 254.10 / 258.95 ยฑ2.82 / 261.65 ms โ 259.35 / 261.52 ยฑ2.53 / 266.27 ms โ no change โ
โ QQuery 66 โ 232.55 / 254.65 ยฑ15.58 / 278.45 ms โ 248.78 / 269.87 ยฑ14.56 / 290.52 ms โ 1.06x slower โ
โ QQuery 67 โ 320.84 / 324.89 ยฑ2.92 / 329.78 ms โ 323.44 / 331.52 ยฑ5.17 / 339.08 ms โ no change โ
โ QQuery 68 โ 284.99 / 289.20 ยฑ3.07 / 293.12 ms โ 289.91 / 291.88 ยฑ2.00 / 295.52 ms โ no change โ
โ QQuery 69 โ 102.47 / 104.27 ยฑ1.17 / 105.95 ms โ 124.42 / 126.22 ยฑ1.47 / 128.49 ms โ 1.21x slower โ
โ QQuery 70 โ 336.64 / 348.92 ยฑ6.85 / 356.50 ms โ 351.19 / 363.98 ยฑ13.68 / 385.64 ms โ no change โ
โ QQuery 71 โ 135.12 / 136.76 ยฑ1.10 / 138.06 ms โ 137.03 / 140.51 ยฑ3.35 / 145.99 ms โ no change โ
โ QQuery 72 โ 731.73 / 734.55 ยฑ2.75 / 738.76 ms โ 739.00 / 747.82 ยฑ10.02 / 761.12 ms โ no change โ
โ QQuery 73 โ 104.84 / 106.77 ยฑ2.01 / 110.11 ms โ 105.10 / 106.39 ยฑ1.42 / 109.17 ms โ no change โ
โ QQuery 74 โ 603.27 / 618.14 ยฑ9.09 / 626.11 ms โ 703.05 / 719.18 ยฑ8.45 / 727.77 ms โ 1.16x slower โ
โ QQuery 75 โ 278.50 / 279.56 ยฑ0.83 / 280.71 ms โ 459.07 / 463.84 ยฑ2.96 / 468.28 ms โ 1.66x slower โ
โ QQuery 76 โ 136.48 / 137.87 ยฑ1.50 / 140.59 ms โ 134.83 / 136.95 ยฑ1.69 / 139.25 ms โ no change โ
โ QQuery 77 โ 189.12 / 191.74 ยฑ1.60 / 193.95 ms โ 191.08 / 194.78 ยฑ2.12 / 197.24 ms โ no change โ
โ QQuery 78 โ 358.00 / 361.08 ยฑ2.37 / 364.07 ms โ 403.52 / 410.05 ยฑ4.93 / 416.53 ms โ 1.14x slower โ
โ QQuery 79 โ 237.12 / 240.41 ยฑ2.28 / 244.05 ms โ 247.05 / 249.20 ยฑ1.94 / 252.61 ms โ no change โ
โ QQuery 80 โ 335.73 / 338.63 ยฑ2.43 / 341.64 ms โ 334.68 / 338.93 ยฑ3.58 / 342.79 ms โ no change โ
โ QQuery 81 โ 27.06 / 28.14 ยฑ0.70 / 29.08 ms โ 36.32 / 39.11 ยฑ3.94 / 46.86 ms โ 1.39x slower โ
โ QQuery 82 โ 204.22 / 208.16 ยฑ2.43 / 211.60 ms โ 202.19 / 204.13 ยฑ2.40 / 208.58 ms โ no change โ
โ QQuery 83 โ 41.12 / 42.41 ยฑ1.15 / 44.18 ms โ 53.15 / 54.19 ยฑ1.07 / 56.04 ms โ 1.28x slower โ
โ QQuery 84 โ 50.42 / 50.93 ยฑ0.42 / 51.68 ms โ 50.34 / 51.44 ยฑ0.84 / 52.77 ms โ no change โ
โ QQuery 85 โ 152.69 / 154.76 ยฑ2.47 / 159.32 ms โ 150.77 / 152.92 ยฑ1.52 / 154.90 ms โ no change โ
โ QQuery 86 โ 40.10 / 41.19 ยฑ0.66 / 42.10 ms โ 39.91 / 41.07 ยฑ0.71 / 42.08 ms โ no change โ
โ QQuery 87 โ 90.50 / 93.73 ยฑ2.65 / 98.42 ms โ 121.66 / 124.38 ยฑ2.49 / 128.31 ms โ 1.33x slower โ
โ QQuery 88 โ 101.54 / 103.01 ยฑ1.18 / 105.08 ms โ 591.65 / 594.48 ยฑ2.25 / 597.31 ms โ 5.77x slower โ
โ QQuery 89 โ 121.79 / 123.14 ยฑ0.84 / 124.43 ms โ 121.51 / 123.08 ยฑ1.41 / 125.57 ms โ no change โ
โ QQuery 90 โ 24.53 / 25.91 ยฑ1.76 / 29.37 ms โ 39.46 / 41.40 ยฑ1.13 / 42.51 ms โ 1.60x slower โ
โ QQuery 91 โ 65.45 / 66.39 ยฑ1.02 / 68.34 ms โ 64.32 / 64.99 ยฑ0.41 / 65.50 ms โ no change โ
โ QQuery 92 โ 60.06 / 61.32 ยฑ1.47 / 64.18 ms โ 59.09 / 60.33 ยฑ0.91 / 61.92 ms โ no change โ
