You are viewing a plain text version of this content. The canonical link for it is here.
Posted to github@datafusion.apache.org by "rluvaton (via GitHub)" <gi...@apache.org> on 2026/07/29 15:05:08 UTC
[PR] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
rluvaton opened a new pull request, #23976:
URL: https://github.com/apache/datafusion/pull/23976
## Which issue does this PR close?
Part of:
- https://github.com/apache/datafusion/issues/23974
## Rationale for this change
Simplify the code and make
## What changes are included in this PR?
Changed to async generators + try to simplify to be similar to textbook, but did not get to that point that when I look at the main function I see the algorithm from the textbook
## Are these changes tested?
existing tests
## Are there any user-facing changes?
nope
--
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] chore: refactor `MaterializingSortMergeJoinStream` into generators and simplify code to be textbook like as possible [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5134543710
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5134520421)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5134520421-1307-v5r7k 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (cb881c764a1ca7d87fb3fb872ee26b944ad387e4) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..cb881c764a1ca7d87fb3fb872ee26b944ad387e4) 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5132847009
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5132631045)
**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 refactor-materializing-smj
--------------------
Benchmark smj.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 8.64 / 8.73 ยฑ0.15 / 9.02 ms โ 8.09 / 8.40 ยฑ0.24 / 8.82 ms โ no change โ
โ QQuery 2 โ 162.14 / 164.25 ยฑ1.69 / 166.49 ms โ 161.72 / 163.73 ยฑ1.94 / 166.85 ms โ no change โ
โ QQuery 3 โ 103.51 / 105.17 ยฑ0.99 / 106.60 ms โ 102.30 / 105.50 ยฑ1.95 / 107.96 ms โ no change โ
โ QQuery 4 โ 28.49 / 28.84 ยฑ0.25 / 29.15 ms โ 28.02 / 28.18 ยฑ0.11 / 28.32 ms โ no change โ
โ QQuery 5 โ 20.52 / 20.76 ยฑ0.14 / 20.94 ms โ 20.43 / 20.47 ยฑ0.05 / 20.55 ms โ no change โ
โ QQuery 6 โ 158.00 / 162.26 ยฑ4.05 / 169.96 ms โ 155.49 / 159.62 ยฑ2.51 / 162.73 ms โ no change โ
โ QQuery 7 โ 200.87 / 203.32 ยฑ1.57 / 205.41 ms โ 199.07 / 200.54 ยฑ0.80 / 201.37 ms โ no change โ
โ QQuery 8 โ 19.67 / 19.93 ยฑ0.14 / 20.10 ms โ 19.10 / 19.26 ยฑ0.13 / 19.40 ms โ no change โ
โ QQuery 9 โ 207.98 / 214.00 ยฑ3.48 / 217.52 ms โ 206.72 / 208.83 ยฑ2.23 / 212.94 ms โ no change โ
โ QQuery 10 โ 57.65 / 58.16 ยฑ0.32 / 58.48 ms โ 59.15 / 59.45 ยฑ0.23 / 59.78 ms โ no change โ
โ QQuery 11 โ 27.30 / 30.07 ยฑ4.43 / 38.87 ms โ 27.63 / 31.22 ยฑ6.00 / 43.15 ms โ no change โ
โ QQuery 12 โ 66.72 / 69.85 ยฑ2.25 / 73.20 ms โ 70.66 / 73.03 ยฑ1.33 / 74.59 ms โ no change โ
โ QQuery 13 โ 96.82 / 105.70 ยฑ4.56 / 109.12 ms โ 101.37 / 108.36 ยฑ3.67 / 111.89 ms โ no change โ
โ QQuery 14 โ 58.22 / 60.00 ยฑ2.90 / 65.77 ms โ 59.37 / 63.08 ยฑ5.35 / 73.40 ms โ 1.05x slower โ
โ QQuery 15 โ 57.69 / 58.27 ยฑ0.34 / 58.69 ms โ 59.08 / 59.40 ยฑ0.26 / 59.81 ms โ no change โ
โ QQuery 16 โ 11.41 / 11.58 ยฑ0.09 / 11.67 ms โ 11.83 / 11.92 ยฑ0.06 / 11.98 ms โ no change โ
โ QQuery 17 โ 145.77 / 147.05 ยฑ0.81 / 148.16 ms โ 147.91 / 152.74 ยฑ7.95 / 168.60 ms โ no change โ
โ QQuery 18 โ 111.31 / 112.11 ยฑ0.54 / 112.81 ms โ 112.44 / 112.94 ยฑ0.32 / 113.36 ms โ no change โ
โ QQuery 19 โ 289.62 / 295.00 ยฑ4.97 / 303.07 ms โ 292.38 / 302.20 ยฑ7.69 / 311.75 ms โ no change โ
โ QQuery 20 โ 1262.67 / 1274.66 ยฑ14.71 / 1301.61 ms โ 1244.53 / 1250.58 ยฑ4.67 / 1256.28 ms โ no change โ
โ QQuery 21 โ 94.18 / 98.10 ยฑ6.23 / 110.50 ms โ 89.32 / 96.57 ยฑ8.72 / 113.56 ms โ no change โ
โ QQuery 22 โ 100.32 / 103.06 ยฑ1.51 / 104.43 ms โ 100.72 / 101.92 ยฑ0.96 / 103.60 ms โ no change โ
โ QQuery 23 โ 108.67 / 110.06 ยฑ1.41 / 112.62 ms โ 104.59 / 107.16 ยฑ2.77 / 112.45 ms โ no change โ
โ QQuery 24 โ 27.35 / 27.44 ยฑ0.09 / 27.58 ms โ 27.95 / 28.07 ยฑ0.07 / 28.14 ms โ no change โ
โ QQuery 25 โ 70.79 / 73.17 ยฑ3.14 / 79.35 ms โ 70.70 / 73.43 ยฑ1.68 / 75.66 ms โ no change โ
โ QQuery 26 โ 100.92 / 109.92 ยฑ4.83 / 115.57 ms โ 109.89 / 113.36 ยฑ2.40 / 116.18 ms โ no change โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 3671.48ms โ
โ Total Time (refactor-materializing-smj) โ 3659.97ms โ
โ Average Time (HEAD) โ 141.21ms โ
โ Average Time (refactor-materializing-smj) โ 140.77ms โ
โ Queries Faster โ 0 โ
โ Queries Slower โ 1 โ
โ Queries with No Change โ 25 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**smj โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 20.0s |
| Peak memory | 1.2 GiB |
| Avg memory | 460.1 MiB |
| CPU user | 157.3s |
| CPU sys | 3.3s |
| Peak spill | 0 B |
**smj โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 20.0s |
| Peak memory | 1.2 GiB |
| Avg memory | 454.1 MiB |
| CPU user | 156.1s |
| CPU sys | 3.6s |
| 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5119690693
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5119660671)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5119660671-1290-2d6h6 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (ccdd36340b13836fbae8289748495bdad236d042) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..ccdd36340b13836fbae8289748495bdad236d042) 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5132675123
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5132631045)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5132631045-1301-xr9hs 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (5bab8f5cacb00de1e8c507a809093de0ada1939a) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..5bab8f5cacb00de1e8c507a809093de0ada1939a) using: smj
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5132866635
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5132649709)
**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 refactor-materializing-smj
--------------------
Benchmark tpcds_sf1.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 5.60 / 6.12 ยฑ0.90 / 7.91 ms โ 5.50 / 6.09 ยฑ0.92 / 7.89 ms โ no change โ
โ QQuery 2 โ 81.56 / 81.82 ยฑ0.20 / 82.11 ms โ 81.52 / 82.06 ยฑ0.42 / 82.54 ms โ no change โ
โ QQuery 3 โ 29.80 / 30.08 ยฑ0.19 / 30.28 ms โ 29.82 / 30.14 ยฑ0.21 / 30.44 ms โ no change โ
โ QQuery 4 โ 496.11 / 499.65 ยฑ2.07 / 501.86 ms โ 494.59 / 500.83 ยฑ4.93 / 507.93 ms โ no change โ
โ QQuery 5 โ 53.14 / 53.59 ยฑ0.41 / 54.34 ms โ 52.29 / 52.77 ยฑ0.26 / 53.09 ms โ no change โ
โ QQuery 6 โ 36.66 / 37.16 ยฑ0.28 / 37.43 ms โ 36.57 / 36.84 ยฑ0.21 / 37.14 ms โ no change โ
โ QQuery 7 โ 95.26 / 97.85 ยฑ2.58 / 102.80 ms โ 95.92 / 97.68 ยฑ2.10 / 101.57 ms โ no change โ
โ QQuery 8 โ 37.26 / 37.80 ยฑ0.60 / 38.96 ms โ 37.20 / 37.61 ยฑ0.25 / 37.91 ms โ no change โ
โ QQuery 9 โ 54.75 / 56.58 ยฑ2.38 / 61.22 ms โ 52.40 / 55.17 ยฑ1.66 / 57.35 ms โ no change โ
โ QQuery 10 โ 63.87 / 64.58 ยฑ0.38 / 64.91 ms โ 63.22 / 63.72 ยฑ0.26 / 63.91 ms โ no change โ
โ QQuery 11 โ 309.84 / 312.33 ยฑ2.30 / 315.56 ms โ 306.55 / 312.37 ยฑ5.53 / 319.30 ms โ no change โ
โ QQuery 12 โ 28.77 / 29.18 ยฑ0.26 / 29.55 ms โ 28.88 / 29.62 ยฑ0.68 / 30.47 ms โ no change โ
โ QQuery 13 โ 119.96 / 121.70 ยฑ1.64 / 124.29 ms โ 119.72 / 121.91 ยฑ3.36 / 128.58 ms โ no change โ
โ QQuery 14 โ 418.39 / 423.35 ยฑ3.61 / 428.98 ms โ 415.75 / 419.43 ยฑ3.96 / 426.62 ms โ no change โ
โ QQuery 15 โ 57.62 / 59.15 ยฑ1.22 / 61.30 ms โ 58.73 / 59.92 ยฑ0.68 / 60.61 ms โ no change โ
โ QQuery 16 โ 6.84 / 6.99 ยฑ0.18 / 7.34 ms โ 6.70 / 6.86 ยฑ0.24 / 7.33 ms โ no change โ
โ QQuery 17 โ 79.60 / 81.54 ยฑ2.13 / 85.65 ms โ 80.68 / 81.88 ยฑ1.09 / 83.66 ms โ no change โ
โ QQuery 18 โ 123.05 / 125.37 ยฑ1.80 / 128.07 ms โ 123.75 / 125.91 ยฑ1.16 / 127.16 ms โ no change โ
โ QQuery 19 โ 42.17 / 42.58 ยฑ0.36 / 43.14 ms โ 42.48 / 42.75 ยฑ0.17 / 42.96 ms โ no change โ
โ QQuery 20 โ 36.76 / 37.91 ยฑ1.24 / 39.68 ms โ 35.88 / 37.84 ยฑ1.49 / 40.18 ms โ no change โ
โ QQuery 21 โ 17.87 / 18.22 ยฑ0.25 / 18.54 ms โ 17.74 / 18.21 ยฑ0.46 / 19.01 ms โ no change โ
โ QQuery 22 โ 63.10 / 64.19 ยฑ0.76 / 65.12 ms โ 63.80 / 64.31 ยฑ0.62 / 65.52 ms โ no change โ
โ QQuery 23 โ 345.04 / 359.96 ยฑ16.57 / 391.48 ms โ 347.63 / 352.58 ยฑ5.77 / 363.65 ms โ no change โ
