Skip to content

perf(plan): reuse grouped sums after early key qualification - #32800

Open
aunjgr wants to merge 9 commits into
matrixorigin:mainfrom
aunjgr:perf/grouped-measure-reuse
Open

aunjgr wants to merge 9 commits into
matrixorigin:mainfrom
aunjgr:perf/grouped-measure-reuse

Conversation

@aunjgr

@aunjgr aunjgr commented Nov 16, 2026

Copy link
Copy Markdown
Contributor

What type of PR is this?

  • API-change
  • BUG
  • Improvement
  • Documentation
  • Feature
  • Test and CI
  • Code Refactoring

Which issue(s) this PR fixes:

Related to #30857

What this PR does / why we need it:

Reuse a qualifying grouped SUM when a later aggregate computes the same exact measure over the same fact rows through key-preserving dimension joins. Carry the existing measure through those joins, remove the redundant fact scan and connector, and replace the repeated aggregate with a projection.

Keep qualification as SEMI through normal join ordering and pushdown. Apply reuse only at a lowered edge whose probe is a direct surviving dimension scan, so qualification filters the probe before the other dimensions materialize its rows. Preserve the dimension predicates and existing runtime-filter planning.

Require matching input bags, snapshots, filters, key and argument types, total exact SUM arithmetic, and catalog-proven dimension uniqueness. Retain the original plan for unsupported or unprofitable cases, including disabled SEMI pushdown. Preserve NULLs, missing dimensions, output metadata, HAVING, ordering and limits.

Design

Reviewed design.

Validation

Focused rewrite tests and the owning planner package pass. The full-planner test verifies that qualification remains below the surviving customer join. Negative tests cover join-spanning and retired-fact probes, disabled pushdown, incompatible bags/types and missing uniqueness proofs.

The changed SQL case passes normal mo-tester comparison: 19 statements, twice on the same task-owned instance, with schema teardown checked after each run. Owning-package static checks and convention checks pass. Loaded-schema operator diagnostics confirm early qualification, one fact scan and normal qualifier parallelism.

The local all22 comparison covers SF10 memory/disk and SF100 disk, with excluded cold suites and five warm suites per process. Disk comparisons control OS data-page residency while retaining MO's configured cache. Results and timing/metric limitations are posted in a separate comment. The full repository pre-push static gate passes with the repository-pinned tools in the default CI build mode. Functional tests and benchmark builds use SIMD/v3. The optional whole-repository amd64 SIMD lint mode also reports two unsafe-allowlist failures on unchanged main in vector-distance files; this PR does not modify them.

@qodo-code-review

Copy link
Copy Markdown

Qodo reviews are paused for this user.

Troubleshooting steps vary by plan Learn more →

On a Teams plan?
Reviews resume once this user has a paid seat and their Git account is linked in Qodo.
Link Git account →

Using GitHub Enterprise Server, GitLab Self-Managed, or Bitbucket Data Center?
These require an Enterprise plan - Contact us
Contact us →

aunjgr commented Nov 16, 2026

Copy link
Copy Markdown
Contributor Author

Validated against clean main 7fcd9bdff8433d72983f0399fd5c10dcab91a00a, candidate 3a3067b9f8ec390514f67e49ad8dcaeb23e339a6. Each configuration uses one persistent process, one excluded cold full suite and five warm Q1–Q22 suites; Go 1.27.1, SIMD/v3, identical release metadata, all 32 CPUs,48 GiB application budget and16 GiB shared CN cache. No spill, swap, OOM or CPU/compiler-interference exclusion occurred in the retained captures. Results are drained; scripts do not verify dataset/schema/row/result contents.

For disk mode, MO cache stays warm while OS data-file pages start cold before each query. The harness evicts only the task clone's original immutable SHARED files, checks zero residency using fincore metadata, then starts query measurements. All 528 cold/warm disk boundaries passed. Eviction/metadata time is outside query latency. Memory-mode evidence is reused from the same unchanged sources/binaries/configuration.

Configuration Q18 wall Q18 instructions Q18 Go allocation bytes Total wall
SF10 memory −4.60% −2.60% −8.70% −1.05%
SF10 disk −3.75% −2.65% −4.32% −1.48%
SF100 disk −24.23% −28.21% −40.19% −1.35%

The native Q18 diagnostic confirms early qualification:624 ORDERS rows and623 CUSTOMER rows produce624 dimension-join rows, with one LINEITEM scan and ordinary32-way qualifier parallelism. The failed earlier prototype materialized15M dimension-join rows before qualification; that implementation is superseded. No executor/cache algorithm was changed.