โ QQuery 93 โ 198.58 / 201.10 ยฑ2.63 / 205.11 ms โ 193.58 / 195.80 ยฑ1.72 / 197.78 ms โ no change โ
โ QQuery 94 โ 62.83 / 64.21 ยฑ0.76 / 65.11 ms โ 63.29 / 63.85 ยฑ0.53 / 64.85 ms โ no change โ
โ QQuery 95 โ 137.78 / 138.86 ยฑ1.26 / 141.22 ms โ 135.15 / 136.98 ยฑ1.54 / 138.83 ms โ no change โ
โ QQuery 96 โ 75.14 / 76.17 ยฑ0.63 / 77.12 ms โ 71.72 / 74.60 ยฑ1.58 / 76.56 ms โ no change โ
โ QQuery 97 โ 132.27 / 137.45 ยฑ3.03 / 141.32 ms โ 129.46 / 133.09 ยฑ1.99 / 135.04 ms โ no change โ
โ QQuery 98 โ 156.44 / 161.32 ยฑ2.95 / 165.35 ms โ 155.30 / 158.06 ยฑ2.39 / 161.30 ms โ no change โ
โ QQuery 99 โ 10847.83 / 10901.62 ยฑ39.37 / 10946.22 ms โ 10795.79 / 10850.41 ยฑ37.54 / 10902.77 ms โ no change โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 34450.98ms โ
โ Total Time (lazy-coalesce-partitions) โ 40120.45ms โ
โ Average Time (HEAD) โ 347.99ms โ
โ Average Time (lazy-coalesce-partitions) โ 405.26ms โ
โ Queries Faster โ 1 โ
โ Queries Slower โ 28 โ
โ Queries with No Change โ 70 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 172.6s |
| Peak memory | 5.8 GiB |
| Avg memory | 4.7 GiB |
| CPU user | 276.2s |
| CPU sys | 19.5s |
| Disk read | 0 B |
| Disk write | 707.5 MiB |
**tpcds โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 200.9s |
| Peak memory | 5.7 GiB |
| Avg memory | 4.4 GiB |
| CPU user | 276.1s |
| CPU sys | 20.9s |
| Disk read | 0 B |
| Disk write | 796.0 KiB |
</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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180001312
run benchmarks
--
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180103816
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21326#issuecomment-4180001312)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<details><summary>CPU Details (lscpu)</summary>
```
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>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and lazy-coalesce-partitions
--------------------
Benchmark clickbench_partitioned.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ lazy-coalesce-partitions โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 0 โ 1.30 / 4.48 ยฑ6.29 / 17.05 ms โ 1.30 / 4.49 ยฑ6.30 / 17.08 ms โ no change โ
โ QQuery 1 โ 14.07 / 14.47 ยฑ0.25 / 14.78 ms โ 14.05 / 14.48 ยฑ0.26 / 14.86 ms โ no change โ
โ QQuery 2 โ 44.02 / 44.50 ยฑ0.42 / 45.16 ms โ 43.96 / 44.31 ยฑ0.25 / 44.61 ms โ no change โ
โ QQuery 3 โ 43.34 / 44.81 ยฑ1.28 / 46.97 ms โ 41.83 / 45.22 ยฑ2.44 / 48.48 ms โ no change โ
โ QQuery 4 โ 286.77 / 294.40 ยฑ4.07 / 298.04 ms โ 292.58 / 303.72 ยฑ7.01 / 312.89 ms โ no change โ
โ QQuery 5 โ 344.12 / 348.24 ยฑ3.08 / 352.55 ms โ 342.68 / 345.99 ยฑ4.48 / 354.84 ms โ no change โ