โ QQuery 24 โ 226.14 / 229.08 ยฑ2.24 / 231.79 ms โ 226.67 / 230.16 ยฑ5.07 / 240.19 ms โ no change โ
โ QQuery 25 โ 110.36 / 111.55 ยฑ1.24 / 113.85 ms โ 111.57 / 112.41 ยฑ1.32 / 115.00 ms โ no change โ
โ QQuery 26 โ 58.72 / 59.12 ยฑ0.30 / 59.58 ms โ 58.26 / 61.26 ยฑ3.41 / 67.38 ms โ no change โ
โ QQuery 27 โ 6.45 / 6.59 ยฑ0.17 / 6.91 ms โ 6.38 / 6.51 ยฑ0.13 / 6.73 ms โ no change โ
โ QQuery 28 โ 57.94 / 61.11 ยฑ2.41 / 64.00 ms โ 58.07 / 61.64 ยฑ1.93 / 63.96 ms โ no change โ
โ QQuery 29 โ 97.81 / 99.16 ยฑ1.79 / 102.61 ms โ 98.20 / 99.71 ยฑ1.66 / 102.80 ms โ no change โ
โ QQuery 30 โ 32.65 / 33.01 ยฑ0.37 / 33.61 ms โ 33.75 / 35.39 ยฑ2.01 / 39.26 ms โ 1.07x slower โ
โ QQuery 31 โ 112.25 / 112.53 ยฑ0.23 / 112.89 ms โ 112.56 / 113.38 ยฑ0.90 / 115.11 ms โ no change โ
โ QQuery 32 โ 20.57 / 20.80 ยฑ0.17 / 21.02 ms โ 20.65 / 21.04 ยฑ0.23 / 21.36 ms โ no change โ
โ QQuery 33 โ 38.46 / 38.66 ยฑ0.27 / 39.19 ms โ 38.58 / 40.01 ยฑ1.35 / 42.26 ms โ no change โ
โ QQuery 34 โ 9.83 / 10.14 ยฑ0.26 / 10.45 ms โ 9.95 / 10.27 ยฑ0.20 / 10.44 ms โ no change โ
โ QQuery 35 โ 73.53 / 73.84 ยฑ0.33 / 74.41 ms โ 73.33 / 74.13 ยฑ0.41 / 74.43 ms โ no change โ
โ QQuery 36 โ 5.85 / 5.97 ยฑ0.18 / 6.32 ms โ 5.82 / 5.95 ยฑ0.18 / 6.30 ms โ no change โ
โ QQuery 37 โ 6.92 / 7.00 ยฑ0.07 / 7.10 ms โ 6.96 / 7.02 ยฑ0.05 / 7.09 ms โ no change โ
โ QQuery 38 โ 62.88 / 63.00 ยฑ0.15 / 63.27 ms โ 62.02 / 62.30 ยฑ0.24 / 62.58 ms โ no change โ
โ QQuery 39 โ 91.69 / 92.69 ยฑ1.32 / 95.27 ms โ 90.76 / 92.22 ยฑ1.70 / 95.47 ms โ no change โ
โ QQuery 40 โ 23.60 / 23.99 ยฑ0.37 / 24.47 ms โ 23.79 / 24.07 ยฑ0.22 / 24.37 ms โ no change โ
โ QQuery 41 โ 11.68 / 11.89 ยฑ0.24 / 12.34 ms โ 11.72 / 12.09 ยฑ0.30 / 12.58 ms โ no change โ
โ QQuery 42 โ 24.35 / 24.69 ยฑ0.36 / 25.37 ms โ 24.37 / 24.88 ยฑ0.47 / 25.57 ms โ no change โ
โ QQuery 43 โ 5.10 / 5.23 ยฑ0.19 / 5.60 ms โ 5.16 / 5.27 ยฑ0.20 / 5.67 ms โ no change โ
โ QQuery 44 โ 9.48 / 9.54 ยฑ0.06 / 9.64 ms โ 9.56 / 9.66 ยฑ0.09 / 9.82 ms โ no change โ
โ QQuery 45 โ 38.31 / 39.32 ยฑ1.43 / 42.13 ms โ 38.57 / 40.34 ยฑ2.04 / 44.22 ms โ no change โ
โ QQuery 46 โ 11.88 / 12.23 ยฑ0.44 / 13.10 ms โ 12.19 / 12.83 ยฑ0.46 / 13.38 ms โ no change โ
โ QQuery 47 โ 231.15 / 235.85 ยฑ4.81 / 244.52 ms โ 234.12 / 237.48 ยฑ2.20 / 239.92 ms โ no change โ
โ QQuery 48 โ 97.40 / 98.15 ยฑ0.54 / 99.08 ms โ 98.01 / 98.48 ยฑ0.25 / 98.71 ms โ no change โ
โ QQuery 49 โ 77.16 / 78.02 ยฑ1.65 / 81.32 ms โ 78.15 / 80.51 ยฑ3.30 / 86.91 ms โ no change โ
โ QQuery 50 โ 59.77 / 60.68 ยฑ0.69 / 61.82 ms โ 60.57 / 60.97 ยฑ0.48 / 61.88 ms โ no change โ
โ QQuery 51 โ 93.03 / 93.90 ยฑ0.72 / 94.98 ms โ 93.92 / 94.47 ยฑ0.38 / 94.93 ms โ no change โ
โ QQuery 52 โ 24.80 / 25.02 ยฑ0.22 / 25.38 ms โ 24.97 / 27.04 ยฑ3.42 / 33.82 ms โ 1.08x slower โ
โ QQuery 53 โ 30.57 / 32.18 ยฑ2.80 / 37.76 ms โ 30.71 / 31.39 ยฑ0.51 / 32.03 ms โ no change โ
โ QQuery 54 โ 56.80 / 57.26 ยฑ0.38 / 57.89 ms โ 57.00 / 57.50 ยฑ0.41 / 57.93 ms โ no change โ
โ QQuery 55 โ 23.92 / 24.55 ยฑ0.79 / 26.10 ms โ 24.47 / 24.59 ยฑ0.14 / 24.79 ms โ no change โ
โ QQuery 56 โ 39.35 / 39.99 ยฑ0.53 / 40.74 ms โ 39.72 / 40.24 ยฑ0.34 / 40.74 ms โ no change โ
โ QQuery 57 โ 176.89 / 179.16 ยฑ2.20 / 183.34 ms โ 179.21 / 182.31 ยฑ3.52 / 189.18 ms โ no change โ
โ QQuery 58 โ 116.43 / 118.71 ยฑ3.23 / 125.06 ms โ 117.45 / 119.88 ยฑ2.26 / 123.56 ms โ no change โ
โ QQuery 59 โ 118.77 / 119.26 ยฑ0.38 / 119.86 ms โ 120.48 / 120.87 ยฑ0.29 / 121.31 ms โ no change โ
โ QQuery 60 โ 40.07 / 40.58 ยฑ0.27 / 40.87 ms โ 40.70 / 42.16 ยฑ1.87 / 45.78 ms โ no change โ
โ QQuery 61 โ 12.66 / 13.32 ยฑ1.20 / 15.73 ms โ 12.26 / 12.46 ยฑ0.18 / 12.77 ms โ +1.07x faster โ
โ QQuery 62 โ 47.17 / 47.60 ยฑ0.33 / 48.07 ms โ 47.37 / 47.78 ยฑ0.29 / 48.15 ms โ no change โ
โ QQuery 63 โ 30.48 / 30.68 ยฑ0.24 / 31.00 ms โ 30.54 / 31.03 ยฑ0.28 / 31.38 ms โ no change โ
โ QQuery 64 โ 407.58 / 410.76 ยฑ2.36 / 413.44 ms โ 416.24 / 422.95 ยฑ6.60 / 432.69 ms โ no change โ
โ QQuery 65 โ 127.76 / 130.35 ยฑ2.41 / 133.66 ms โ 128.24 / 131.95 ยฑ2.47 / 135.60 ms โ no change โ
โ QQuery 66 โ 80.38 / 81.43 ยฑ0.83 / 82.80 ms โ 82.86 / 83.02 ยฑ0.13 / 83.22 ms โ no change โ
โ QQuery 67 โ 244.27 / 247.12 ยฑ3.18 / 251.02 ms โ 244.37 / 249.10 ยฑ4.22 / 256.90 ms โ no change โ
โ QQuery 68 โ 11.88 / 12.15 ยฑ0.24 / 12.55 ms โ 12.09 / 13.83 ยฑ3.10 / 20.00 ms โ 1.14x slower โ
โ QQuery 69 โ 58.34 / 61.10 ยฑ2.14 / 63.71 ms โ 58.73 / 60.42 ยฑ1.77 / 63.78 ms โ no change โ
โ QQuery 70 โ 106.86 / 108.55 ยฑ2.41 / 113.28 ms โ 107.61 / 109.18 ยฑ1.65 / 111.92 ms โ no change โ
โ QQuery 71 โ 35.66 / 36.40 ยฑ0.56 / 37.33 ms โ 35.71 / 37.92 ยฑ3.83 / 45.58 ms โ no change โ
โ QQuery 72 โ 2148.41 / 2190.76 ยฑ22.54 / 2214.29 ms โ 1984.41 / 2124.42 ยฑ83.77 / 2230.21 ms โ no change โ
โ QQuery 73 โ 9.53 / 9.75 ยฑ0.29 / 10.33 ms โ 9.71 / 9.94 ยฑ0.29 / 10.51 ms โ no change โ
โ QQuery 74 โ 171.57 / 175.52 ยฑ2.60 / 177.96 ms โ 172.89 / 177.96 ยฑ5.56 / 188.42 ms โ no change โ
โ QQuery 75 โ 149.40 / 154.04 ยฑ6.87 / 167.64 ms โ 151.37 / 152.64 ยฑ0.92 / 154.11 ms โ no change โ
โ QQuery 76 โ 35.82 / 36.10 ยฑ0.37 / 36.81 ms โ 36.06 / 36.45 ยฑ0.42 / 37.15 ms โ no change โ
โ QQuery 77 โ 62.34 / 63.51 ยฑ1.68 / 66.83 ms โ 62.21 / 62.43 ยฑ0.17 / 62.68 ms โ no change โ
โ QQuery 78 โ 198.51 / 200.52 ยฑ2.01 / 204.33 ms โ 201.98 / 203.83 ยฑ2.45 / 208.56 ms โ no change โ
โ QQuery 79 โ 68.09 / 69.50 ยฑ1.09 / 71.46 ms โ 68.54 / 71.18 ยฑ4.26 / 79.67 ms โ no change โ
โ QQuery 80 โ 98.92 / 100.58 ยฑ1.32 / 102.49 ms โ 101.78 / 103.92 ยฑ1.42 / 106.05 ms โ no change โ
โ QQuery 81 โ 26.06 / 27.66 ยฑ2.56 / 32.76 ms โ 26.17 / 26.40 ยฑ0.13 / 26.54 ms โ no change โ
โ QQuery 82 โ 17.28 / 19.04 ยฑ3.12 / 25.27 ms โ 16.48 / 17.38 ยฑ1.01 / 18.83 ms โ +1.10x faster โ
โ QQuery 83 โ 40.57 / 41.47 ยฑ0.63 / 42.35 ms โ 40.86 / 43.50 ยฑ3.52 / 50.48 ms โ no change โ
โ QQuery 84 โ 30.84 / 31.16 ยฑ0.23 / 31.53 ms โ 30.65 / 31.16 ยฑ0.49 / 31.92 ms โ no change โ
โ QQuery 85 โ 107.24 / 109.25 ยฑ1.53 / 111.90 ms โ 107.13 / 108.67 ยฑ1.91 / 112.16 ms โ no change โ
โ QQuery 86 โ 26.33 / 28.29 ยฑ2.97 / 34.19 ms โ 25.33 / 25.95 ยฑ0.41 / 26.56 ms โ +1.09x faster โ
โ QQuery 87 โ 63.15 / 64.22 ยฑ0.86 / 65.61 ms โ 63.38 / 64.90 ยฑ2.16 / 69.17 ms โ no change โ
โ QQuery 88 โ 64.80 / 65.09 ยฑ0.28 / 65.48 ms โ 63.90 / 64.86 ยฑ1.08 / 66.95 ms โ no change โ
โ QQuery 89 โ 36.46 / 36.55 ยฑ0.08 / 36.69 ms โ 36.48 / 36.67 ยฑ0.12 / 36.82 ms โ no change โ
โ QQuery 90 โ 17.44 / 18.52 ยฑ1.93 / 22.38 ms โ 17.20 / 17.41 ยฑ0.16 / 17.63 ms โ +1.06x faster โ
โ QQuery 91 โ 46.70 / 47.42 ยฑ0.51 / 47.99 ms โ 45.72 / 47.31 ยฑ2.22 / 51.72 ms โ no change โ
โ QQuery 92 โ 30.02 / 30.76 ยฑ0.44 / 31.37 ms โ 29.90 / 30.60 ยฑ0.36 / 30.91 ms โ no change โ
โ QQuery 93 โ 50.61 / 50.95 ยฑ0.26 / 51.27 ms โ 50.53 / 51.23 ยฑ0.50 / 52.00 ms โ no change โ
โ QQuery 94 โ 38.68 / 39.06 ยฑ0.32 / 39.51 ms โ 38.39 / 38.81 ยฑ0.34 / 39.22 ms โ no change โ
โ QQuery 95 โ 81.31 / 82.62 ยฑ0.96 / 83.64 ms โ 80.41 / 82.66 ยฑ2.12 / 86.42 ms โ no change โ
โ QQuery 96 โ 24.37 / 24.58 ยฑ0.15 / 24.81 ms โ 24.49 / 24.70 ยฑ0.19 / 24.96 ms โ no change โ
โ QQuery 97 โ 46.91 / 47.25 ยฑ0.22 / 47.58 ms โ 47.24 / 47.73 ยฑ0.67 / 49.06 ms โ no change โ
โ QQuery 98 โ 43.48 / 43.99 ยฑ0.47 / 44.82 ms โ 43.35 / 43.84 ยฑ0.58 / 44.91 ms โ no change โ
โ QQuery 99 โ 71.14 / 72.66 ยฑ1.57 / 75.53 ms โ 70.90 / 72.91 ยฑ2.57 / 77.39 ms โ no change โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 10062.44ms โ
โ Total Time (refactor-materializing-smj) โ 10036.05ms โ
โ Average Time (HEAD) โ 101.64ms โ
โ Average Time (refactor-materializing-smj) โ 101.37ms โ
โ Queries Faster โ 4 โ
โ Queries Slower โ 3 โ
โ Queries with No Change โ 92 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 55.0s |
| Peak memory | 2.4 GiB |
| Avg memory | 1.6 GiB |
| CPU user | 226.7s |
| CPU sys | 6.2s |
| Peak spill | 0 B |
**tpcds โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 55.0s |
| Peak memory | 2.3 GiB |
| Avg memory | 1.5 GiB |
| CPU user | 225.6s |
| CPU sys | 6.0s |
| 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] chore: refactor `MaterializingSortMergeJoinStream` into generators and simplify code to be textbook like as possible [datafusion]
Posted by "rluvaton (via GitHub)" <gi...@apache.org>.