Timing variation is disclosed, not subtracted from wall scores. SF100 Q17 raw medians are+11.39%, but three paired suite positions are nearly unchanged/faster; slow episodes also occur in main. Main five wall samples:5.0210,3.5677,3.5972,3.5991,3.5937s; candidate:4.9698,4.0068,3.6074,4.8195,3.5991s. Physical reads remain9.38–9.48 GB and CPU work remains about 33 CPU-seconds; estimated non-GC instructions differ+0.63%. The rule does not rewrite Q17. These captures do not show a stable increase in Q17 work; the raw slower median remains visible below. SF10 memory Q6's+26.61% wall/+27.64% raw instruction difference coincides with GC, while estimated non-GC instructions differ+0.06%.

Before controlling OS pages, unchanged main varied79.46 s→71.12 s with stable MO cache hits/requested bytes but different physical I/O. Those captures are retained as diagnostics and are superseded for disk latency acceptance; they are not pooled with these tables.

sf10-memory: all22 queries, Total, CPU and allocations
Query Main wall s Candidate wall s Wall Δ Main CPU s Candidate CPU s Main instructions G Candidate instructions G Instructions Δ
1 0.210789 0.203272 -3.57% 5.92527 5.82988 67.9618 66.7803 -1.74%
2 0.027307 0.027636 +1.20% 0.36479 0.34713 2.5722 2.5777 +0.21%
3 0.099490 0.095429 -4.08% 2.25740 2.14889 8.9593 8.9587 -0.01%
4 0.059298 0.057317 -3.34% 1.45143 1.42657 10.6672 10.6647 -0.02%
5 0.075049 0.074708 -0.45% 1.56855 1.55130 8.9198 8.9150 -0.05%
6 0.030927 0.039157 +26.61% 0.65824 0.80372 3.9022 4.9809 +27.64%
7 0.076878 0.076249 -0.82% 1.82560 1.78736 7.7223 7.7185 -0.05%
8 0.239372 0.231746 -3.19% 2.07349 2.03649 12.9676 13.0609 +0.72%
9 0.230664 0.228662 -0.87% 5.20299 5.21369 37.6517 37.5037 -0.39%
10 0.181203 0.177116 -2.26% 4.37687 4.34134 24.8697 24.8619 -0.03%
11 0.029973 0.029928 -0.15% 0.42622 0.41662 2.9080 2.9033 -0.16%
12 0.073807 0.079481 +7.69% 1.92250 2.11861 12.5035 13.7390 +9.88%
13 0.217795 0.222472 +2.15% 6.02241 5.99518 44.3991 44.3565 -0.10%
14 0.078142 0.078439 +0.38% 0.97213 1.04677 7.8681 7.8692 +0.01%
15 0.148961 0.150288 +0.89% 1.49240 1.50943 11.0256 11.0704 +0.41%
16 0.124179 0.124455 +0.22% 1.38133 1.39458 7.8545 7.8628 +0.11%
17 0.073438 0.074610 +1.60% 1.81851 1.80223 15.2758 15.2757 -0.00%
18 0.364254 0.347495 -4.60% 7.05553 6.62630 21.6856 21.1214 -2.60%
19 0.105493 0.104703 -0.75% 2.57963 2.57609 21.7390 21.7406 +0.01%
20 0.198839 0.195995 -1.43% 3.94134 3.82392 18.0995 18.0338 -0.36%
21 0.312528 0.306040 -2.08% 8.08804 7.96020 68.9698 68.9722 +0.00%
22 0.039943 0.041558 +4.04% 0.76241 0.75273 6.8922 6.8914 -0.01%
Total 2.998329 2.966756 -1.05% 62.16709 61.50902 425.4148 425.8585 +0.10%

Allocation counters include common process/scrape and asynchronous engine work. No globally exact copy-byte counter is available. GC CPU is measured process GC CPU, not a wall-time subtraction.