โ QQuery 6 โ 5.28 / 6.14 ยฑ0.55 / 7.01 ms โ 5.21 / 5.96 ยฑ0.51 / 6.51 ms โ no change โ
โ QQuery 7 โ 16.45 / 17.07 ยฑ0.84 / 18.74 ms โ 16.29 / 16.91 ยฑ0.80 / 18.46 ms โ no change โ
โ QQuery 8 โ 416.65 / 425.74 ยฑ7.51 / 435.13 ms โ 424.12 / 430.53 ยฑ4.73 / 438.67 ms โ no change โ
โ QQuery 9 โ 626.14 / 650.39 ยฑ46.04 / 742.47 ms โ 625.23 / 639.55 ยฑ10.06 / 651.58 ms โ no change โ
โ QQuery 10 โ 91.82 / 94.87 ยฑ2.21 / 98.04 ms โ 92.24 / 95.14 ยฑ2.00 / 98.12 ms โ no change โ
โ QQuery 11 โ 103.65 / 105.56 ยฑ1.10 / 106.64 ms โ 102.58 / 103.43 ยฑ0.73 / 104.60 ms โ no change โ
โ QQuery 12 โ 339.60 / 344.79 ยฑ3.21 / 348.87 ms โ 347.61 / 351.58 ยฑ2.95 / 355.38 ms โ no change โ
โ QQuery 13 โ 459.53 / 477.69 ยฑ14.85 / 495.69 ms โ 442.96 / 472.88 ยฑ16.69 / 492.94 ms โ no change โ
โ QQuery 14 โ 350.36 / 356.42 ยฑ4.60 / 363.76 ms โ 350.69 / 357.19 ยฑ5.49 / 366.32 ms โ no change โ
โ QQuery 15 โ 358.79 / 381.35 ยฑ24.32 / 428.52 ms โ 367.85 / 380.39 ยฑ11.28 / 400.61 ms โ no change โ
โ QQuery 16 โ 738.12 / 892.19 ยฑ269.77 / 1430.95 ms โ 731.52 / 755.38 ยฑ25.42 / 801.76 ms โ +1.18x faster โ
โ QQuery 17 โ 711.91 / 721.03 ยฑ7.24 / 732.87 ms โ 728.02 / 734.35 ยฑ4.18 / 739.54 ms โ no change โ
โ QQuery 18 โ 1439.32 / 1510.03 ยฑ37.82 / 1551.57 ms โ 1445.45 / 1472.06 ยฑ26.29 / 1518.53 ms โ no change โ
โ QQuery 19 โ 35.72 / 37.28 ยฑ1.26 / 38.97 ms โ 35.87 / 64.04 ยฑ35.76 / 127.11 ms โ 1.72x slower โ
โ QQuery 20 โ 713.27 / 727.75 ยฑ14.20 / 749.14 ms โ 714.19 / 735.34 ยฑ24.50 / 780.16 ms โ no change โ
โ QQuery 21 โ 758.01 / 773.45 ยฑ21.01 / 814.30 ms โ 756.78 / 763.36 ยฑ4.96 / 772.17 ms โ no change โ
โ QQuery 22 โ 1124.82 / 1130.75 ยฑ5.73 / 1140.69 ms โ 1126.16 / 1129.07 ยฑ2.56 / 1133.75 ms โ no change โ
โ QQuery 23 โ 3053.39 / 3072.91 ยฑ19.07 / 3099.67 ms โ 3042.73 / 3066.87 ยฑ15.36 / 3084.04 ms โ no change โ
โ QQuery 24 โ 97.14 / 101.91 ยฑ2.83 / 105.35 ms โ 99.76 / 102.95 ยฑ3.14 / 106.81 ms โ no change โ
โ QQuery 25 โ 136.61 / 139.36 ยฑ1.95 / 142.45 ms โ 137.78 / 140.03 ยฑ1.23 / 141.06 ms โ no change โ
โ QQuery 26 โ 97.05 / 100.46 ยฑ2.24 / 104.02 ms โ 95.49 / 99.81 ยฑ2.47 / 102.49 ms โ no change โ
โ QQuery 27 โ 847.44 / 852.61 ยฑ5.15 / 861.78 ms โ 846.17 / 853.55 ยฑ8.86 / 868.46 ms โ no change โ
โ QQuery 28 โ 7745.51 / 7823.45 ยฑ50.59 / 7890.84 ms โ 7678.20 / 7731.42 ยฑ30.23 / 7766.80 ms โ no change โ
โ QQuery 29 โ 50.64 / 55.82 ยฑ5.09 / 64.62 ms โ 49.98 / 53.92 ยฑ4.40 / 59.46 ms โ no change โ
โ QQuery 30 โ 362.11 / 367.82 ยฑ6.19 / 378.74 ms โ 364.09 / 369.39 ยฑ6.95 / 382.59 ms โ no change โ
โ QQuery 31 โ 364.99 / 381.51 ยฑ8.96 / 390.87 ms โ 368.02 / 380.47 ยฑ11.97 / 399.87 ms โ no change โ
โ QQuery 32 โ 1039.12 / 1051.88 ยฑ16.98 / 1084.97 ms โ 1170.50 / 1222.72 ยฑ53.04 / 1299.37 ms โ 1.16x slower โ
โ QQuery 33 โ 1449.21 / 1470.76 ยฑ11.36 / 1480.17 ms โ 1438.55 / 1458.49 ยฑ16.79 / 1483.50 ms โ no change โ