rluvaton commented on code in PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#discussion_r3686530474
##########
datafusion/physical-plan/src/joins/sort_merge_join/materializing_stream.rs:
##########
@@ -995,15 +984,12 @@ impl MaterializingSortMergeJoinStream {
let bb = &mut self.buffered_data.batches[idx];
if let BufferedBatchState::Spilled(spill_file) = &bb.batch {
- if self.spill_stream.is_none() {
- let stream = self
- .spill_manager
- .read_spill_as_stream(Arc::clone(spill_file), None)?;
- self.spill_stream = Some(stream);
- }
+ let mut spill_stream = self
+ .spill_manager
+ .read_spill_as_stream(Arc::clone(spill_file), None)?;
- match ready!(self.spill_stream.as_mut().unwrap().poll_next_unpin(cx)) {
- Some(Ok(batch)) => {
+ match spill_stream.next().await.transpose()? {
Review Comment:
done
--
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5119905842
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5119659315)
**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 refactor-materializing-smj
--------------------
Benchmark smj.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 8.52 / 8.72 ยฑ0.14 / 8.92 ms โ 8.00 / 8.21 ยฑ0.19 / 8.45 ms โ +1.06x faster โ
โ QQuery 2 โ 163.16 / 164.09 ยฑ0.80 / 165.22 ms โ 158.54 / 159.63 ยฑ1.12 / 161.78 ms โ no change โ
โ QQuery 3 โ 105.44 / 106.72 ยฑ0.76 / 107.66 ms โ 102.00 / 103.87 ยฑ1.26 / 105.97 ms โ no change โ
โ QQuery 4 โ 28.36 / 28.52 ยฑ0.15 / 28.75 ms โ 28.20 / 28.33 ยฑ0.12 / 28.51 ms โ no change โ
โ QQuery 5 โ 20.79 / 20.97 ยฑ0.18 / 21.29 ms โ 20.60 / 20.68 ยฑ0.09 / 20.86 ms โ no change โ
โ QQuery 6 โ 160.06 / 163.18 ยฑ3.53 / 170.04 ms โ 155.67 / 157.04 ยฑ1.40 / 159.42 ms โ no change โ
โ QQuery 7 โ 198.64 / 201.97 ยฑ1.74 / 203.75 ms โ 193.45 / 197.79 ยฑ5.44 / 208.31 ms โ no change โ
โ QQuery 8 โ 19.05 / 19.25 ยฑ0.17 / 19.52 ms โ 18.75 / 19.78 ยฑ1.05 / 21.46 ms โ no change โ
โ QQuery 9 โ 207.12 / 214.96 ยฑ10.12 / 234.42 ms โ 202.09 / 203.86 ยฑ1.15 / 205.55 ms โ +1.05x faster โ
โ QQuery 10 โ 58.05 / 58.33 ยฑ0.19 / 58.64 ms โ 57.95 / 58.40 ยฑ0.23 / 58.56 ms โ no change โ
โ QQuery 11 โ 27.10 / 30.49 ยฑ6.31 / 43.10 ms โ 27.36 / 27.51 ยฑ0.15 / 27.78 ms โ +1.11x faster โ
โ QQuery 12 โ 70.76 / 72.19 ยฑ1.62 / 75.27 ms โ 67.13 / 71.77 ยฑ2.56 / 74.92 ms โ no change โ
โ QQuery 13 โ 96.69 / 103.83 ยฑ4.88 / 108.23 ms โ 95.39 / 103.18 ยฑ6.59 / 111.93 ms โ no change โ
โ QQuery 14 โ 58.48 / 61.55 ยฑ4.42 / 70.11 ms โ 57.75 / 59.64 ยฑ2.40 / 64.35 ms โ no change โ
โ QQuery 15 โ 58.00 / 58.14 ยฑ0.14 / 58.39 ms โ 57.94 / 58.20 ยฑ0.24 / 58.62 ms โ no change โ
โ QQuery 16 โ 11.55 / 11.83 ยฑ0.19 / 12.14 ms โ 11.55 / 11.72 ยฑ0.12 / 11.84 ms โ no change โ
โ QQuery 17 โ 145.96 / 151.28 ยฑ8.47 / 168.15 ms โ 145.70 / 146.69 ยฑ1.67 / 150.01 ms โ no change โ
โ QQuery 18 โ 111.04 / 111.46 ยฑ0.35 / 111.98 ms โ 111.97 / 112.33 ยฑ0.25 / 112.73 ms โ no change โ
โ QQuery 19 โ 290.20 / 296.30 ยฑ6.29 / 308.04 ms โ 290.65 / 299.33 ยฑ7.38 / 308.95 ms โ no change โ
โ QQuery 20 โ 1262.70 / 1267.90 ยฑ5.40 / 1274.78 ms โ 1236.53 / 1244.40 ยฑ5.39 / 1251.81 ms โ no change โ
โ QQuery 21 โ 94.46 / 95.84 ยฑ0.94 / 97.08 ms โ 88.86 / 89.73 ยฑ0.65 / 90.64 ms โ +1.07x faster โ
โ QQuery 22 โ 102.67 / 103.61 ยฑ0.78 / 105.00 ms โ 96.32 / 99.06 ยฑ3.98 / 106.82 ms โ no change โ
โ QQuery 23 โ 110.90 / 112.48 ยฑ1.39 / 114.29 ms โ 104.36 / 105.48 ยฑ0.72 / 106.57 ms โ +1.07x faster โ
โ QQuery 24 โ 27.46 / 27.54 ยฑ0.08 / 27.66 ms โ 27.41 / 27.84 ยฑ0.42 / 28.63 ms โ no change โ
โ QQuery 25 โ 65.74 / 71.26 ยฑ3.85 / 77.16 ms โ 69.69 / 73.68 ยฑ3.53 / 80.15 ms โ no change โ
โ QQuery 26 โ 98.03 / 107.04 ยฑ7.11 / 118.25 ms โ 98.33 / 109.05 ยฑ5.52 / 113.61 ms โ no change โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 3669.47ms โ
โ Total Time (refactor-materializing-smj) โ 3597.21ms โ
โ Average Time (HEAD) โ 141.13ms โ
โ Average Time (refactor-materializing-smj) โ 138.35ms โ
โ Queries Faster โ 5 โ
โ Queries Slower โ 0 โ
โ Queries with No Change โ 21 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**smj โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 20.0s |
| Peak memory | 1.1 GiB |
| Avg memory | 445.0 MiB |
| CPU user | 156.9s |
| CPU sys | 3.5s |
| Peak spill | 0 B |
**smj โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 20.0s |
| Peak memory | 1.1 GiB |
| Avg memory | 441.0 MiB |
| CPU user | 157.2s |
| CPU sys | 3.5s |
| 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] chore: refactor `MaterializingSortMergeJoinStream` into generators and simplify code to be textbook like as possible [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5133692079
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5133512422)
**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 refactor-materializing-smj
--------------------
Benchmark smj.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 8.37 / 8.78 ยฑ0.27 / 9.19 ms โ 8.65 / 8.81 ยฑ0.22 / 9.24 ms โ no change โ
โ QQuery 2 โ 162.12 / 164.55 ยฑ1.93 / 167.19 ms โ 158.47 / 160.95 ยฑ2.06 / 164.32 ms โ no change โ
โ QQuery 3 โ 102.55 / 105.30 ยฑ1.44 / 106.30 ms โ 103.17 / 104.74 ยฑ1.03 / 106.24 ms โ no change โ
โ QQuery 4 โ 28.08 / 28.35 ยฑ0.18 / 28.66 ms โ 28.20 / 28.52 ยฑ0.19 / 28.76 ms โ no change โ
โ QQuery 5 โ 20.75 / 21.12 ยฑ0.21 / 21.35 ms โ 20.41 / 20.59 ยฑ0.17 / 20.91 ms โ no change โ
โ QQuery 6 โ 159.14 / 164.41 ยฑ5.76 / 175.18 ms โ 156.46 / 157.60 ยฑ1.01 / 159.42 ms โ no change โ
โ QQuery 7 โ 198.07 / 200.74 ยฑ1.84 / 202.48 ms โ 195.58 / 198.12 ยฑ1.81 / 200.44 ms โ no change โ
โ QQuery 8 โ 18.96 / 19.17 ยฑ0.16 / 19.34 ms โ 18.43 / 18.67 ยฑ0.16 / 18.86 ms โ no change โ
โ QQuery 9 โ 206.62 / 213.19 ยฑ11.50 / 236.12 ms โ 202.32 / 203.59 ยฑ1.50 / 206.44 ms โ no change โ
โ QQuery 10 โ 57.75 / 57.96 ยฑ0.25 / 58.42 ms โ 58.10 / 58.42 ยฑ0.33 / 59.05 ms โ no change โ
โ QQuery 11 โ 27.13 / 31.70 ยฑ9.02 / 49.74 ms โ 27.31 / 27.40 ยฑ0.06 / 27.50 ms โ +1.16x faster โ
โ QQuery 12 โ 64.81 / 70.00 ยฑ3.29 / 74.21 ms โ 66.48 / 70.23 ยฑ2.26 / 73.50 ms โ no change โ
โ QQuery 13 โ 95.63 / 104.60 ยฑ6.78 / 113.43 ms โ 103.15 / 104.56 ยฑ1.01 / 106.20 ms โ no change โ
โ QQuery 14 โ 58.25 / 61.93 ยฑ6.05 / 73.94 ms โ 58.12 / 61.33 ยฑ4.95 / 71.11 ms โ no change โ
โ QQuery 15 โ 57.86 / 58.54 ยฑ0.55 / 59.26 ms โ 58.48 / 58.65 ยฑ0.11 / 58.76 ms โ no change โ
โ QQuery 16 โ 11.50 / 11.63 ยฑ0.10 / 11.79 ms โ 11.45 / 11.65 ยฑ0.11 / 11.76 ms โ no change โ
โ QQuery 17 โ 146.42 / 151.27 ยฑ8.59 / 168.43 ms โ 146.00 / 149.50 ยฑ6.23 / 161.94 ms โ no change โ
โ QQuery 18 โ 111.50 / 111.70 ยฑ0.19 / 112.02 ms โ 112.37 / 113.11 ยฑ0.59 / 113.75 ms โ no change โ
โ QQuery 19 โ 288.35 / 296.50 ยฑ8.25 / 311.51 ms โ 292.84 / 309.56 ยฑ9.65 / 318.62 ms โ no change โ
โ QQuery 20 โ 1258.96 / 1266.00 ยฑ7.98 / 1281.22 ms โ 1240.29 / 1247.36 ยฑ6.44 / 1255.92 ms โ no change โ
โ QQuery 21 โ 94.92 / 96.10 ยฑ1.37 / 98.56 ms โ 90.92 / 91.80 ยฑ0.58 / 92.54 ms โ no change โ
โ QQuery 22 โ 98.82 / 102.74 ยฑ2.24 / 105.39 ms โ 98.09 / 100.39 ยฑ1.47 / 101.69 ms โ no change โ
โ QQuery 23 โ 107.14 / 110.51 ยฑ2.08 / 113.63 ms โ 103.01 / 105.08 ยฑ1.21 / 106.70 ms โ no change โ
โ QQuery 24 โ 27.33 / 27.49 ยฑ0.08 / 27.57 ms โ 27.78 / 27.97 ยฑ0.16 / 28.15 ms โ no change โ
โ QQuery 25 โ 63.53 / 67.59 ยฑ3.36 / 71.09 ms โ 69.38 / 72.32 ยฑ4.20 / 80.64 ms โ 1.07x slower โ
โ QQuery 26 โ 96.58 / 108.68 ยฑ7.70 / 119.06 ms โ 94.61 / 112.70 ยฑ17.19 / 145.44 ms โ no change โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 3660.55ms โ
โ Total Time (refactor-materializing-smj) โ 3623.60ms โ
โ Average Time (HEAD) โ 140.79ms โ
โ Average Time (refactor-materializing-smj) โ 139.37ms โ
โ Queries Faster โ 1 โ
โ Queries Slower โ 1 โ
โ Queries with No Change โ 24 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**smj โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 20.0s |
| Peak memory | 1.1 GiB |
| Avg memory | 449.4 MiB |
| CPU user | 157.0s |
| CPU sys | 3.6s |
| Peak spill | 0 B |
**smj โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 20.0s |
| Peak memory | 1.1 GiB |
| Avg memory | 450.5 MiB |
| CPU user | 155.9s |
| CPU sys | 3.6s |
| 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5132899405
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5132649709)