| Query | Main Go MB | Candidate Go MB | Go bytes Δ | Main Go objects | Candidate Go objects | Main GC CPU s | Candidate GC CPU s |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| 1 | 68.278 | 56.819 | -16.78% | 468457 | 440767 | 0.1712 | 0.0000 |
| 2 | 22.666 | 24.819 | +9.50% | 297745 | 336296 | 0.0000 | 0.0000 |
| 3 | 36.800 | 38.153 | +3.68% | 461621 | 479695 | 0.0000 | 0.0000 |
| 4 | 40.946 | 40.296 | -1.59% | 368759 | 363096 | 0.0000 | 0.0000 |
| 5 | 50.050 | 50.347 | +0.59% | 431135 | 433797 | 0.0000 | 0.0000 |
| 6 | 40.246 | 51.754 | +28.59% | 278682 | 300626 | 0.0000 | 0.1967 |
| 7 | 56.201 | 54.973 | -2.19% | 444814 | 439425 | 0.0000 | 0.0000 |
| 8 | 286.047 | 284.352 | -0.59% | 2459146 | 2437331 | 0.2480 | 0.2577 |
| 9 | 156.942 | 156.770 | -0.11% | 2501618 | 2507160 | 0.1857 | 0.1861 |
| 10 | 73.755 | 59.921 | -18.76% | 817361 | 797378 | 0.0000 | 0.0000 |
| 11 | 33.104 | 32.050 | -3.19% | 551892 | 545413 | 0.0000 | 0.0000 |
| 12 | 53.530 | 68.094 | +27.21% | 444177 | 468601 | 0.0000 | 0.1950 |
| 13 | 863.768 | 864.733 | +0.11% | 15310932 | 15296374 | 0.5366 | 0.5281 |
| 14 | 48.286 | 49.030 | +1.54% | 1036125 | 1043535 | 0.0000 | 0.0000 |
| 15 | 55.417 | 58.644 | +5.82% | 668970 | 702319 | 0.0000 | 0.0000 |
| 16 | 38.900 | 40.354 | +3.74% | 659803 | 670727 | 0.0000 | 0.0000 |
| 17 | 80.382 | 79.944 | -0.55% | 552501 | 574026 | 0.1784 | 0.1670 |
| 18 | 44.397 | 40.534 | -8.70% | 404279 | 368926 | 0.0000 | 0.0000 |
| 19 | 61.711 | 62.296 | +0.95% | 565604 | 582164 | 0.0000 | 0.0000 |
| 20 | 130.188 | 130.814 | +0.48% | 2394921 | 2406997 | 0.1931 | 0.2002 |
| 21 | 101.190 | 100.722 | -0.46% | 740307 | 765100 | 0.0000 | 0.0000 |
| 22 | 21.972 | 21.842 | -0.60% | 232884 | 229831 | 0.0000 | 0.0000 |
| Total | 2364.777 | 2367.258 | +0.10% | 32091733 | 32189584 | 1.5130 | 1.7309 |

Native allocation attribution limit: the default allocator publishes counters after a 100 ms window; MPool publishes after a one-second window. Their query-boundary increments are retained in raw JSON, but can belong to neighboring queries. They are not exact per-query allocation evidence and are not used to declare a per-query allocation pass. IO/cache counters overlap the default allocator and must not be added again. Long-window native totals also have endpoint lag; exact per-query native costs need separate operator/component evidence.

External process cores: main 0.042, candidate 0.029; interrupt cores separately: 0.337, 0.334.

Period-weighted instruction samples estimate GC attribution. These are sampling estimates, not exact foreground counters, GC-subtracted wall times, or independent-process confidence intervals. Raw hardware counters are above.