โ QQuery 34 โ 1450.04 / 1457.91 ยฑ8.55 / 1473.82 ms โ 1426.98 / 1446.02 ยฑ19.37 / 1483.35 ms โ no change โ
โ QQuery 35 โ 383.70 / 391.35 ยฑ7.03 / 400.84 ms โ 385.45 / 391.07 ยฑ3.71 / 394.80 ms โ no change โ
โ QQuery 36 โ 118.47 / 122.91 ยฑ2.48 / 125.17 ms โ 122.38 / 123.87 ยฑ1.64 / 126.39 ms โ no change โ
โ QQuery 37 โ 47.49 / 48.94 ยฑ1.68 / 52.04 ms โ 48.63 / 50.08 ยฑ1.23 / 52.25 ms โ no change โ
โ QQuery 38 โ 75.02 / 76.46 ยฑ1.74 / 79.76 ms โ 74.78 / 76.22 ยฑ0.91 / 77.55 ms โ no change โ
โ QQuery 39 โ 219.93 / 222.85 ยฑ3.56 / 229.82 ms โ 208.06 / 213.70 ยฑ4.45 / 220.04 ms โ no change โ
โ QQuery 40 โ 23.58 / 25.70 ยฑ1.75 / 28.38 ms โ 22.97 / 25.63 ยฑ1.54 / 27.26 ms โ no change โ
โ QQuery 41 โ 18.74 / 20.57 ยฑ1.18 / 22.08 ms โ 20.59 / 20.80 ยฑ0.25 / 21.28 ms โ no change โ
โ QQuery 42 โ 18.92 / 19.38 ยฑ0.32 / 19.78 ms โ 18.51 / 19.49 ยฑ0.61 / 20.30 ms โ no change โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 27207.98ms โ
โ Total Time (lazy-coalesce-partitions) โ 27111.86ms โ
โ Average Time (HEAD) โ 632.74ms โ
โ Average Time (lazy-coalesce-partitions) โ 630.51ms โ
โ Queries Faster โ 1 โ
โ Queries Slower โ 2 โ
โ Queries with No Change โ 40 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 137.2s |
| Peak memory | 44.5 GiB |
| Avg memory | 32.5 GiB |
| CPU user | 1289.7s |
| CPU sys | 91.9s |
| Disk read | 0 B |
| Disk write | 4.4 GiB |
**clickbench_partitioned โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 136.6s |
| Peak memory | 40.3 GiB |
| Avg memory | 30.1 GiB |
| CPU user | 1277.2s |
| CPU sys | 96.9s |
| Disk read | 0 B |
| Disk write | 748.0 KiB |
</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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180220240
run benchmark tpcds10
--
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180010940
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21326#issuecomment-4180001312)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4180001312-667-8p9qw 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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): 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 lazy-coalesce-partitions (e2339fb28f11ecb28d17232f6d11984422123281) to 4084a18 (merge-base) [diff](https://github.com/apache/datafusion/compare/4084a18bc709d65f9901fa0d51932867ce9f6342..e2339fb28f11ecb28d17232f6d11984422123281) using: tpcds
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180188536