**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 refactor-materializing-smj
--------------------
Benchmark clickbench_partitioned.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 0 โ 1.28 / 4.05 ยฑ5.45 / 14.95 ms โ 1.28 / 4.05 ยฑ5.47 / 14.98 ms โ no change โ
โ QQuery 1 โ 13.49 / 13.68 ยฑ0.22 / 14.09 ms โ 13.21 / 13.61 ยฑ0.21 / 13.80 ms โ no change โ
โ QQuery 2 โ 36.71 / 36.96 ยฑ0.27 / 37.40 ms โ 36.31 / 36.55 ยฑ0.17 / 36.84 ms โ no change โ
โ QQuery 3 โ 31.25 / 32.01 ยฑ0.63 / 33.17 ms โ 30.77 / 31.16 ยฑ0.31 / 31.63 ms โ no change โ
โ QQuery 4 โ 224.68 / 227.21 ยฑ3.35 / 233.73 ms โ 223.02 / 225.75 ยฑ2.98 / 231.18 ms โ no change โ
โ QQuery 5 โ 271.23 / 274.16 ยฑ1.58 / 275.57 ms โ 273.01 / 274.05 ยฑ1.20 / 276.27 ms โ no change โ
โ QQuery 6 โ 1.30 / 1.45 ยฑ0.23 / 1.90 ms โ 1.32 / 1.51 ยฑ0.30 / 2.11 ms โ no change โ
โ QQuery 7 โ 14.42 / 14.46 ยฑ0.03 / 14.51 ms โ 14.52 / 14.65 ยฑ0.07 / 14.75 ms โ no change โ
โ QQuery 8 โ 321.40 / 327.76 ยฑ3.97 / 333.55 ms โ 322.82 / 326.39 ยฑ2.10 / 328.17 ms โ no change โ
โ QQuery 9 โ 453.55 / 460.03 ยฑ5.96 / 467.49 ms โ 453.85 / 461.66 ยฑ7.59 / 474.75 ms โ no change โ
โ QQuery 10 โ 69.82 / 70.69 ยฑ0.83 / 71.72 ms โ 69.75 / 70.67 ยฑ0.77 / 71.73 ms โ no change โ
โ QQuery 11 โ 81.47 / 82.19 ยฑ0.46 / 82.87 ms โ 82.02 / 83.11 ยฑ0.96 / 84.81 ms โ no change โ
โ QQuery 12 โ 263.56 / 268.19 ยฑ4.32 / 275.12 ms โ 268.01 / 271.93 ยฑ4.10 / 278.53 ms โ no change โ
โ QQuery 13 โ 362.10 / 368.88 ยฑ5.26 / 375.62 ms โ 372.95 / 386.95 ยฑ11.96 / 408.80 ms โ no change โ
โ QQuery 14 โ 282.92 / 287.59 ยฑ5.53 / 298.02 ms โ 283.73 / 285.07 ยฑ0.98 / 285.99 ms โ no change โ
โ QQuery 15 โ 265.14 / 277.39 ยฑ7.05 / 285.64 ms โ 267.25 / 281.30 ยฑ9.69 / 297.18 ms โ no change โ
โ QQuery 16 โ 611.43 / 619.09 ยฑ5.86 / 628.07 ms โ 608.25 / 619.72 ยฑ6.53 / 626.43 ms โ no change โ
โ QQuery 17 โ 613.93 / 622.20 ยฑ6.59 / 629.33 ms โ 619.03 / 626.82 ยฑ5.18 / 632.30 ms โ no change โ
โ QQuery 18 โ 1249.80 / 1266.69 ยฑ15.90 / 1291.56 ms โ 1240.61 / 1274.58 ยฑ17.90 / 1292.19 ms โ no change โ
โ QQuery 19 โ 28.62 / 31.88 ยฑ5.93 / 43.73 ms โ 28.75 / 39.45 ยฑ15.90 / 69.82 ms โ 1.24x slower โ
โ QQuery 20 โ 517.08 / 520.67 ยฑ6.20 / 533.03 ms โ 520.44 / 529.72 ยฑ5.13 / 535.01 ms โ no change โ
โ QQuery 21 โ 514.37 / 522.37 ยฑ6.25 / 533.09 ms โ 512.62 / 518.56 ยฑ3.54 / 522.54 ms โ no change โ
โ QQuery 22 โ 981.94 / 993.03 ยฑ11.08 / 1014.28 ms โ 988.46 / 996.86 ยฑ5.92 / 1005.34 ms โ no change โ
โ QQuery 23 โ 3010.96 / 3066.67 ยฑ29.29 / 3094.84 ms โ 3021.81 / 3072.66 ยฑ31.96 / 3106.05 ms โ no change โ
โ QQuery 24 โ 41.33 / 41.99 ยฑ0.67 / 43.14 ms โ 42.48 / 50.70 ยฑ14.37 / 79.36 ms โ 1.21x slower โ
โ QQuery 25 โ 112.35 / 114.13 ยฑ2.01 / 117.95 ms โ 111.86 / 115.52 ยฑ6.22 / 127.91 ms โ no change โ
โ QQuery 26 โ 41.95 / 44.28 ยฑ3.95 / 52.16 ms โ 41.94 / 46.14 ยฑ7.81 / 61.76 ms โ no change โ
โ QQuery 27 โ 510.07 / 516.23 ยฑ3.23 / 518.79 ms โ 511.59 / 526.15 ยฑ7.91 / 533.40 ms โ no change โ
โ QQuery 28 โ 2866.58 / 2909.48 ยฑ23.60 / 2934.20 ms โ 2875.78 / 2920.68 ยฑ24.27 / 2945.53 ms โ no change โ
โ QQuery 29 โ 41.53 / 51.95 ยฑ20.14 / 92.23 ms โ 41.44 / 51.71 ยฑ14.25 / 78.41 ms โ no change โ
โ QQuery 30 โ 309.44 / 312.45 ยฑ2.27 / 315.80 ms โ 299.90 / 310.35 ยฑ6.39 / 319.43 ms โ no change โ
โ QQuery 31 โ 284.43 / 288.85 ยฑ2.62 / 292.46 ms โ 287.33 / 295.02 ยฑ6.01 / 302.91 ms โ no change โ
โ QQuery 32 โ 925.78 / 966.34 ยฑ23.17 / 997.17 ms โ 924.24 / 954.56 ยฑ19.49 / 975.43 ms โ no change โ
โ QQuery 33 โ 1436.40 / 1465.37 ยฑ19.05 / 1488.07 ms โ 1446.18 / 1483.20 ยฑ23.03 / 1513.43 ms โ no change โ
โ QQuery 34 โ 1464.85 / 1495.14 ยฑ20.14 / 1524.96 ms โ 1451.97 / 1480.90 ยฑ19.08 / 1503.03 ms โ no change โ
โ QQuery 35 โ 280.49 / 314.00 ยฑ40.01 / 370.05 ms โ 279.46 / 317.30 ยฑ36.35 / 378.68 ms โ no change โ
โ QQuery 36 โ 66.88 / 74.66 ยฑ6.56 / 83.52 ms โ 65.59 / 72.63 ยฑ6.10 / 80.13 ms โ no change โ
โ QQuery 37 โ 35.42 / 37.93 ยฑ3.48 / 44.76 ms โ 36.02 / 37.64 ยฑ1.67 / 40.70 ms โ no change โ
โ QQuery 38 โ 40.08 / 44.69 ยฑ3.96 / 49.53 ms โ 43.06 / 45.63 ยฑ2.05 / 48.46 ms โ no change โ
โ QQuery 39 โ 148.86 / 153.47 ยฑ2.72 / 156.24 ms โ 136.01 / 149.49 ยฑ7.73 / 157.19 ms โ no change โ
โ QQuery 40 โ 14.30 / 15.95 ยฑ2.18 / 20.22 ms โ 14.65 / 16.29 ยฑ1.80 / 19.70 ms โ no change โ
โ QQuery 41 โ 14.04 / 14.35 ยฑ0.26 / 14.75 ms โ 14.30 / 15.89 ยฑ1.52 / 18.43 ms โ 1.11x slower โ
โ QQuery 42 โ 13.42 / 15.55 ยฑ3.84 / 23.23 ms โ 13.40 / 13.80 ยฑ0.33 / 14.25 ms โ +1.13x faster โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 19266.13ms โ
โ Total Time (refactor-materializing-smj) โ 19350.36ms โ
โ Average Time (HEAD) โ 448.05ms โ
โ Average Time (refactor-materializing-smj) โ 450.01ms โ
โ Queries Faster โ 1 โ
โ Queries Slower โ 3 โ
โ Queries with No Change โ 39 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 100.0s |
| Peak memory | 11.9 GiB |
| Avg memory | 4.5 GiB |
| CPU user | 993.6s |
| CPU sys | 66.6s |
| Peak spill | 0 B |
**clickbench_partitioned โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 100.0s |
| Peak memory | 12.2 GiB |
| Avg memory | 4.4 GiB |
| CPU user | 991.2s |
| CPU sys | 68.7s |
| 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
Re: [PR] chore: refactor `MaterializingSortMergeJoinStream` into generators and simplify code to be textbook like as possible [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5133540801
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5133512422)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5133512422-1306-lhxv7 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (cb881c764a1ca7d87fb3fb872ee26b944ad387e4) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..cb881c764a1ca7d87fb3fb872ee26b944ad387e4) using: smj
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5119916127
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5119660671)
**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 refactor-materializing-smj
--------------------
Benchmark tpcds_sf1.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 5.56 / 6.07 ยฑ0.89 / 7.85 ms โ 5.67 / 6.18 ยฑ0.92 / 8.01 ms โ no change โ
โ QQuery 2 โ 82.05 / 82.18 ยฑ0.13 / 82.40 ms โ 82.55 / 83.15 ยฑ0.45 / 83.83 ms โ no change โ
โ QQuery 3 โ 30.29 / 30.92 ยฑ1.14 / 33.19 ms โ 30.34 / 30.62 ยฑ0.21 / 30.92 ms โ no change โ
โ QQuery 4 โ 506.64 / 512.45 ยฑ3.50 / 517.53 ms โ 511.17 / 515.43 ยฑ3.90 / 522.68 ms โ no change โ
โ QQuery 5 โ 52.84 / 53.53 ยฑ0.79 / 54.97 ms โ 54.39 / 54.93 ยฑ0.34 / 55.32 ms โ no change โ
โ QQuery 6 โ 36.77 / 37.14 ยฑ0.30 / 37.63 ms โ 38.22 / 38.65 ยฑ0.49 / 39.46 ms โ no change โ
โ QQuery 7 โ 95.44 / 97.83 ยฑ3.57 / 104.90 ms โ 96.70 / 98.57 ยฑ2.42 / 103.25 ms โ no change โ
โ QQuery 8 โ 37.37 / 37.80 ยฑ0.30 / 38.19 ms โ 37.96 / 38.52 ยฑ0.39 / 38.97 ms โ no change โ
โ QQuery 9 โ 53.98 / 56.29 ยฑ1.21 / 57.23 ms โ 55.49 / 56.53 ยฑ0.64 / 57.30 ms โ no change โ
โ QQuery 10 โ 63.74 / 63.89 ยฑ0.13 / 64.07 ms โ 64.84 / 64.99 ยฑ0.15 / 65.25 ms โ no change โ
โ QQuery 11 โ 315.17 / 326.75 ยฑ11.14 / 340.57 ms โ 320.07 / 324.01 ยฑ3.92 / 330.58 ms โ no change โ
โ QQuery 12 โ 29.41 / 29.98 ยฑ0.40 / 30.65 ms โ 29.33 / 29.88 ยฑ0.35 / 30.37 ms โ no change โ
โ QQuery 13 โ 120.83 / 122.10 ยฑ1.52 / 125.01 ms โ 120.62 / 122.07 ยฑ1.60 / 125.11 ms โ no change โ
โ QQuery 14 โ 415.42 / 421.13 ยฑ6.62 / 433.50 ms โ 422.50 / 426.05 ยฑ5.25 / 436.48 ms โ no change โ
โ QQuery 15 โ 59.34 / 60.01 ยฑ0.42 / 60.48 ms โ 58.76 / 60.36 ยฑ1.44 / 62.93 ms โ no change โ
โ QQuery 16 โ 6.66 / 6.86 ยฑ0.21 / 7.24 ms โ 6.69 / 6.92 ยฑ0.20 / 7.23 ms โ no change โ
โ QQuery 17 โ 81.08 / 83.04 ยฑ2.66 / 88.18 ms โ 82.68 / 84.16 ยฑ1.78 / 87.42 ms โ no change โ
โ QQuery 18 โ 123.47 / 124.44 ยฑ0.85 / 125.98 ms โ 124.00 / 125.51 ยฑ1.23 / 127.75 ms โ no change โ
โ QQuery 19 โ 42.37 / 42.77 ยฑ0.38 / 43.31 ms โ 42.99 / 44.01 ยฑ1.60 / 47.19 ms โ no change โ
โ QQuery 20 โ 38.13 / 38.74 ยฑ0.71 / 40.07 ms โ 37.15 / 38.15 ยฑ0.60 / 38.86 ms โ no change โ
โ QQuery 21 โ 18.09 / 18.33 ยฑ0.23 / 18.70 ms โ 18.37 / 18.50 ยฑ0.14 / 18.77 ms โ no change โ
โ QQuery 22 โ 64.84 / 65.70 ยฑ0.49 / 66.23 ms โ 63.72 / 64.85 ยฑ0.83 / 66.28 ms โ no change โ
โ QQuery 23 โ 354.84 / 361.42 ยฑ5.56 / 369.49 ms โ 357.80 / 361.45 ยฑ2.41 / 365.11 ms โ no change โ