Query Main estimated non-GC instructions G Candidate estimated non-GC instructions G Δ
1 66.9377 66.7781 -0.24%
2 2.5722 2.5777 +0.21%
3 8.9593 8.9587 -0.01%
4 10.6672 10.6615 -0.05%
5 8.9198 8.9150 -0.05%
6 3.9022 3.9046 +0.06%
7 7.7223 7.7185 -0.05%
8 12.0094 11.8266 -1.52%
9 36.0546 36.4017 +0.96%
10 24.8697 24.8619 -0.03%
11 2.9052 2.9033 -0.07%
12 12.5035 12.7364 +1.86%
13 39.9560 39.3743 -1.46%
14 7.8681 7.8644 -0.05%
15 11.0256 11.0704 +0.41%
16 7.8545 7.8568 +0.03%
17 14.1123 14.1139 +0.01%
18 21.6856 21.1214 -2.60%
19 21.7390 21.7406 +0.01%
20 16.8097 16.8181 +0.05%
21 68.9698 68.9722 +0.00%
22 6.8922 6.8914 -0.01%
Total 414.9362 414.0673 -0.21%
sf10-disk: all22 queries, Total, CPU and allocations
Query Main wall s Candidate wall s Wall Δ Main CPU s Candidate CPU s Main instructions G Candidate instructions G Instructions Δ
1 0.212374 0.199037 -6.28% 5.99793 5.81829 68.0013 66.7770 -1.80%
2 0.027810 0.027959 +0.53% 0.37494 0.34571 2.5817 2.5744 -0.28%
3 0.093580 0.094200 +0.66% 2.18404 2.18768 8.9560 9.0420 +0.96%
4 0.059974 0.059535 -0.73% 1.43015 1.43007 10.6684 10.6686 +0.00%
5 0.083360 0.075100 -9.91% 1.77796 1.56807 10.1151 8.9203 -11.81%
6 0.032828 0.031869 -2.92% 0.67111 0.64314 3.9011 3.9034 +0.06%
7 0.076653 0.077212 +0.73% 1.80798 1.79902 7.7179 7.7175 -0.01%
8 0.231316 0.219148 -5.26% 2.06918 2.05696 13.2183 13.3138 +0.72%
9 0.223584 0.224257 +0.30% 5.05198 4.97255 36.3798 36.3749 -0.01%
10 0.182647 0.179547 -1.70% 4.44489 4.38946 24.8609 24.8648 +0.02%
11 0.028946 0.029449 +1.74% 0.42264 0.41842 2.9069 2.9044 -0.09%
12 0.072314 0.074473 +2.98% 1.87830 1.84187 12.5016 12.5036 +0.02%
13 0.213453 0.214440 +0.46% 5.90684 5.96952 44.4379 44.4528 +0.03%
14 0.077026 0.081703 +6.07% 0.99345 0.94977 7.8748 7.8645 -0.13%
15 0.151872 0.152175 +0.20% 1.48805 1.46889 11.0399 11.0312 -0.08%
16 0.123236 0.123937 +0.57% 1.35236 1.36855 7.8535 7.8583 +0.06%
17 0.074077 0.075980 +2.57% 1.74048 1.81291 15.4322 15.3291 -0.67%
18 0.335472 0.322906 -3.75% 6.57530 6.00736 21.7052 21.1299 -2.65%
19 0.105901 0.108811 +2.75% 2.54032 2.57102 21.7499 21.7497 -0.00%
20 0.192934 0.188258 -2.42% 3.93465 3.78748 17.9951 18.0081 +0.07%
21 0.303348 0.299725 -1.19% 7.96686 7.92589 68.9425 68.9466 +0.01%
22 0.042135 0.041455 -1.61% 0.74388 0.77997 6.8901 6.8962 +0.09%
Total 2.944840 2.901175 -1.48% 61.35329 60.11259 425.7301 422.8310 -0.68%

Allocation counters include common process/scrape and asynchronous engine work. No globally exact copy-byte counter is available. GC CPU is measured process GC CPU, not a wall-time subtraction.