๐ค Criterion benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21326#issuecomment-4180178576)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4180178576-675-csc82 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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): 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 lazy-coalesce-partitions (e2339fb28f11ecb28d17232f6d11984422123281) to 4084a18 (merge-base) [diff](https://github.com/apache/datafusion/compare/4084a18bc709d65f9901fa0d51932867ce9f6342..e2339fb28f11ecb28d17232f6d11984422123281)
BENCH_NAME=tpcd10
BENCH_COMMAND=cargo bench --features=parquet --bench tpcd10
BENCH_FILTER=
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "Dandandan (via GitHub)" <gi...@apache.org>.
Dandandan commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180178576
run benchmark tpcd10 tpch10
--
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180190080
Benchmark for [this request](https://github.com/apache/datafusion/pull/21326#issuecomment-4180178576) failed.
Last 20 lines of output:
<details><summary>Click to expand</summary>
```
stats_merge
strpos
struct_query_sql
substr
substr_index
substring
sum
to_char
to_hex
to_timestamp
topk_aggregate
topk_repartition
translate
trim
trunc
unhex
upper
uuid
window_query_sql
with_hashes
```
</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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180011376
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/21326#issuecomment-4180001312)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c4180001312-668-7n4cg 6.12.55+ #1 SMP Sun Feb 1 08:59:41 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): 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 lazy-coalesce-partitions (e2339fb28f11ecb28d17232f6d11984422123281) to 4084a18 (merge-base) [diff](https://github.com/apache/datafusion/compare/4084a18bc709d65f9901fa0d51932867ce9f6342..e2339fb28f11ecb28d17232f6d11984422123281) using: tpch
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] Defer task spawning in CoalescePartitionsExec to first poll [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #21326:
URL: https://github.com/apache/datafusion/pull/21326#issuecomment-4180240764
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/21326#issuecomment-4180178576)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB)
<details><summary>CPU Details (lscpu)</summary>
```
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>
<details><summary>Details</summary>
<p>
```
Comparing HEAD and lazy-coalesce-partitions
--------------------
Benchmark tpch_sf10.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโ
โ Query โ HEAD โ lazy-coalesce-partitions โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 375.54 / 380.19 ยฑ4.02 / 385.25 ms โ 374.01 / 381.51 ยฑ5.07 / 386.71 ms โ no change โ
โ QQuery 2 โ 139.11 / 143.43 ยฑ2.45 / 146.08 ms โ 142.51 / 146.69 ยฑ2.60 / 150.67 ms โ no change โ
โ QQuery 3 โ 291.44 / 301.66 ยฑ8.41 / 310.63 ms โ 307.38 / 311.28 ยฑ2.93 / 315.87 ms โ no change โ