โ QQuery 24 โ 227.22 / 229.30 ยฑ2.24 / 232.81 ms โ 229.87 / 231.30 ยฑ0.79 / 232.17 ms โ no change โ
โ QQuery 25 โ 111.28 / 113.67 ยฑ1.97 / 117.11 ms โ 112.76 / 114.88 ยฑ1.58 / 116.92 ms โ no change โ
โ QQuery 26 โ 58.50 / 59.02 ยฑ0.29 / 59.38 ms โ 58.67 / 58.97 ยฑ0.23 / 59.25 ms โ no change โ
โ QQuery 27 โ 6.52 / 6.78 ยฑ0.18 / 7.06 ms โ 6.28 / 6.39 ยฑ0.12 / 6.62 ms โ +1.06x faster โ
โ QQuery 28 โ 57.29 / 61.69 ยฑ2.24 / 63.58 ms โ 62.15 / 62.64 ยฑ0.35 / 63.04 ms โ no change โ
โ QQuery 29 โ 98.68 / 101.85 ยฑ2.84 / 105.53 ms โ 99.38 / 102.55 ยฑ2.58 / 107.00 ms โ no change โ
โ QQuery 30 โ 32.92 / 33.21 ยฑ0.39 / 33.98 ms โ 33.07 / 33.23 ยฑ0.12 / 33.42 ms โ no change โ
โ QQuery 31 โ 113.69 / 114.27 ยฑ0.45 / 114.80 ms โ 114.56 / 116.95 ยฑ2.70 / 122.21 ms โ no change โ
โ QQuery 32 โ 21.08 / 21.68 ยฑ0.51 / 22.35 ms โ 21.57 / 21.89 ยฑ0.20 / 22.18 ms โ no change โ
โ QQuery 33 โ 38.56 / 39.14 ยฑ0.63 / 40.29 ms โ 39.40 / 39.78 ยฑ0.23 / 40.06 ms โ no change โ
โ QQuery 34 โ 10.11 / 10.58 ยฑ0.39 / 11.22 ms โ 10.29 / 10.60 ยฑ0.36 / 11.18 ms โ no change โ
โ QQuery 35 โ 74.16 / 75.08 ยฑ0.82 / 76.30 ms โ 73.93 / 75.12 ยฑ0.82 / 76.45 ms โ no change โ
โ QQuery 36 โ 5.86 / 6.06 ยฑ0.22 / 6.47 ms โ 6.10 / 6.27 ยฑ0.19 / 6.62 ms โ no change โ
โ QQuery 37 โ 6.88 / 7.01 ยฑ0.08 / 7.14 ms โ 7.00 / 7.15 ยฑ0.09 / 7.26 ms โ no change โ
โ QQuery 38 โ 63.40 / 65.04 ยฑ2.84 / 70.71 ms โ 65.23 / 66.40 ยฑ1.06 / 68.07 ms โ no change โ
โ QQuery 39 โ 90.78 / 92.28 ยฑ0.83 / 93.15 ms โ 91.34 / 92.65 ยฑ0.90 / 94.13 ms โ no change โ
โ QQuery 40 โ 23.80 / 24.92 ยฑ1.05 / 26.84 ms โ 23.89 / 24.28 ยฑ0.36 / 24.79 ms โ no change โ
โ QQuery 41 โ 11.84 / 12.01 ยฑ0.15 / 12.27 ms โ 11.85 / 12.07 ยฑ0.21 / 12.43 ms โ no change โ
โ QQuery 42 โ 24.43 / 24.95 ยฑ0.50 / 25.81 ms โ 24.87 / 25.17 ยฑ0.32 / 25.77 ms โ no change โ
โ QQuery 43 โ 4.81 / 4.99 ยฑ0.19 / 5.36 ms โ 5.11 / 6.19 ยฑ1.90 / 9.98 ms โ 1.24x slower โ
โ QQuery 44 โ 9.24 / 9.34 ยฑ0.06 / 9.39 ms โ 9.48 / 10.29 ยฑ1.22 / 12.71 ms โ 1.10x slower โ
โ QQuery 45 โ 38.54 / 41.51 ยฑ3.03 / 47.18 ms โ 40.55 / 40.99 ยฑ0.34 / 41.29 ms โ no change โ
โ QQuery 46 โ 12.16 / 12.58 ยฑ0.29 / 12.95 ms โ 12.19 / 12.52 ยฑ0.25 / 12.84 ms โ no change โ
โ QQuery 47 โ 235.27 / 241.14 ยฑ3.87 / 246.06 ms โ 237.16 / 243.41 ยฑ5.86 / 250.96 ms โ no change โ
โ QQuery 48 โ 96.82 / 97.84 ยฑ0.96 / 99.65 ms โ 98.76 / 99.90 ยฑ1.24 / 102.15 ms โ no change โ
โ QQuery 49 โ 78.01 / 79.00 ยฑ0.84 / 80.17 ms โ 78.07 / 79.59 ยฑ1.60 / 82.04 ms โ no change โ
โ QQuery 50 โ 59.72 / 60.26 ยฑ0.43 / 61.03 ms โ 60.36 / 60.96 ยฑ0.37 / 61.46 ms โ no change โ
โ QQuery 51 โ 93.79 / 95.03 ยฑ1.24 / 96.92 ms โ 93.78 / 98.43 ยฑ4.16 / 105.61 ms โ no change โ
โ QQuery 52 โ 24.44 / 25.11 ยฑ0.45 / 25.83 ms โ 25.36 / 25.54 ยฑ0.11 / 25.65 ms โ no change โ
โ QQuery 53 โ 29.97 / 30.33 ยฑ0.20 / 30.59 ms โ 30.52 / 30.84 ยฑ0.22 / 31.17 ms โ no change โ
โ QQuery 54 โ 56.13 / 56.46 ยฑ0.25 / 56.85 ms โ 56.71 / 57.26 ยฑ0.37 / 57.76 ms โ no change โ
โ QQuery 55 โ 23.97 / 24.18 ยฑ0.22 / 24.60 ms โ 24.66 / 24.85 ยฑ0.22 / 25.24 ms โ no change โ
โ QQuery 56 โ 40.19 / 40.67 ยฑ0.48 / 41.41 ms โ 40.13 / 40.68 ยฑ0.67 / 41.93 ms โ no change โ
โ QQuery 57 โ 178.37 / 182.02 ยฑ3.70 / 189.12 ms โ 180.49 / 182.62 ยฑ2.16 / 186.53 ms โ no change โ
โ QQuery 58 โ 116.98 / 119.47 ยฑ2.61 / 124.41 ms โ 118.05 / 119.53 ยฑ1.17 / 121.25 ms โ no change โ
โ QQuery 59 โ 119.16 / 119.72 ยฑ0.47 / 120.52 ms โ 119.45 / 120.40 ยฑ0.82 / 121.44 ms โ no change โ
โ QQuery 60 โ 40.49 / 42.83 ยฑ3.37 / 49.53 ms โ 40.79 / 41.94 ยฑ0.95 / 43.57 ms โ no change โ
โ QQuery 61 โ 12.45 / 12.60 ยฑ0.14 / 12.85 ms โ 12.65 / 12.75 ยฑ0.13 / 12.96 ms โ no change โ
โ QQuery 62 โ 47.02 / 47.63 ยฑ0.61 / 48.66 ms โ 47.35 / 47.65 ยฑ0.32 / 48.16 ms โ no change โ
โ QQuery 63 โ 30.05 / 30.26 ยฑ0.16 / 30.46 ms โ 30.83 / 30.97 ยฑ0.15 / 31.17 ms โ no change โ
โ QQuery 64 โ 414.93 / 417.96 ยฑ2.89 / 422.88 ms โ 413.83 / 416.58 ยฑ3.02 / 422.36 ms โ no change โ
โ QQuery 65 โ 129.21 / 134.21 ยฑ5.96 / 145.62 ms โ 131.41 / 132.73 ยฑ1.54 / 134.70 ms โ no change โ
โ QQuery 66 โ 80.69 / 82.72 ยฑ1.92 / 85.99 ms โ 82.40 / 83.63 ยฑ1.34 / 86.24 ms โ no change โ
โ QQuery 67 โ 253.51 / 257.34 ยฑ3.89 / 264.12 ms โ 252.92 / 257.05 ยฑ5.19 / 266.30 ms โ no change โ
โ QQuery 68 โ 12.22 / 12.39 ยฑ0.12 / 12.55 ms โ 12.28 / 12.56 ยฑ0.20 / 12.80 ms โ no change โ
โ QQuery 69 โ 58.07 / 58.41 ยฑ0.27 / 58.82 ms โ 58.58 / 59.26 ยฑ0.50 / 59.99 ms โ no change โ
โ QQuery 70 โ 106.46 / 108.12 ยฑ1.28 / 109.76 ms โ 107.83 / 109.22 ยฑ1.19 / 111.23 ms โ no change โ
โ QQuery 71 โ 37.10 / 40.13 ยฑ5.60 / 51.32 ms โ 36.39 / 38.78 ยฑ4.29 / 47.35 ms โ no change โ
โ QQuery 72 โ 2231.14 / 2331.94 ยฑ80.31 / 2436.04 ms โ 2186.64 / 2244.81 ยฑ45.53 / 2326.38 ms โ no change โ
โ QQuery 73 โ 9.65 / 9.89 ยฑ0.18 / 10.16 ms โ 9.83 / 10.13 ยฑ0.19 / 10.41 ms โ no change โ
โ QQuery 74 โ 178.57 / 183.13 ยฑ4.64 / 192.03 ms โ 181.90 / 185.94 ยฑ4.91 / 195.56 ms โ no change โ
โ QQuery 75 โ 149.26 / 151.51 ยฑ2.06 / 155.38 ms โ 152.99 / 154.78 ยฑ1.29 / 156.42 ms โ no change โ
โ QQuery 76 โ 35.93 / 36.28 ยฑ0.31 / 36.81 ms โ 36.40 / 36.65 ยฑ0.23 / 36.95 ms โ no change โ
โ QQuery 77 โ 62.65 / 66.30 ยฑ4.93 / 75.34 ms โ 63.37 / 64.00 ยฑ0.63 / 65.11 ms โ no change โ
โ QQuery 78 โ 202.48 / 209.00 ยฑ5.22 / 217.82 ms โ 203.98 / 206.51 ยฑ2.26 / 210.35 ms โ no change โ
โ QQuery 79 โ 68.61 / 69.52 ยฑ0.63 / 70.60 ms โ 69.96 / 72.06 ยฑ2.53 / 76.89 ms โ no change โ
โ QQuery 80 โ 101.57 / 102.10 ยฑ0.32 / 102.54 ms โ 100.35 / 102.67 ยฑ1.82 / 105.51 ms โ no change โ
โ QQuery 81 โ 26.69 / 30.61 ยฑ5.00 / 40.16 ms โ 27.09 / 27.35 ยฑ0.14 / 27.50 ms โ +1.12x faster โ
โ QQuery 82 โ 17.08 / 17.40 ยฑ0.18 / 17.61 ms โ 17.41 / 17.65 ยฑ0.17 / 17.82 ms โ no change โ
โ QQuery 83 โ 40.77 / 41.33 ยฑ0.31 / 41.64 ms โ 41.70 / 44.12 ยฑ3.43 / 50.95 ms โ 1.07x slower โ
โ QQuery 84 โ 30.54 / 30.87 ยฑ0.29 / 31.20 ms โ 31.27 / 31.57 ยฑ0.28 / 32.08 ms โ no change โ
โ QQuery 85 โ 107.73 / 111.68 ยฑ5.80 / 123.02 ms โ 108.70 / 111.01 ยฑ2.97 / 116.85 ms โ no change โ
โ QQuery 86 โ 25.96 / 26.78 ยฑ0.55 / 27.54 ms โ 26.54 / 28.50 ยฑ2.14 / 31.72 ms โ 1.06x slower โ
โ QQuery 87 โ 63.29 / 64.84 ยฑ1.19 / 66.29 ms โ 65.10 / 66.91 ยฑ2.04 / 70.76 ms โ no change โ
โ QQuery 88 โ 64.38 / 64.72 ยฑ0.21 / 64.99 ms โ 65.11 / 66.54 ยฑ2.20 / 70.89 ms โ no change โ
โ QQuery 89 โ 36.17 / 38.85 ยฑ3.42 / 45.45 ms โ 37.08 / 37.54 ยฑ0.25 / 37.85 ms โ no change โ
โ QQuery 90 โ 17.32 / 17.72 ยฑ0.27 / 18.07 ms โ 17.96 / 18.54 ยฑ0.86 / 20.24 ms โ no change โ
โ QQuery 91 โ 46.02 / 46.81 ยฑ0.77 / 48.28 ms โ 46.60 / 48.23 ยฑ1.55 / 51.14 ms โ no change โ
โ QQuery 92 โ 29.90 / 30.56 ยฑ0.56 / 31.24 ms โ 30.85 / 31.55 ยฑ0.59 / 32.55 ms โ no change โ
โ QQuery 93 โ 50.73 / 51.55 ยฑ0.44 / 52.01 ms โ 51.17 / 52.28 ยฑ0.82 / 53.67 ms โ no change โ
โ QQuery 94 โ 38.87 / 40.38 ยฑ2.62 / 45.60 ms โ 39.61 / 39.97 ยฑ0.38 / 40.67 ms โ no change โ
โ QQuery 95 โ 83.00 / 85.13 ยฑ3.42 / 91.93 ms โ 83.87 / 85.47 ยฑ1.51 / 87.93 ms โ no change โ
โ QQuery 96 โ 24.80 / 24.93 ยฑ0.08 / 25.03 ms โ 25.24 / 25.35 ยฑ0.07 / 25.44 ms โ no change โ
โ QQuery 97 โ 47.11 / 47.43 ยฑ0.19 / 47.65 ms โ 47.94 / 48.43 ยฑ0.36 / 48.86 ms โ no change โ
โ QQuery 98 โ 43.43 / 45.01 ยฑ1.39 / 47.63 ms โ 43.98 / 44.56 ยฑ0.39 / 45.18 ms โ no change โ
โ QQuery 99 โ 70.74 / 71.16 ยฑ0.42 / 71.94 ms โ 71.65 / 73.14 ยฑ1.02 / 74.56 ms โ no change โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 10309.60ms โ
โ Total Time (refactor-materializing-smj) โ 10275.15ms โ
โ Average Time (HEAD) โ 104.14ms โ
โ Average Time (refactor-materializing-smj) โ 103.79ms โ
โ Queries Faster โ 2 โ
โ Queries Slower โ 4 โ
โ Queries with No Change โ 93 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpcds โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 55.0s |
| Peak memory | 2.0 GiB |
| Avg memory | 1.3 GiB |
| CPU user | 233.9s |
| CPU sys | 6.2s |
| Peak spill | 0 B |
**tpcds โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 55.0s |
| Peak memory | 2.3 GiB |
| Avg memory | 1.6 GiB |
| CPU user | 233.9s |
| CPU sys | 6.2s |
| 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5133317449
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5133171057)