| Query | Main Go MB | Candidate Go MB | Go bytes Δ | Main Go objects | Candidate Go objects | Main GC CPU s | Candidate GC CPU s |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| 1 | 70.654 | 53.347 | -24.50% | 453584 | 404576 | 0.1612 | 0.0000 |
| 2 | 24.865 | 23.270 | -6.42% | 329315 | 317037 | 0.0000 | 0.0000 |
| 3 | 37.450 | 49.542 | +32.29% | 477044 | 498706 | 0.0000 | 0.0000 |
| 4 | 39.214 | 39.615 | +1.02% | 331842 | 358225 | 0.0000 | 0.0000 |
| 5 | 61.253 | 50.094 | -18.22% | 492157 | 452709 | 0.1766 | 0.0000 |
| 6 | 38.687 | 51.577 | +33.32% | 280930 | 295439 | 0.0000 | 0.1869 |
| 7 | 54.959 | 54.579 | -0.69% | 430502 | 431541 | 0.0000 | 0.0000 |
| 8 | 285.302 | 287.064 | +0.62% | 2447898 | 2456082 | 0.2815 | 0.3008 |
| 9 | 155.194 | 151.943 | -2.09% | 2504456 | 2481916 | 0.0000 | 0.0000 |
| 10 | 61.089 | 61.781 | +1.13% | 819857 | 820762 | 0.0000 | 0.0000 |
| 11 | 29.844 | 31.984 | +7.17% | 506856 | 539749 | 0.0000 | 0.0000 |
| 12 | 53.158 | 53.157 | -0.00% | 425564 | 423004 | 0.0000 | 0.0000 |
| 13 | 869.278 | 866.445 | -0.33% | 15313730 | 15299738 | 0.5425 | 0.5449 |
| 14 | 49.894 | 47.729 | -4.34% | 1062510 | 1039468 | 0.0000 | 0.0000 |
| 15 | 56.737 | 56.912 | +0.31% | 674905 | 678342 | 0.0000 | 0.0000 |
| 16 | 38.889 | 41.290 | +6.17% | 660883 | 687056 | 0.0000 | 0.0000 |
| 17 | 83.539 | 79.328 | -5.04% | 582972 | 576471 | 0.1712 | 0.1829 |
| 18 | 43.779 | 41.889 | -4.32% | 410859 | 374489 | 0.0000 | 0.0000 |
| 19 | 63.916 | 62.460 | -2.28% | 589538 | 574078 | 0.0000 | 0.0000 |
| 20 | 129.904 | 129.831 | -0.06% | 2392441 | 2412999 | 0.1918 | 0.1798 |
| 21 | 98.802 | 98.369 | -0.44% | 730251 | 734077 | 0.0000 | 0.0000 |
| 22 | 21.909 | 22.295 | +1.76% | 230491 | 241127 | 0.0000 | 0.0000 |
| Total | 2368.317 | 2354.500 | -0.58% | 32148585 | 32097591 | 1.5248 | 1.3952 |

Native allocation attribution limit: the default allocator publishes counters after a 100 ms window; MPool publishes after a one-second window. Their query-boundary increments are retained in raw JSON, but can belong to neighboring queries. They are not exact per-query allocation evidence and are not used to declare a per-query allocation pass. IO/cache counters overlap the default allocator and must not be added again. Long-window native totals also have endpoint lag; exact per-query native costs need separate operator/component evidence.

External process cores: main 0.096, candidate 0.047; interrupt cores separately: 0.324, 0.305.

Period-weighted instruction samples estimate GC attribution. These are sampling estimates, not exact foreground counters, GC-subtracted wall times, or independent-process confidence intervals. Raw hardware counters are above.