โ QQuery 4 โ 155.68 / 157.66 ยฑ2.09 / 161.64 ms โ 158.59 / 163.28 ยฑ2.89 / 166.34 ms โ no change โ
โ QQuery 5 โ 433.43 / 457.81 ยฑ12.53 / 468.98 ms โ 476.18 / 482.11 ยฑ8.52 / 498.95 ms โ 1.05x slower โ
โ QQuery 6 โ 133.20 / 134.93 ยฑ1.00 / 136.12 ms โ 140.06 / 140.99 ยฑ0.82 / 142.51 ms โ no change โ
โ QQuery 7 โ 573.24 / 587.07 ยฑ9.34 / 600.60 ms โ 590.60 / 612.64 ยฑ18.48 / 640.42 ms โ no change โ
โ QQuery 8 โ 475.53 / 498.44 ยฑ17.22 / 522.40 ms โ 481.04 / 499.65 ยฑ15.94 / 523.67 ms โ no change โ
โ QQuery 9 โ 699.10 / 711.96 ยฑ13.30 / 734.30 ms โ 691.48 / 707.61 ยฑ11.99 / 726.40 ms โ no change โ
โ QQuery 10 โ 342.21 / 357.16 ยฑ12.60 / 377.59 ms โ 337.50 / 349.70 ยฑ8.68 / 364.39 ms โ no change โ
โ QQuery 11 โ 107.09 / 114.32 ยฑ4.56 / 120.49 ms โ 109.98 / 115.89 ยฑ4.77 / 122.74 ms โ no change โ
โ QQuery 12 โ 211.31 / 215.33 ยฑ4.36 / 222.89 ms โ 208.04 / 212.86 ยฑ3.48 / 218.13 ms โ no change โ
โ QQuery 13 โ 306.02 / 327.74 ยฑ13.99 / 346.79 ms โ 316.70 / 334.96 ยฑ14.77 / 361.51 ms โ no change โ
โ QQuery 14 โ 184.76 / 189.95 ยฑ3.70 / 195.79 ms โ 186.58 / 193.78 ยฑ4.26 / 199.06 ms โ no change โ
โ QQuery 15 โ 344.70 / 351.45 ยฑ5.63 / 361.16 ms โ 340.53 / 347.38 ยฑ4.09 / 352.38 ms โ no change โ
โ QQuery 16 โ 84.03 / 86.69 ยฑ2.29 / 89.46 ms โ 83.18 / 90.16 ยฑ4.26 / 95.32 ms โ no change โ
โ QQuery 17 โ 768.73 / 793.22 ยฑ18.43 / 813.46 ms โ 773.48 / 799.10 ยฑ15.98 / 823.40 ms โ no change โ
โ QQuery 18 โ 877.35 / 898.91 ยฑ11.82 / 911.79 ms โ 881.69 / 902.54 ยฑ10.92 / 912.13 ms โ no change โ
โ QQuery 19 โ 278.22 / 283.62 ยฑ5.73 / 294.65 ms โ 271.59 / 282.25 ยฑ12.53 / 305.74 ms โ no change โ
โ QQuery 20 โ 327.55 / 335.36 ยฑ6.79 / 346.40 ms โ 328.31 / 341.28 ยฑ8.93 / 351.43 ms โ no change โ
โ QQuery 21 โ 837.46 / 873.32 ยฑ18.71 / 890.70 ms โ 856.88 / 886.85 ยฑ22.27 / 918.29 ms โ no change โ
โ QQuery 22 โ 82.27 / 83.99 ยฑ1.93 / 87.02 ms โ 81.51 / 86.84 ยฑ4.19 / 90.62 ms โ no change โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 8284.21ms โ
โ Total Time (lazy-coalesce-partitions) โ 8389.33ms โ
โ Average Time (HEAD) โ 376.56ms โ
โ Average Time (lazy-coalesce-partitions) โ 381.33ms โ
โ Queries Faster โ 0 โ
โ Queries Slower โ 1 โ
โ Queries with No Change โ 21 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpch10 โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 41.8s |
| Peak memory | 10.0 GiB |
| Avg memory | 7.7 GiB |
| CPU user | 434.8s |
| CPU sys | 31.6s |
| Disk read | 0 B |
| Disk write | 2.9 GiB |
**tpch10 โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 42.3s |
| Peak memory | 10.0 GiB |
| Avg memory | 7.7 GiB |
| CPU user | 440.1s |
| CPU sys | 32.5s |
| Disk read | 0 B |
| Disk write | 204.0 KiB |
</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