**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 refactor-materializing-smj
--------------------
Benchmark smj.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 8.48 / 8.82 ยฑ0.24 / 9.15 ms โ 8.36 / 8.49 ยฑ0.15 / 8.78 ms โ no change โ
โ QQuery 2 โ 162.13 / 163.05 ยฑ0.82 / 164.30 ms โ 160.51 / 162.01 ยฑ1.20 / 163.95 ms โ no change โ
โ QQuery 3 โ 103.43 / 104.76 ยฑ1.14 / 106.87 ms โ 101.78 / 115.74 ยฑ14.44 / 133.85 ms โ 1.10x slower โ
โ QQuery 4 โ 28.31 / 28.51 ยฑ0.11 / 28.64 ms โ 28.07 / 28.23 ยฑ0.09 / 28.32 ms โ no change โ
โ QQuery 5 โ 20.70 / 21.00 ยฑ0.20 / 21.24 ms โ 20.31 / 20.64 ยฑ0.17 / 20.74 ms โ no change โ
โ QQuery 6 โ 158.78 / 160.40 ยฑ2.04 / 164.34 ms โ 157.78 / 160.52 ยฑ3.11 / 166.48 ms โ no change โ
โ QQuery 7 โ 199.92 / 201.71 ยฑ0.98 / 202.57 ms โ 195.61 / 198.03 ยฑ1.75 / 200.62 ms โ no change โ
โ QQuery 8 โ 18.80 / 19.16 ยฑ0.22 / 19.48 ms โ 18.59 / 18.90 ยฑ0.21 / 19.21 ms โ no change โ
โ QQuery 9 โ 206.47 / 208.02 ยฑ0.91 / 209.22 ms โ 204.07 / 210.90 ยฑ10.89 / 232.52 ms โ no change โ
โ QQuery 10 โ 57.96 / 58.12 ยฑ0.09 / 58.22 ms โ 57.94 / 58.12 ยฑ0.10 / 58.25 ms โ no change โ
โ QQuery 11 โ 27.44 / 29.96 ยฑ4.98 / 39.91 ms โ 27.25 / 31.52 ยฑ8.25 / 48.02 ms โ 1.05x slower โ
โ QQuery 12 โ 65.20 / 70.15 ยฑ2.75 / 73.23 ms โ 63.17 / 66.37 ยฑ2.60 / 68.90 ms โ +1.06x faster โ
โ QQuery 13 โ 96.78 / 106.52 ยฑ5.29 / 112.18 ms โ 104.60 / 106.92 ยฑ1.55 / 108.47 ms โ no change โ
โ QQuery 14 โ 57.11 / 59.63 ยฑ2.32 / 63.98 ms โ 58.29 / 62.13 ยฑ5.96 / 73.86 ms โ no change โ
โ QQuery 15 โ 57.96 / 58.40 ยฑ0.27 / 58.69 ms โ 58.04 / 58.28 ยฑ0.28 / 58.78 ms โ no change โ
โ QQuery 16 โ 11.53 / 11.66 ยฑ0.10 / 11.85 ms โ 11.44 / 11.60 ยฑ0.13 / 11.78 ms โ no change โ
โ QQuery 17 โ 146.60 / 148.89 ยฑ3.33 / 155.46 ms โ 145.32 / 147.83 ยฑ3.50 / 154.76 ms โ no change โ
โ QQuery 18 โ 111.70 / 112.16 ยฑ0.29 / 112.51 ms โ 111.86 / 112.36 ยฑ0.31 / 112.78 ms โ no change โ
โ QQuery 19 โ 290.45 / 298.34 ยฑ6.59 / 305.40 ms โ 289.52 / 297.47 ยฑ7.10 / 309.21 ms โ no change โ
โ QQuery 20 โ 1259.66 / 1269.68 ยฑ16.25 / 1302.08 ms โ 1240.74 / 1247.42 ยฑ4.40 / 1252.58 ms โ no change โ
โ QQuery 21 โ 95.16 / 99.36 ยฑ6.11 / 111.38 ms โ 90.75 / 95.16 ยฑ7.21 / 109.54 ms โ no change โ
โ QQuery 22 โ 102.37 / 103.64 ยฑ0.83 / 104.87 ms โ 97.28 / 98.83 ยฑ1.05 / 100.39 ms โ no change โ
โ QQuery 23 โ 109.77 / 110.93 ยฑ1.07 / 112.77 ms โ 102.75 / 103.19 ยฑ0.54 / 104.26 ms โ +1.08x faster โ
โ QQuery 24 โ 27.37 / 27.45 ยฑ0.08 / 27.58 ms โ 27.44 / 27.60 ยฑ0.13 / 27.84 ms โ no change โ
โ QQuery 25 โ 65.24 / 70.88 ยฑ4.78 / 79.12 ms โ 68.39 / 72.37 ยฑ3.44 / 78.69 ms โ no change โ
โ QQuery 26 โ 103.95 / 107.66 ยฑ2.70 / 111.97 ms โ 96.81 / 107.37 ยฑ6.04 / 112.68 ms โ no change โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 3658.87ms โ
โ Total Time (refactor-materializing-smj) โ 3628.01ms โ
โ Average Time (HEAD) โ 140.73ms โ
โ Average Time (refactor-materializing-smj) โ 139.54ms โ
โ Queries Faster โ 2 โ
โ Queries Slower โ 2 โ
โ Queries with No Change โ 22 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**smj โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 20.0s |
| Peak memory | 1.2 GiB |
| Avg memory | 462.6 MiB |
| CPU user | 157.3s |
| CPU sys | 3.3s |
| Peak spill | 0 B |
**smj โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 20.0s |
| Peak memory | 1.2 GiB |
| Avg memory | 468.8 MiB |
| CPU user | 155.7s |
| CPU sys | 3.6s |
| 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5132849711
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5132649709)
**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 refactor-materializing-smj
--------------------
Benchmark tpch_sf1.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 39.54 / 40.62 ยฑ1.36 / 43.29 ms โ 39.42 / 40.11 ยฑ1.07 / 42.23 ms โ no change โ
โ QQuery 2 โ 19.95 / 20.28 ยฑ0.30 / 20.84 ms โ 19.85 / 20.53 ยฑ0.65 / 21.76 ms โ no change โ
โ QQuery 3 โ 32.09 / 34.73 ยฑ2.41 / 38.69 ms โ 33.52 / 34.35 ยฑ0.47 / 34.94 ms โ no change โ
โ QQuery 4 โ 18.05 / 18.53 ยฑ0.61 / 19.71 ms โ 18.09 / 18.46 ยฑ0.51 / 19.46 ms โ no change โ
โ QQuery 5 โ 39.90 / 43.12 ยฑ2.26 / 45.91 ms โ 42.20 / 42.53 ยฑ0.32 / 43.10 ms โ no change โ
โ QQuery 6 โ 17.57 / 17.84 ยฑ0.28 / 18.33 ms โ 17.00 / 17.26 ยฑ0.21 / 17.53 ms โ no change โ
โ QQuery 7 โ 46.39 / 48.43 ยฑ1.13 / 49.71 ms โ 47.15 / 49.01 ยฑ1.02 / 50.06 ms โ no change โ
โ QQuery 8 โ 44.39 / 45.45 ยฑ0.88 / 46.91 ms โ 44.31 / 45.24 ยฑ0.91 / 46.93 ms โ no change โ
โ QQuery 9 โ 51.39 / 53.93 ยฑ1.73 / 56.58 ms โ 51.51 / 52.47 ยฑ0.64 / 53.07 ms โ no change โ
โ QQuery 10 โ 44.97 / 45.58 ยฑ0.39 / 46.20 ms โ 43.92 / 44.58 ยฑ0.46 / 45.35 ms โ no change โ
โ QQuery 11 โ 14.29 / 14.76 ยฑ0.26 / 15.05 ms โ 14.11 / 14.33 ยฑ0.17 / 14.50 ms โ no change โ
โ QQuery 12 โ 25.26 / 26.69 ยฑ1.98 / 30.60 ms โ 24.33 / 25.94 ยฑ1.72 / 29.20 ms โ no change โ
โ QQuery 13 โ 32.86 / 35.84 ยฑ2.43 / 39.94 ms โ 32.54 / 35.03 ยฑ3.05 / 40.97 ms โ no change โ
โ QQuery 14 โ 24.46 / 25.33 ยฑ1.27 / 27.83 ms โ 24.32 / 24.63 ยฑ0.20 / 24.86 ms โ no change โ
โ QQuery 15 โ 32.50 / 32.88 ยฑ0.20 / 33.12 ms โ 32.12 / 32.85 ยฑ0.81 / 34.40 ms โ no change โ
โ QQuery 16 โ 14.50 / 14.62 ยฑ0.11 / 14.79 ms โ 14.70 / 14.84 ยฑ0.17 / 15.16 ms โ no change โ
โ QQuery 17 โ 77.91 / 79.75 ยฑ1.28 / 81.82 ms โ 75.08 / 75.83 ยฑ0.56 / 76.38 ms โ no change โ
โ QQuery 18 โ 63.38 / 64.93 ยฑ1.38 / 66.75 ms โ 61.52 / 62.86 ยฑ0.73 / 63.60 ms โ no change โ
โ QQuery 19 โ 34.50 / 35.06 ยฑ0.50 / 35.99 ms โ 33.74 / 34.23 ยฑ0.42 / 35.01 ms โ no change โ
โ QQuery 20 โ 33.33 / 34.08 ยฑ0.55 / 34.97 ms โ 33.42 / 33.69 ยฑ0.32 / 34.27 ms โ no change โ
โ QQuery 21 โ 58.84 / 59.40 ยฑ0.75 / 60.88 ms โ 57.96 / 58.81 ยฑ0.58 / 59.69 ms โ no change โ
โ QQuery 22 โ 14.95 / 16.41 ยฑ2.51 / 21.42 ms โ 14.64 / 14.99 ยฑ0.27 / 15.32 ms โ +1.09x faster โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 808.28ms โ
โ Total Time (refactor-materializing-smj) โ 792.57ms โ
โ Average Time (HEAD) โ 36.74ms โ
โ Average Time (refactor-materializing-smj) โ 36.03ms โ
โ Queries Faster โ 1 โ
โ Queries Slower โ 0 โ
โ Queries with No Change โ 21 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpch โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 5.0s |
| Peak memory | 1.2 GiB |
| Avg memory | 687.5 MiB |
| CPU user | 23.4s |
| CPU sys | 1.9s |
| Peak spill | 0 B |
**tpch โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 5.0s |
| Peak memory | 1.2 GiB |
| Avg memory | 514.6 MiB |
| CPU user | 23.1s |
| CPU sys | 1.9s |
| 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "rluvaton (via GitHub)" <gi...@apache.org>.
rluvaton commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5119659315
run benchmark smj
--
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5119706692
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5119659315)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5119659315-1289-ds2lx 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (ccdd36340b13836fbae8289748495bdad236d042) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..ccdd36340b13836fbae8289748495bdad236d042) using: smj
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5132684212
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5132649709)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5132649709-1302-krvm7 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (5bab8f5cacb00de1e8c507a809093de0ada1939a) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..5bab8f5cacb00de1e8c507a809093de0ada1939a) 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5132694263
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5132649709)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5132649709-1304-pvp6s 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (5bab8f5cacb00de1e8c507a809093de0ada1939a) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..5bab8f5cacb00de1e8c507a809093de0ada1939a) 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "rluvaton (via GitHub)" <gi...@apache.org>.
rluvaton commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5133171057
run benchmark smj
--
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "rluvaton (via GitHub)" <gi...@apache.org>.
rluvaton commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5133512422
run benchmark smj
--
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] chore: refactor `MaterializingSortMergeJoinStream` into generators and simplify code to be textbook like as possible [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5134658866
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5134520421)
**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 refactor-materializing-smj
--------------------
Benchmark tpch_sf10.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 315.56 / 317.48 ยฑ1.64 / 320.36 ms โ 316.61 / 317.92 ยฑ1.96 / 321.81 ms โ no change โ
โ QQuery 2 โ 103.40 / 106.34 ยฑ2.32 / 110.41 ms โ 105.55 / 107.13 ยฑ2.11 / 111.23 ms โ no change โ
โ QQuery 3 โ 236.21 / 239.54 ยฑ2.93 / 243.74 ms โ 231.76 / 238.24 ยฑ4.40 / 243.78 ms โ no change โ
โ QQuery 4 โ 114.08 / 115.93 ยฑ1.29 / 117.93 ms โ 113.54 / 115.91 ยฑ1.72 / 118.57 ms โ no change โ
โ QQuery 5 โ 353.13 / 370.64 ยฑ10.89 / 382.47 ms โ 359.34 / 362.17 ยฑ2.41 / 365.91 ms โ no change โ
โ QQuery 6 โ 128.88 / 131.80 ยฑ4.03 / 139.74 ms โ 129.12 / 131.88 ยฑ3.47 / 138.68 ms โ no change โ
โ QQuery 7 โ 466.09 / 474.65 ยฑ5.35 / 480.92 ms โ 473.74 / 484.60 ยฑ8.25 / 494.95 ms โ no change โ
โ QQuery 8 โ 387.93 / 397.03 ยฑ5.63 / 404.50 ms โ 387.85 / 393.57 ยฑ4.33 / 399.15 ms โ no change โ
โ QQuery 9 โ 563.23 / 572.05 ยฑ12.20 / 596.14 ms โ 557.78 / 568.11 ยฑ6.47 / 577.13 ms โ no change โ
โ QQuery 10 โ 296.74 / 306.42 ยฑ5.48 / 312.21 ms โ 296.33 / 305.54 ยฑ7.17 / 318.22 ms โ no change โ
โ QQuery 11 โ 82.81 / 92.37 ยฑ8.52 / 107.87 ms โ 86.40 / 88.67 ยฑ2.75 / 93.93 ms โ no change โ
โ QQuery 12 โ 178.46 / 186.48 ยฑ10.83 / 207.88 ms โ 182.03 / 189.62 ยฑ9.42 / 207.74 ms โ no change โ
โ QQuery 13 โ 289.23 / 304.24 ยฑ18.99 / 339.49 ms โ 293.91 / 306.49 ยฑ11.98 / 329.18 ms โ no change โ
โ QQuery 14 โ 173.23 / 177.62 ยฑ5.11 / 187.36 ms โ 173.33 / 177.85 ยฑ3.87 / 182.78 ms โ no change โ
โ QQuery 15 โ 307.13 / 311.03 ยฑ2.41 / 314.45 ms โ 308.33 / 310.47 ยฑ1.70 / 312.60 ms โ no change โ
โ QQuery 16 โ 65.50 / 67.79 ยฑ2.02 / 71.58 ms โ 64.02 / 67.72 ยฑ2.35 / 70.73 ms โ no change โ
โ QQuery 17 โ 650.14 / 662.01 ยฑ10.13 / 678.05 ms โ 640.47 / 652.84 ยฑ9.54 / 663.01 ms โ no change โ
โ QQuery 18 โ 707.26 / 724.25 ยฑ20.75 / 760.77 ms โ 699.20 / 717.38 ยฑ19.23 / 752.03 ms โ no change โ
โ QQuery 19 โ 247.08 / 262.60 ยฑ13.93 / 283.77 ms โ 249.84 / 263.01 ยฑ14.09 / 285.24 ms โ no change โ
โ QQuery 20 โ 287.60 / 295.91 ยฑ4.68 / 300.20 ms โ 273.91 / 292.31 ยฑ12.44 / 311.82 ms โ no change โ
โ QQuery 21 โ 668.39 / 673.81 ยฑ4.02 / 680.35 ms โ 665.72 / 677.55 ยฑ7.75 / 685.02 ms โ no change โ
โ QQuery 22 โ 59.94 / 63.87 ยฑ3.99 / 71.32 ms โ 61.61 / 64.63 ยฑ2.75 / 68.03 ms โ no change โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 6853.86ms โ
โ Total Time (refactor-materializing-smj) โ 6833.63ms โ
โ Average Time (HEAD) โ 311.54ms โ
โ Average Time (refactor-materializing-smj) โ 310.62ms โ
โ Queries Faster โ 0 โ
โ Queries Slower โ 0 โ
โ Queries with No Change โ 22 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpch10 โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 35.0s |
| Peak memory | 5.0 GiB |
| Avg memory | 1.5 GiB |
| CPU user | 344.8s |
| CPU sys | 19.6s |
| Peak spill | 0 B |
**tpch10 โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 35.0s |
| Peak memory | 5.1 GiB |
| Avg memory | 1.6 GiB |
| CPU user | 345.8s |
| CPU sys | 19.8s |
| 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] chore: refactor `MaterializingSortMergeJoinStream` into generators and simplify code to be textbook like as possible [datafusion]
Posted by "mbutrovich (via GitHub)" <gi...@apache.org>.
mbutrovich commented on code in PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#discussion_r3686398479
##########
datafusion/physical-plan/src/joins/sort_merge_join/materializing_stream.rs:
##########
@@ -995,15 +984,12 @@ impl MaterializingSortMergeJoinStream {
let bb = &mut self.buffered_data.batches[idx];
if let BufferedBatchState::Spilled(spill_file) = &bb.batch {
- if self.spill_stream.is_none() {
- let stream = self
- .spill_manager
- .read_spill_as_stream(Arc::clone(spill_file), None)?;
- self.spill_stream = Some(stream);
- }
+ let mut spill_stream = self
+ .spill_manager
+ .read_spill_as_stream(Arc::clone(spill_file), None)?;
- match ready!(self.spill_stream.as_mut().unwrap().poll_next_unpin(cx)) {
- Some(Ok(batch)) => {
+ match spill_stream.next().await.transpose()? {
Review Comment:
`restore_spilled_batches` awaits the spill stream here without `stop_join_time()`/`start_join_time()` around it, per the `stop_join_time` doc comment: "The join's own spill write and read-back are NOT excluded โ that time is join work." That's a real change from the old `poll_next`, which returned out via `ready!(self.poll_spilled_batches(cx, ...))` on a genuine `Poll::Pending`, stopping `_timer` for that wait.
This is the same `join_time` semantics change #23761 required calling out before merge (that PR's description was updated to note the metric now includes spill read-back time). This PR's description currently says "Are there any user-facing changes? nope" โ worth adding the same note here, since `join_time` is visible in `EXPLAIN ANALYZE`.
--
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5119700646
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5119660671)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5119660671-1291-bzfnj 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (ccdd36340b13836fbae8289748495bdad236d042) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..ccdd36340b13836fbae8289748495bdad236d042) 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5133203280
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5133171057)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5133171057-1305-7zf7z 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (5bab8f5cacb00de1e8c507a809093de0ada1939a) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..5bab8f5cacb00de1e8c507a809093de0ada1939a) using: smj
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5119703807
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5119660671)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5119660671-1292-rgbxd 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (ccdd36340b13836fbae8289748495bdad236d042) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..ccdd36340b13836fbae8289748495bdad236d042) 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5119975242
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5119660671)
**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 refactor-materializing-smj
--------------------
Benchmark clickbench_partitioned.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 0 โ 1.63 / 4.66 ยฑ5.87 / 16.40 ms โ 1.59 / 4.53 ยฑ5.71 / 15.94 ms โ no change โ
โ QQuery 1 โ 13.91 / 14.41 ยฑ0.42 / 15.02 ms โ 13.85 / 13.98 ยฑ0.07 / 14.05 ms โ no change โ
โ QQuery 2 โ 37.67 / 38.39 ยฑ0.97 / 40.31 ms โ 37.37 / 37.69 ยฑ0.32 / 38.13 ms โ no change โ
โ QQuery 3 โ 33.17 / 34.20 ยฑ0.95 / 35.79 ms โ 32.72 / 33.22 ยฑ0.30 / 33.53 ms โ no change โ
โ QQuery 4 โ 243.44 / 255.91 ยฑ12.23 / 278.48 ms โ 224.13 / 242.47 ยฑ9.34 / 249.44 ms โ +1.06x faster โ
โ QQuery 5 โ 272.87 / 276.46 ยฑ3.67 / 283.27 ms โ 271.69 / 273.61 ยฑ1.95 / 277.17 ms โ no change โ
โ QQuery 6 โ 1.30 / 1.45 ยฑ0.22 / 1.88 ms โ 1.32 / 1.47 ยฑ0.23 / 1.92 ms โ no change โ
โ QQuery 7 โ 14.34 / 15.53 ยฑ2.17 / 19.86 ms โ 13.70 / 13.87 ยฑ0.12 / 14.08 ms โ +1.12x faster โ
โ QQuery 8 โ 328.55 / 331.08 ยฑ3.24 / 336.97 ms โ 325.45 / 330.70 ยฑ4.08 / 337.23 ms โ no change โ
โ QQuery 9 โ 456.72 / 464.27 ยฑ8.16 / 479.43 ms โ 451.42 / 464.30 ยฑ9.06 / 478.37 ms โ no change โ
โ QQuery 10 โ 69.67 / 71.36 ยฑ1.33 / 73.64 ms โ 70.16 / 70.79 ยฑ0.74 / 71.82 ms โ no change โ
โ QQuery 11 โ 81.74 / 82.85 ยฑ0.64 / 83.51 ms โ 81.47 / 83.41 ยฑ1.59 / 86.23 ms โ no change โ
โ QQuery 12 โ 269.60 / 280.21 ยฑ8.97 / 296.27 ms โ 267.10 / 273.34 ยฑ5.31 / 279.36 ms โ no change โ
โ QQuery 13 โ 407.77 / 433.96 ยฑ24.27 / 477.12 ms โ 370.22 / 373.68 ยฑ4.70 / 382.73 ms โ +1.16x faster โ
โ QQuery 14 โ 287.17 / 295.88 ยฑ9.05 / 313.31 ms โ 280.01 / 285.63 ยฑ5.44 / 293.29 ms โ no change โ
โ QQuery 15 โ 274.47 / 280.46 ยฑ5.36 / 288.68 ms โ 266.48 / 272.33 ยฑ4.48 / 277.55 ms โ no change โ
โ QQuery 16 โ 610.22 / 623.80 ยฑ7.33 / 632.51 ms โ 611.15 / 633.43 ยฑ14.07 / 654.39 ms โ no change โ
โ QQuery 17 โ 613.35 / 628.81 ยฑ11.68 / 639.60 ms โ 628.88 / 664.48 ยฑ34.48 / 718.14 ms โ 1.06x slower โ
โ QQuery 18 โ 1248.43 / 1325.31 ยฑ81.62 / 1449.44 ms โ 1264.37 / 1308.50 ยฑ33.77 / 1363.06 ms โ no change โ
โ QQuery 19 โ 29.37 / 29.87 ยฑ0.59 / 30.98 ms โ 28.42 / 30.63 ยฑ3.09 / 36.59 ms โ no change โ
โ QQuery 20 โ 517.98 / 532.48 ยฑ18.05 / 567.68 ms โ 521.28 / 553.41 ยฑ28.56 / 605.93 ms โ no change โ
โ QQuery 21 โ 515.15 / 522.74 ยฑ5.82 / 532.19 ms โ 520.27 / 531.61 ยฑ8.41 / 544.66 ms โ no change โ
โ QQuery 22 โ 989.05 / 997.40 ยฑ7.35 / 1010.64 ms โ 989.80 / 998.89 ยฑ9.03 / 1014.92 ms โ no change โ
โ QQuery 23 โ 3124.26 / 3157.92 ยฑ39.32 / 3233.59 ms โ 3090.65 / 3136.35 ยฑ63.85 / 3262.52 ms โ no change โ
โ QQuery 24 โ 41.71 / 47.75 ยฑ10.33 / 68.33 ms โ 41.43 / 42.56 ยฑ0.80 / 43.86 ms โ +1.12x faster โ
โ QQuery 25 โ 111.07 / 115.70 ยฑ5.10 / 122.43 ms โ 111.05 / 113.39 ยฑ1.75 / 116.35 ms โ no change โ
โ QQuery 26 โ 41.87 / 50.27 ยฑ10.01 / 68.56 ms โ 41.68 / 42.17 ยฑ0.46 / 43.02 ms โ +1.19x faster โ
โ QQuery 27 โ 512.16 / 529.64 ยฑ9.39 / 537.03 ms โ 517.91 / 531.34 ยฑ9.42 / 543.82 ms โ no change โ
โ QQuery 28 โ 2891.26 / 2968.39 ยฑ50.51 / 3023.80 ms โ 2917.22 / 2932.91 ยฑ17.86 / 2966.65 ms โ no change โ
โ QQuery 29 โ 42.31 / 45.24 ยฑ4.83 / 54.89 ms โ 41.09 / 41.38 ยฑ0.23 / 41.76 ms โ +1.09x faster โ
โ QQuery 30 โ 298.31 / 317.51 ยฑ23.68 / 363.93 ms โ 303.90 / 316.00 ยฑ9.84 / 328.66 ms โ no change โ
โ QQuery 31 โ 275.74 / 288.49 ยฑ7.98 / 299.54 ms โ 284.87 / 299.69 ยฑ14.41 / 326.54 ms โ no change โ
โ QQuery 32 โ 921.87 / 939.13 ยฑ11.14 / 954.71 ms โ 921.71 / 967.84 ยฑ41.36 / 1028.23 ms โ no change โ
โ QQuery 33 โ 1462.13 / 1534.72 ยฑ97.05 / 1721.29 ms โ 1496.32 / 1575.32 ยฑ110.15 / 1789.57 ms โ no change โ
โ QQuery 34 โ 1466.24 / 1519.28 ยฑ34.52 / 1561.38 ms โ 1526.35 / 1600.91 ยฑ89.65 / 1760.60 ms โ 1.05x slower โ
โ QQuery 35 โ 281.91 / 300.12 ยฑ21.02 / 335.30 ms โ 278.54 / 296.00 ยฑ25.29 / 346.15 ms โ no change โ
โ QQuery 36 โ 67.21 / 73.40 ยฑ4.65 / 79.05 ms โ 65.31 / 72.03 ยฑ3.50 / 75.31 ms โ no change โ
โ QQuery 37 โ 35.83 / 37.77 ยฑ2.54 / 42.75 ms โ 36.42 / 40.14 ยฑ3.97 / 45.93 ms โ 1.06x slower โ
โ QQuery 38 โ 40.53 / 45.19 ยฑ5.32 / 55.09 ms โ 41.22 / 45.13 ยฑ2.67 / 49.51 ms โ no change โ
โ QQuery 39 โ 137.16 / 151.07 ยฑ8.54 / 161.35 ms โ 143.36 / 156.30 ยฑ7.56 / 166.80 ms โ no change โ
โ QQuery 40 โ 14.65 / 16.68 ยฑ3.00 / 22.53 ms โ 14.78 / 17.32 ยฑ4.50 / 26.30 ms โ no change โ
โ QQuery 41 โ 14.07 / 20.00 ยฑ10.93 / 41.82 ms โ 13.99 / 14.14 ยฑ0.13 / 14.35 ms โ +1.41x faster โ
โ QQuery 42 โ 13.39 / 15.48 ยฑ2.44 / 19.18 ms โ 13.16 / 14.11 ยฑ1.45 / 16.98 ms โ +1.10x faster โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 19715.23ms โ
โ Total Time (refactor-materializing-smj) โ 19750.99ms โ
โ Average Time (HEAD) โ 458.49ms โ
โ Average Time (refactor-materializing-smj) โ 459.33ms โ
โ Queries Faster โ 8 โ
โ Queries Slower โ 3 โ
โ Queries with No Change โ 32 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**clickbench_partitioned โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 100.0s |
| Peak memory | 10.5 GiB |
| Avg memory | 4.5 GiB |
| CPU user | 1009.2s |
| CPU sys | 73.8s |
| Peak spill | 0 B |
**clickbench_partitioned โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 100.0s |
| Peak memory | 11.9 GiB |
| Avg memory | 4.6 GiB |
| CPU user | 1007.7s |
| CPU sys | 75.9s |
| 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "rluvaton (via GitHub)" <gi...@apache.org>.
rluvaton commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5132649709
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] chore: refactor `MaterializingSortMergeJoinStream` into generators and simplify code to be textbook like as possible [datafusion]
Posted by "rluvaton (via GitHub)" <gi...@apache.org>.
rluvaton merged PR #23976:
URL: https://github.com/apache/datafusion/pull/23976
--
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] chore: refactor `MaterializingSortMergeJoinStream` into generators and simplify code to be textbook like as possible [datafusion]
Posted by "mbutrovich (via GitHub)" <gi...@apache.org>.
mbutrovich commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5134520421
run benchmark tpch10
```
env:
PREFER_HASH_JOIN: false
```
--
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "rluvaton (via GitHub)" <gi...@apache.org>.
rluvaton commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5132631045
run benchmark smj
--
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5132686896
๐ค Benchmark running (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5132649709)
**Instance:** `c4a-highmem-16` (12 vCPU / 65 GiB) | `Linux bench-c5132649709-1303-44pmk 6.12.85+ #1 SMP Mon May 11 08:17:35 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 refactor-materializing-smj (5bab8f5cacb00de1e8c507a809093de0ada1939a) to 68d5874 (merge-base) [diff](https://github.com/apache/datafusion/compare/68d587468073c1fc84ffe00db66f35d506dcc828..5bab8f5cacb00de1e8c507a809093de0ada1939a) 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "rluvaton (via GitHub)" <gi...@apache.org>.
rluvaton commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5119660671
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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "adriangbot (via GitHub)" <gi...@apache.org>.
adriangbot commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5119894062
๐ค Benchmark completed (GKE) | [trigger](https://github.com/apache/datafusion/pull/23976#issuecomment-5119660671)
**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 refactor-materializing-smj
--------------------
Benchmark tpch_sf1.json
--------------------
โโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโโโโโโ
โ Query โ HEAD โ refactor-materializing-smj โ Change โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ QQuery 1 โ 38.53 / 39.94 ยฑ1.24 / 41.73 ms โ 38.67 / 39.75 ยฑ1.80 / 43.34 ms โ no change โ
โ QQuery 2 โ 19.38 / 19.57 ยฑ0.19 / 19.91 ms โ 19.15 / 19.52 ยฑ0.36 / 20.21 ms โ no change โ
โ QQuery 3 โ 30.79 / 32.80 ยฑ1.85 / 35.75 ms โ 31.42 / 33.54 ยฑ1.60 / 36.38 ms โ no change โ
โ QQuery 4 โ 17.35 / 17.53 ยฑ0.12 / 17.68 ms โ 17.43 / 17.69 ยฑ0.19 / 17.93 ms โ no change โ
โ QQuery 5 โ 39.81 / 40.52 ยฑ0.44 / 41.09 ms โ 37.69 / 39.88 ยฑ2.27 / 43.81 ms โ no change โ
โ QQuery 6 โ 16.42 / 16.97 ยฑ0.90 / 18.77 ms โ 16.30 / 16.70 ยฑ0.48 / 17.64 ms โ no change โ
โ QQuery 7 โ 43.03 / 44.00 ยฑ1.24 / 46.44 ms โ 43.19 / 45.03 ยฑ1.41 / 47.18 ms โ no change โ
โ QQuery 8 โ 42.62 / 42.78 ยฑ0.21 / 43.18 ms โ 42.44 / 42.74 ยฑ0.27 / 43.23 ms โ no change โ
โ QQuery 9 โ 49.16 / 49.87 ยฑ0.57 / 50.80 ms โ 48.70 / 51.32 ยฑ1.82 / 53.55 ms โ no change โ
โ QQuery 10 โ 41.78 / 42.20 ยฑ0.33 / 42.76 ms โ 42.15 / 42.45 ยฑ0.31 / 43.02 ms โ no change โ
โ QQuery 11 โ 13.31 / 13.59 ยฑ0.23 / 13.98 ms โ 13.40 / 14.00 ยฑ0.68 / 14.93 ms โ no change โ
โ QQuery 12 โ 23.69 / 24.54 ยฑ1.03 / 26.48 ms โ 23.61 / 24.11 ยฑ0.31 / 24.58 ms โ no change โ
โ QQuery 13 โ 32.06 / 33.96 ยฑ1.30 / 35.85 ms โ 33.04 / 34.17 ยฑ0.80 / 35.08 ms โ no change โ
โ QQuery 14 โ 23.45 / 23.72 ยฑ0.20 / 23.95 ms โ 23.62 / 23.73 ยฑ0.09 / 23.85 ms โ no change โ
โ QQuery 15 โ 31.24 / 32.20 ยฑ1.00 / 33.47 ms โ 31.05 / 31.36 ยฑ0.21 / 31.68 ms โ no change โ
โ QQuery 16 โ 13.94 / 14.01 ยฑ0.06 / 14.08 ms โ 13.68 / 13.88 ยฑ0.22 / 14.21 ms โ no change โ
โ QQuery 17 โ 72.72 / 79.06 ยฑ3.75 / 84.14 ms โ 70.47 / 70.79 ยฑ0.29 / 71.30 ms โ +1.12x faster โ
โ QQuery 18 โ 64.20 / 65.71 ยฑ1.60 / 68.38 ms โ 57.86 / 60.07 ยฑ1.33 / 61.57 ms โ +1.09x faster โ
โ QQuery 19 โ 34.74 / 35.39 ยฑ0.56 / 36.27 ms โ 32.99 / 33.43 ยฑ0.40 / 34.07 ms โ +1.06x faster โ
โ QQuery 20 โ 34.50 / 34.72 ยฑ0.23 / 35.15 ms โ 31.81 / 32.05 ยฑ0.14 / 32.20 ms โ +1.08x faster โ
โ QQuery 21 โ 61.37 / 62.68 ยฑ1.85 / 66.36 ms โ 55.02 / 57.26 ยฑ1.71 / 59.59 ms โ +1.09x faster โ
โ QQuery 22 โ 15.32 / 16.02 ยฑ0.70 / 17.05 ms โ 13.90 / 14.13 ยฑ0.16 / 14.38 ms โ +1.13x faster โ
โโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโณโโโโโโโโโโโ
โ Benchmark Summary โ โ
โกโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฉ
โ Total Time (HEAD) โ 781.79ms โ
โ Total Time (refactor-materializing-smj) โ 757.60ms โ
โ Average Time (HEAD) โ 35.54ms โ
โ Average Time (refactor-materializing-smj) โ 34.44ms โ
โ Queries Faster โ 6 โ
โ Queries Slower โ 0 โ
โ Queries with No Change โ 16 โ
โ Queries with Failure โ 0 โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโดโโโโโโโโโโโ
```
</p>
</details>
<details><summary>Resource Usage</summary>
**tpch โ base (merge-base)**
| Metric | Value |
|--------|-------|
| Wall time | 5.0s |
| Peak memory | 1.2 GiB |
| Avg memory | 519.7 MiB |
| CPU user | 22.8s |
| CPU sys | 1.8s |
| Peak spill | 0 B |
**tpch โ branch**
| Metric | Value |
|--------|-------|
| Wall time | 5.0s |
| Peak memory | 1.1 GiB |
| Avg memory | 488.1 MiB |
| CPU user | 21.7s |
| CPU sys | 1.6s |
| 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] chore: refactor `MaterializingSortMergeJoinStream` into generators [datafusion]
Posted by "codecov-commenter (via GitHub)" <gi...@apache.org>.
codecov-commenter commented on PR #23976:
URL: https://github.com/apache/datafusion/pull/23976#issuecomment-5120278371
## [Codecov](https://app.codecov.io/gh/apache/datafusion/pull/23976?dropdown=coverage&src=pr&el=h1&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache) Report
:x: Patch coverage is `86.71587%` with `36 lines` in your changes missing coverage. Please review.
:white_check_mark: Project coverage is 80.75%. Comparing base ([`68d5874`](https://app.codecov.io/gh/apache/datafusion/commit/68d587468073c1fc84ffe00db66f35d506dcc828?dropdown=coverage&el=desc&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache)) to head ([`ccdd363`](https://app.codecov.io/gh/apache/datafusion/commit/ccdd36340b13836fbae8289748495bdad236d042?dropdown=coverage&el=desc&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache)).
:warning: Report is 5 commits behind head on main.
| [Files with missing lines](https://app.codecov.io/gh/apache/datafusion/pull/23976?dropdown=coverage&src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache) | Patch % | Lines |
|---|---|---|
| [.../src/joins/sort\_merge\_join/materializing\_stream.rs](https://app.codecov.io/gh/apache/datafusion/pull/23976?src=pr&el=tree&filepath=datafusion%2Fphysical-plan%2Fsrc%2Fjoins%2Fsort_merge_join%2Fmaterializing_stream.rs&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache#diff-ZGF0YWZ1c2lvbi9waHlzaWNhbC1wbGFuL3NyYy9qb2lucy9zb3J0X21lcmdlX2pvaW4vbWF0ZXJpYWxpemluZ19zdHJlYW0ucnM=) | 86.66% | [10 Missing and 26 partials :warning: ](https://app.codecov.io/gh/apache/datafusion/pull/23976?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache) |
<details><summary>Additional details and impacted files</summary>
```diff
@@ Coverage Diff @@
## main #23976 +/- ##
========================================
Coverage 80.75% 80.75%
========================================
Files 1096 1096
Lines 373282 373497 +215
Branches 373282 373497 +215
========================================
+ Hits 301440 301617 +177
- Misses 53869 53887 +18
- Partials 17973 17993 +20
```
</details>
[:umbrella: View full report in Codecov by Harness](https://app.codecov.io/gh/apache/datafusion/pull/23976?dropdown=coverage&src=pr&el=continue&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache).
:loudspeaker: Have feedback on the report? [Share it here](https://about.codecov.io/codecov-pr-comment-feedback/?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=apache).
<details><summary> :rocket: New features to boost your workflow: </summary>
- :snowflake: [Test Analytics](https://docs.codecov.com/docs/test-analytics): Detect flaky tests, report on failures, and find test suite problems.
- :package: [JS Bundle Analysis](https://docs.codecov.com/docs/javascript-bundle-analysis): Save yourself from yourself by tracking and limiting bundle sizes in JS merges.
</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