Query Main estimated non-GC instructions G Candidate estimated non-GC instructions G Δ
1 66.7838 66.7692 -0.02%
2 2.5817 2.5744 -0.28%
3 8.9560 9.0420 +0.96%
4 10.6684 10.6686 +0.00%
5 9.0248 8.9203 -1.16%
6 3.9007 3.9034 +0.07%
7 7.7179 7.7175 -0.01%
8 11.5511 11.7730 +1.92%
9 36.3767 36.3703 -0.02%
10 24.8609 24.8648 +0.02%
11 2.9069 2.9044 -0.09%
12 12.5016 12.5036 +0.02%
13 39.8667 39.5937 -0.68%
14 7.8748 7.8645 -0.13%
15 11.0399 11.0312 -0.08%
16 7.8535 7.8583 +0.06%
17 14.2229 14.1177 -0.74%
18 21.7021 21.1299 -2.64%
19 21.7499 21.7497 -0.00%
20 16.7969 16.7862 -0.06%
21 68.9425 68.9309 -0.02%
22 6.8901 6.8962 +0.09%
Total 414.7699 413.9697 -0.19%
sf100-disk: all22 queries, Total, CPU and allocations
Query Main wall s Candidate wall s Wall Δ Main CPU s Candidate CPU s Main instructions G Candidate instructions G Instructions Δ
1 5.607918 5.537550 -1.25% 91.64826 91.21056 1037.3672 1032.9403 -0.43%
2 0.470551 0.450286 -4.31% 5.21031 4.96279 36.4945 36.3801 -0.31%
3 3.129885 3.130131 +0.01% 39.97475 39.15617 202.3039 202.5797 +0.14%
4 2.862489 2.859637 -0.10% 35.49394 34.72067 221.0854 221.0233 -0.03%
5 3.427368 3.457601 +0.88% 37.73021 37.47399 183.1307 183.5055 +0.20%
6 2.110619 2.099625 -0.52% 16.59698 16.47945 116.3207 116.1495 -0.15%
7 3.573726 3.589754 +0.45% 36.00522 35.91132 177.7633 177.8794 +0.07%
8 3.956431 3.924777 -0.80% 35.20820 35.59547 204.2915 204.4648 +0.08%
9 5.969186 6.017872 +0.82% 82.76058 83.52299 523.0007 523.1906 +0.04%
10 4.633936 4.652337 +0.40% 70.64927 70.22441 479.1331 478.7803 -0.07%
11 0.577605 0.570917 -1.16% 5.24975 5.20192 34.4821 34.1445 -0.98%
12 4.014130 4.022292 +0.20% 46.94804 46.65887 369.8554 370.3363 +0.13%
13 3.024741 3.001765 -0.76% 78.05577 77.93487 497.2175 497.9119 +0.14%
14 2.522844 2.533372 +0.42% 22.27034 22.45150 136.7906 138.1804 +1.02%
15 4.744526 4.762380 +0.38% 41.15588 40.32763 229.4247 229.5670 +0.06%
16 0.654794 0.562885 -14.04% 10.78051 10.12879 61.4257 61.2171 -0.34%
17 3.597249 4.006803 +11.39% 32.79950 33.08897 222.5920 224.9443 +1.06%
18 5.645838 4.277701 -24.23% 88.71253 74.13851 372.6760 267.5584 -28.21%
19 5.044750 5.060133 +0.30% 60.69573 60.72783 573.9564 573.9201 -0.01%
20 3.840463 3.817214 -0.61% 50.30993 49.85518 223.3373 223.5120 +0.08%
21 9.225770 9.228763 +0.03% 120.97755 119.25363 909.1515 911.1222 +0.22%
22 0.471293 0.473080 +0.38% 9.63983 9.65213 73.2199 73.2350 +0.02%
Total 79.106110 78.036873 -1.35% 1018.87310 998.67766 6885.0201 6782.5425 -1.49%

Allocation counters include common process/scrape and asynchronous engine work. No globally exact copy-byte counter is available. GC CPU is measured process GC CPU, not a wall-time subtraction.

| Query | Main Go MB | Candidate Go MB | Go bytes Δ | Main Go objects | Candidate Go objects | Main GC CPU s | Candidate GC CPU s |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| 1 | 952.004 | 944.233 | -0.82% | 8681955 | 8585753 | 0.4023 | 0.4165 |
| 2 | 179.504 | 178.654 | -0.47% | 2223829 | 2218964 | 0.0000 | 0.0000 |
| 3 | 641.633 | 644.884 | +0.51% | 7499626 | 7457596 | 0.3816 | 0.3723 |
| 4 | 618.530 | 627.174 | +1.40% | 5836601 | 5818352 | 0.3749 | 0.3735 |
| 5 | 695.610 | 694.249 | -0.20% | 6535249 | 6485290 | 0.0000 | 0.0000 |
| 6 | 581.319 | 581.610 | +0.05% | 4542146 | 4541838 | 0.4075 | 0.3598 |
| 7 | 825.010 | 835.301 | +1.25% | 7300268 | 7320707 | 0.4036 | 0.3909 |
| 8 | 1173.405 | 1173.716 | +0.03% | 10271424 | 10277300 | 0.3907 | 0.4130 |
| 9 | 1840.554 | 1839.405 | -0.06% | 28685800 | 28709725 | 0.7714 | 0.7762 |
| 10 | 834.206 | 833.602 | -0.07% | 10491016 | 10477639 | 0.4126 | 0.4116 |
| 11 | 259.428 | 256.774 | -1.02% | 4368071 | 4316096 | 0.0000 | 0.0000 |
| 12 | 834.248 | 849.092 | +1.78% | 7899633 | 7957401 | 0.3802 | 0.3917 |
| 13 | 8475.718 | 8473.176 | -0.03% | 151653153 | 151605296 | 3.2635 | 3.4289 |
| 14 | 660.648 | 666.033 | +0.82% | 12358796 | 12421802 | 0.3603 | 0.3425 |
| 15 | 1006.643 | 1010.049 | +0.34% | 11608275 | 11611960 | 0.4104 | 0.3992 |
| 16 | 226.877 | 226.879 | +0.00% | 4782602 | 4793516 | 0.0000 | 0.0000 |
| 17 | 942.885 | 963.789 | +2.22% | 8598589 | 8684993 | 0.3777 | 0.3834 |
| 18 | 930.786 | 556.685 | -40.19% | 8394582 | 4846650 | 0.4096 | 0.0000 |
| 19 | 975.981 | 982.947 | +0.71% | 9546045 | 9552008 | 0.3904 | 0.3938 |
| 20 | 1385.987 | 1389.613 | +0.26% | 25822544 | 25844471 | 0.4630 | 0.5359 |
| 21 | 1713.816 | 1718.224 | +0.26% | 15938288 | 15955065 | 0.8696 | 0.8471 |
| 22 | 89.861 | 88.149 | -1.91% | 925491 | 900490 | 0.0000 | 0.0000 |
| Total | 25844.654 | 25534.239 | -1.20% | 353963983 | 350382912 | 10.4694 | 10.2364 |

Native allocation attribution limit: the default allocator publishes counters after a 100 ms window; MPool publishes after a one-second window. Their query-boundary increments are retained in raw JSON, but can belong to neighboring queries. They are not exact per-query allocation evidence and are not used to declare a per-query allocation pass. IO/cache counters overlap the default allocator and must not be added again. Long-window native totals also have endpoint lag; exact per-query native costs need separate operator/component evidence.

External process cores: main 0.000, candidate 0.116; interrupt cores separately: 0.548, 0.552.

Period-weighted instruction samples estimate GC attribution. These are sampling estimates, not exact foreground counters, GC-subtracted wall times, or independent-process confidence intervals. Raw hardware counters are above.

Query Main estimated non-GC instructions G Candidate estimated non-GC instructions G Δ
1 1035.1816 1030.8632 -0.42%
2 36.0753 36.3801 +0.85%
3 200.3257 201.0194 +0.35%
4 219.6379 219.6785 +0.02%
5 183.1307 183.5025 +0.20%
6 114.0660 114.2424 +0.15%
7 175.7508 175.9898 +0.14%
8 201.9718 201.5934 -0.19%
9 519.5319 520.5730 +0.20%
10 476.7749 476.3665 -0.09%
11 33.8535 34.1445 +0.86%
12 368.2466 367.9192 -0.09%
13 473.3717 473.7565 +0.08%
14 134.9190 135.6698 +0.56%
15 227.5167 227.8686 +0.15%
16 60.9499 61.1725 +0.37%
17 221.5771 222.9767 +0.63%
18 371.3201 267.5584 -27.94%
19 571.7542 571.8988 +0.03%
20 219.6281 219.9513 +0.15%
21 906.5293 906.0869 -0.05%
22 73.2199 73.2350 +0.02%
Total 6825.3326 6722.4470 -1.51%

@matrix-meow matrix-meow added the size/L Denotes a PR that changes [500,999] lines label Nov 16, 2026
@mergify

mergify Bot commented Nov 16, 2026 •

Copy link
Copy Markdown
Contributor

Queue requested — the merge queue status continues in this comment ↓.

@mergify

mergify Bot commented Nov 16, 2026 •

Copy link
Copy Markdown
Contributor

⏳ Waiting for queue conditions

Enters the queue as soon as one of these 7 queue rules has nothing left.

Queue rule main · 1 of 8 left

Queue rule release-2.1 · 1 of 8 left

Queue rule release-2.2 · 1 of 8 left

4 other queue rules

Queue rule release-3.0 · 1 of 8 left

Queue rule release-4.0 · 1 of 8 left

Queue rule release-4.1 · 1 of 8 left

Queue rule release-4.2 · 1 of 8 left

Queue rule main: 7 met
  • #review-threads-unresolved = 0 · required by GitHub branch protection
  • any of · required by GitHub branch protection:
    • check-success = CI Required
    • check-neutral = CI Required
    • check-skipped = CI Required
  • -closed
  • -conflict
  • -draft
  • any of:
    • -mergify-configuration-changed
    • check-success = @mergify/Configuration changed
    • check-success = @mergify/Configuration has been deleted
  • any of:
    • check-neutral = @mergify/Mergify Merge Protections
    • check-skipped = @mergify/Mergify Merge Protections
    • check-success = @mergify/Mergify Merge Protections
Queue rule release-2.1: 7 met
  • #review-threads-unresolved = 0 · required by GitHub branch protection
  • any of · required by GitHub branch protection:
    • check-success = CI Required
    • check-neutral = CI Required
    • check-skipped = CI Required
  • -closed
  • -conflict
  • -draft
  • any of:
    • -mergify-configuration-changed
    • check-success = @mergify/Configuration changed
    • check-success = @mergify/Configuration has been deleted
  • any of:
    • check-neutral = @mergify/Mergify Merge Protections
    • check-skipped = @mergify/Mergify Merge Protections
    • check-success = @mergify/Mergify Merge Protections
Queue rule release-2.2: 7 met
  • #review-threads-unresolved = 0 · required by GitHub branch protection
  • any of · required by GitHub branch protection:
    • check-success = CI Required
    • check-neutral = CI Required
    • check-skipped = CI Required
  • -closed
  • -conflict
  • -draft
  • any of:
    • -mergify-configuration-changed
    • check-success = @mergify/Configuration changed
    • check-success = @mergify/Configuration has been deleted
  • any of:
    • check-neutral = @mergify/Mergify Merge Protections
    • check-skipped = @mergify/Mergify Merge Protections
    • check-success = @mergify/Mergify Merge Protections
Queue rule release-3.0: 7 met
  • #review-threads-unresolved = 0 · required by GitHub branch protection
  • any of · required by GitHub branch protection:
    • check-success = CI Required
    • check-neutral = CI Required
    • check-skipped = CI Required
  • -closed
  • -conflict
  • -draft
  • any of:
    • -mergify-configuration-changed
    • check-success = @mergify/Configuration changed
    • check-success = @mergify/Configuration has been deleted
  • any of:
    • check-neutral = @mergify/Mergify Merge Protections
    • check-skipped = @mergify/Mergify Merge Protections
    • check-success = @mergify/Mergify Merge Protections
Queue rule release-4.0: 7 met
  • #review-threads-unresolved = 0 · required by GitHub branch protection
  • any of · required by GitHub branch protection:
    • check-success = CI Required
    • check-neutral = CI Required
    • check-skipped = CI Required
  • -closed
  • -conflict
  • -draft
  • any of:
    • -mergify-configuration-changed
    • check-success = @mergify/Configuration changed
    • check-success = @mergify/Configuration has been deleted
  • any of:
    • check-neutral = @mergify/Mergify Merge Protections
    • check-skipped = @mergify/Mergify Merge Protections
    • check-success = @mergify/Mergify Merge Protections
Queue rule release-4.1: 7 met
  • #review-threads-unresolved = 0 · required by GitHub branch protection
  • any of · required by GitHub branch protection:
    • check-success = CI Required
    • check-neutral = CI Required
    • check-skipped = CI Required
  • -closed
  • -conflict
  • -draft
  • any of:
    • -mergify-configuration-changed
    • check-success = @mergify/Configuration changed
    • check-success = @mergify/Configuration has been deleted
  • any of:
    • check-neutral = @mergify/Mergify Merge Protections
    • check-skipped = @mergify/Mergify Merge Protections
    • check-success = @mergify/Mergify Merge Protections
Queue rule release-4.2: 7 met
  • #review-threads-unresolved = 0 · required by GitHub branch protection
  • any of · required by GitHub branch protection:
    • check-success = CI Required
    • check-neutral = CI Required
    • check-skipped = CI Required
  • -closed
  • -conflict
  • -draft
  • any of:
    • -mergify-configuration-changed
    • check-success = @mergify/Configuration changed
    • check-success = @mergify/Configuration has been deleted
  • any of:
    • check-neutral = @mergify/Mergify Merge Protections
    • check-skipped = @mergify/Mergify Merge Protections
    • check-success = @mergify/Mergify Merge Protections

Times in UTC · dashboard · how the merge queue works

This branch was previously deployed

1 inactive deployment
ci — 3a3067b9 Deployed Nov 16, 2026 by aunjgr via Matrixone Coverage execution / Coverage #64622
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

kind/enhancement size/L Denotes a PR that changes [500,999] lines

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants