Skip to content

Store accumulator-stage lookups directly#10645

Open
charlesmyu wants to merge 3 commits intomasterfrom
charles.yu/djm-0000/stop-storing-stage-accumulators
Open

Store accumulator-stage lookups directly#10645
charlesmyu wants to merge 3 commits intomasterfrom
charles.yu/djm-0000/stop-storing-stage-accumulators

Conversation

@charlesmyu
Copy link
Contributor

@charlesmyu charlesmyu commented Feb 19, 2026

What Does This Do

Stores the map of accumulator ID to stage ID directly in a Long - Integer map, rather than storing the stage-level accumulators themselves.

Motivation

The stage-level accumulators naively roll up metrics from the task level by summing all values. This means they are not accurate in all cases, particular for metrics that are better visualized as a distribution of values across all tasks instead of a single sum.

In #10553 we rollup task-level metrics ourselves and encode them into the Spark SQL metrics as distributions in order to improve the granularity of information collected. This left one remaining use of the stage-level accumulators - mapping operations in Spark SQL plans to their respective stages.

Since we do not need the entire accumulator to accomplish this, we should simplify that to a ID-ID map instead to save space. We can also remove the 50k limit since we are not storing the entire accumulator, allowing us to avoid creating orphaned operations (previously occurred when the EldestHashMap overflowed).

Additional Notes

I didn't feel the explicit need for a FF here, since the change is fairly straightforwards and shouldn't negatively impact perf. However, if we feel strongly otherwise I'm happy to make that happen.

Contributor Checklist

Jira ticket: [PROJ-IDENT]

Note: Once your PR is ready to merge, add it to the merge queue by commenting /merge. /merge -c cancels the queue request. /merge -f --reason "reason" skips all merge queue checks; please use this judiciously, as some checks do not run at the PR-level. For more information, see this doc.

@charlesmyu charlesmyu added the inst: apache spark Apache Spark instrumentation label Feb 19, 2026
@charlesmyu charlesmyu mentioned this pull request Feb 19, 2026
Merged
@charlesmyu charlesmyu force-pushed the charles.yu/djm-0000/fix-spark-plan-metrics branch from e52fbc5 to e413d1d Compare February 19, 2026 22:02
@charlesmyu charlesmyu force-pushed the charles.yu/djm-0000/stop-storing-stage-accumulators branch from e1c532c to 9594586 Compare February 19, 2026 22:02
@pr-commenter
Copy link

pr-commenter bot commented Feb 19, 2026 •
edited
Loading

Benchmarks

Startup

Parameters

Baseline Candidate
baseline_or_candidate baseline candidate
git_branch master charles.yu/djm-0000/stop-storing-stage-accumulators
git_commit_date 1773169596 1773179914
git_commit_sha cc12228 4c54b6b
release_version 1.61.0-SNAPSHOT~cc122288e5 1.61.0-SNAPSHOT~4c54b6b488
See matching parameters
Baseline Candidate
application insecure-bank insecure-bank
ci_job_date 1773181637 1773181637
ci_job_id 1494902080 1494902080
ci_pipeline_id 101755727 101755727
cpu_model Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz
kernel_version Linux runner-zfyrx7zua-project-304-concurrent-0-pu8d1064 6.8.0-1031-aws #33~22.04.1-Ubuntu SMP Thu Jun 26 14:22:30 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux Linux runner-zfyrx7zua-project-304-concurrent-0-pu8d1064 6.8.0-1031-aws #33~22.04.1-Ubuntu SMP Thu Jun 26 14:22:30 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux
module Agent Agent
parent None None

Summary

Found 0 performance improvements and 0 performance regressions! Performance is the same for 66 metrics, 5 unstable metrics.

Startup time reports for insecure-bank 
gantt
    title insecure-bank - global startup overhead: candidate=1.61.0-SNAPSHOT~4c54b6b488, baseline=1.61.0-SNAPSHOT~cc122288e5

    dateFormat X
    axisFormat %s
section tracing
Agent [baseline] (1.057 s) : 0, 1057197
Total [baseline] (8.839 s) : 0, 8838951
Agent [candidate] (1.061 s) : 0, 1060529
Total [candidate] (8.826 s) : 0, 8825501
section iast
Agent [baseline] (1.228 s) : 0, 1228202
Total [baseline] (9.562 s) : 0, 9561941
Agent [candidate] (1.24 s) : 0, 1239552
Total [candidate] (9.58 s) : 0, 9580283
Loading
  • baseline results
Module Variant Duration Δ tracing
Agent tracing 1.057 s -
Agent iast 1.228 s 171.006 ms (16.2%)
Total tracing 8.839 s -
Total iast 9.562 s 722.99 ms (8.2%)
  • candidate results
Module Variant Duration Δ tracing
Agent tracing 1.061 s -
Agent iast 1.24 s 179.023 ms (16.9%)
Total tracing 8.826 s -
Total iast 9.58 s 754.782 ms (8.6%) 
gantt
    title insecure-bank - break down per module: candidate=1.61.0-SNAPSHOT~4c54b6b488, baseline=1.61.0-SNAPSHOT~cc122288e5

    dateFormat X
    axisFormat %s
section tracing
crashtracking [baseline] (1.177 ms) : 0, 1177
crashtracking [candidate] (1.201 ms) : 0, 1201
BytebuddyAgent [baseline] (626.917 ms) : 0, 626917
BytebuddyAgent [candidate] (628.885 ms) : 0, 628885
AgentMeter [baseline] (29.027 ms) : 0, 29027
AgentMeter [candidate] (29.105 ms) : 0, 29105
GlobalTracer [baseline] (256.276 ms) : 0, 256276
GlobalTracer [candidate] (257.699 ms) : 0, 257699
AppSec [baseline] (31.52 ms) : 0, 31520
AppSec [candidate] (31.566 ms) : 0, 31566
Debugger [baseline] (58.462 ms) : 0, 58462
Debugger [candidate] (58.782 ms) : 0, 58782
Remote Config [baseline] (584.325 µs) : 0, 584
Remote Config [candidate] (589.061 µs) : 0, 589
Telemetry [baseline] (8.656 ms) : 0, 8656
Telemetry [candidate] (8.724 ms) : 0, 8724
Flare Poller [baseline] (8.583 ms) : 0, 8583
Flare Poller [candidate] (7.931 ms) : 0, 7931
section iast
crashtracking [baseline] (1.195 ms) : 0, 1195
crashtracking [candidate] (1.22 ms) : 0, 1220
BytebuddyAgent [baseline] (797.115 ms) : 0, 797115
BytebuddyAgent [candidate] (806.043 ms) : 0, 806043
AgentMeter [baseline] (11.323 ms) : 0, 11323
AgentMeter [candidate] (11.593 ms) : 0, 11593
GlobalTracer [baseline] (247.613 ms) : 0, 247613
GlobalTracer [candidate] (249.182 ms) : 0, 249182
AppSec [baseline] (26.429 ms) : 0, 26429
AppSec [candidate] (27.295 ms) : 0, 27295
Debugger [baseline] (62.922 ms) : 0, 62922
Debugger [candidate] (62.261 ms) : 0, 62261
Remote Config [baseline] (519.083 µs) : 0, 519
Remote Config [candidate] (536.438 µs) : 0, 536
Telemetry [baseline] (14.892 ms) : 0, 14892
Telemetry [candidate] (14.905 ms) : 0, 14905
Flare Poller [baseline] (4.928 ms) : 0, 4928
Flare Poller [candidate] (4.86 ms) : 0, 4860
IAST [baseline] (25.184 ms) : 0, 25184
IAST [candidate] (25.35 ms) : 0, 25350
Loading
Startup time reports for petclinic 
gantt
    title petclinic - global startup overhead: candidate=1.61.0-SNAPSHOT~4c54b6b488, baseline=1.61.0-SNAPSHOT~cc122288e5

    dateFormat X
    axisFormat %s
section tracing
Agent [baseline] (1.07 s) : 0, 1069528
Total [baseline] (11.142 s) : 0, 11141810
Agent [candidate] (1.067 s) : 0, 1066995
Total [candidate] (11.17 s) : 0, 11170191
section appsec
Agent [baseline] (1.243 s) : 0, 1242947
Total [baseline] (11.149 s) : 0, 11148507
Agent [candidate] (1.245 s) : 0, 1245413
Total [candidate] (11.125 s) : 0, 11124694
section iast
Agent [baseline] (1.238 s) : 0, 1237926
Total [baseline] (11.291 s) : 0, 11291196
Agent [candidate] (1.228 s) : 0, 1228480
Total [candidate] (11.372 s) : 0, 11371948
section profiling
Agent [baseline] (1.182 s) : 0, 1182394
Total [baseline] (10.977 s) : 0, 10976588
Agent [candidate] (1.188 s) : 0, 1187847
Total [candidate] (11.055 s) : 0, 11055256
Loading
  • baseline results
Module Variant Duration Δ tracing
Agent tracing 1.07 s -
Agent appsec 1.243 s 173.42 ms (16.2%)
Agent iast 1.238 s 168.399 ms (15.7%)
Agent profiling 1.182 s 112.866 ms (10.6%)
Total tracing 11.142 s -
Total appsec 11.149 s 6.698 ms (0.1%)
Total iast 11.291 s 149.386 ms (1.3%)
Total profiling 10.977 s -165.221 ms (-1.5%)
  • candidate results
Module Variant Duration Δ tracing
Agent tracing 1.067 s -
Agent appsec 1.245 s 178.418 ms (16.7%)
Agent iast 1.228 s 161.485 ms (15.1%)
Agent profiling 1.188 s 120.851 ms (11.3%)
Total tracing 11.17 s -
Total appsec 11.125 s -45.497 ms (-0.4%)
Total iast 11.372 s 201.757 ms (1.8%)
Total profiling 11.055 s -114.935 ms (-1.0%) 
gantt
    title petclinic - break down per module: candidate=1.61.0-SNAPSHOT~4c54b6b488, baseline=1.61.0-SNAPSHOT~cc122288e5

    dateFormat X
    axisFormat %s
section tracing
crashtracking [baseline] (1.197 ms) : 0, 1197
crashtracking [candidate] (1.199 ms) : 0, 1199
BytebuddyAgent [baseline] (634.476 ms) : 0, 634476
BytebuddyAgent [candidate] (633.475 ms) : 0, 633475
AgentMeter [baseline] (29.4 ms) : 0, 29400
AgentMeter [candidate] (29.319 ms) : 0, 29319
GlobalTracer [baseline] (259.307 ms) : 0, 259307
GlobalTracer [candidate] (259.182 ms) : 0, 259182
AppSec [baseline] (31.845 ms) : 0, 31845
AppSec [candidate] (31.839 ms) : 0, 31839
Debugger [baseline] (59.921 ms) : 0, 59921
Debugger [candidate] (59.923 ms) : 0, 59923
Remote Config [baseline] (593.595 µs) : 0, 594
Remote Config [candidate] (598.377 µs) : 0, 598
Telemetry [baseline] (8.696 ms) : 0, 8696
Telemetry [candidate] (8.674 ms) : 0, 8674
Flare Poller [baseline] (7.884 ms) : 0, 7884
Flare Poller [candidate] (6.583 ms) : 0, 6583
section appsec
crashtracking [baseline] (1.181 ms) : 0, 1181
crashtracking [candidate] (1.188 ms) : 0, 1188
BytebuddyAgent [baseline] (655.606 ms) : 0, 655606
BytebuddyAgent [candidate] (656.589 ms) : 0, 656589
AgentMeter [baseline] (12.065 ms) : 0, 12065
AgentMeter [candidate] (12.059 ms) : 0, 12059
GlobalTracer [baseline] (258.092 ms) : 0, 258092
GlobalTracer [candidate] (258.927 ms) : 0, 258927
IAST [baseline] (23.858 ms) : 0, 23858
IAST [candidate] (23.935 ms) : 0, 23935
AppSec [baseline] (177.098 ms) : 0, 177098
AppSec [candidate] (177.602 ms) : 0, 177602
Debugger [baseline] (65.7 ms) : 0, 65700
Debugger [candidate] (65.73 ms) : 0, 65730
Remote Config [baseline] (568.268 µs) : 0, 568
Remote Config [candidate] (566.962 µs) : 0, 567
Telemetry [baseline] (8.909 ms) : 0, 8909
Telemetry [candidate] (8.916 ms) : 0, 8916
Flare Poller [baseline] (3.607 ms) : 0, 3607
Flare Poller [candidate] (3.636 ms) : 0, 3636
section iast
crashtracking [baseline] (1.216 ms) : 0, 1216
crashtracking [candidate] (1.19 ms) : 0, 1190
BytebuddyAgent [baseline] (803.829 ms) : 0, 803829
BytebuddyAgent [candidate] (797.12 ms) : 0, 797120
AgentMeter [baseline] (11.646 ms) : 0, 11646
AgentMeter [candidate] (11.342 ms) : 0, 11342
GlobalTracer [baseline] (249.08 ms) : 0, 249080
GlobalTracer [candidate] (247.398 ms) : 0, 247398
IAST [baseline] (25.449 ms) : 0, 25449
IAST [candidate] (25.124 ms) : 0, 25124
AppSec [baseline] (26.725 ms) : 0, 26725
AppSec [candidate] (26.394 ms) : 0, 26394
Debugger [baseline] (63.473 ms) : 0, 63473
Debugger [candidate] (63.654 ms) : 0, 63654
Remote Config [baseline] (531.874 µs) : 0, 532
Remote Config [candidate] (527.722 µs) : 0, 528
Telemetry [baseline] (14.816 ms) : 0, 14816
Telemetry [candidate] (14.856 ms) : 0, 14856
Flare Poller [baseline] (4.928 ms) : 0, 4928
Flare Poller [candidate] (4.875 ms) : 0, 4875
section profiling
crashtracking [baseline] (1.168 ms) : 0, 1168
crashtracking [candidate] (1.183 ms) : 0, 1183
BytebuddyAgent [baseline] (683.299 ms) : 0, 683299
BytebuddyAgent [candidate] (685.944 ms) : 0, 685944
AgentMeter [baseline] (8.621 ms) : 0, 8621
AgentMeter [candidate] (8.653 ms) : 0, 8653
GlobalTracer [baseline] (215.099 ms) : 0, 215099
GlobalTracer [candidate] (216.618 ms) : 0, 216618
AppSec [baseline] (32.0 ms) : 0, 32000
AppSec [candidate] (32.052 ms) : 0, 32052
Debugger [baseline] (63.596 ms) : 0, 63596
Debugger [candidate] (63.838 ms) : 0, 63838
Remote Config [baseline] (588.626 µs) : 0, 589
Remote Config [candidate] (589.348 µs) : 0, 589
Telemetry [baseline] (9.071 ms) : 0, 9071
Telemetry [candidate] (9.826 ms) : 0, 9826
Flare Poller [baseline] (4.273 ms) : 0, 4273
Flare Poller [candidate] (3.521 ms) : 0, 3521
ProfilingAgent [baseline] (93.978 ms) : 0, 93978
ProfilingAgent [candidate] (94.479 ms) : 0, 94479
Profiling [baseline] (94.556 ms) : 0, 94556
Profiling [candidate] (95.036 ms) : 0, 95036
Loading

Load

Parameters

Baseline Candidate
baseline_or_candidate baseline candidate
git_branch master charles.yu/djm-0000/stop-storing-stage-accumulators
git_commit_date 1773169596 1773179914
git_commit_sha cc12228 4c54b6b
release_version 1.61.0-SNAPSHOT~cc122288e5 1.61.0-SNAPSHOT~4c54b6b488
See matching parameters
Baseline Candidate
application insecure-bank insecure-bank
ci_job_date 1773182116 1773182116
ci_job_id 1494902083 1494902083
ci_pipeline_id 101755727 101755727
cpu_model Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz
kernel_version Linux runner-zfyrx7zua-project-304-concurrent-2-qofyn5kg 6.8.0-1031-aws #33~22.04.1-Ubuntu SMP Thu Jun 26 14:22:30 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux Linux runner-zfyrx7zua-project-304-concurrent-2-qofyn5kg 6.8.0-1031-aws #33~22.04.1-Ubuntu SMP Thu Jun 26 14:22:30 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux

Summary

Found 0 performance improvements and 2 performance regressions! Performance is the same for 18 metrics, 16 unstable metrics.

scenario Δ mean agg_http_req_duration_p50 Δ mean agg_http_req_duration_p95 Δ mean throughput candidate mean agg_http_req_duration_p50 candidate mean agg_http_req_duration_p95 candidate mean throughput baseline mean agg_http_req_duration_p50 baseline mean agg_http_req_duration_p95 baseline mean throughput
scenario:load:insecure-bank:iast_GLOBAL:high_load worse
[+111.184µs; +205.668µs] or [+4.171%; +7.716%]		 unsure
[+74.763µs; +445.581µs] or [+0.966%; +5.760%]		 unstable
[-205.152op/s; +72.590op/s] or [-15.324%; +5.422%]		 2.824ms 7.996ms 1272.469op/s 2.665ms 7.736ms 1338.750op/s
scenario:load:petclinic:no_agent:high_load worse
[+0.459ms; +1.805ms] or [+2.795%; +11.003%]		 unstable
[+0.332ms; +3.303ms] or [+1.198%; +11.912%]		 unstable
[-45.070op/s; +16.008op/s] or [-16.400%; +5.825%]		 17.542ms 29.548ms 260.281op/s 16.410ms 27.730ms 274.812op/s
Request duration reports for petclinic 
gantt
    title petclinic - request duration [CI 0.99] : candidate=1.61.0-SNAPSHOT~4c54b6b488, baseline=1.61.0-SNAPSHOT~cc122288e5
    dateFormat X
    axisFormat %s
section baseline
no_agent (16.977 ms) : 16811, 17143
.   : milestone, 16977,
appsec (18.653 ms) : 18461, 18845
.   : milestone, 18653,
code_origins (17.905 ms) : 17726, 18085
.   : milestone, 17905,
iast (18.51 ms) : 18325, 18695
.   : milestone, 18510,
profiling (18.67 ms) : 18483, 18857
.   : milestone, 18670,
tracing (17.722 ms) : 17546, 17898
.   : milestone, 17722,
section candidate
no_agent (17.928 ms) : 17746, 18110
.   : milestone, 17928,
appsec (18.519 ms) : 18331, 18707
.   : milestone, 18519,
code_origins (18.122 ms) : 17943, 18302
.   : milestone, 18122,
iast (18.703 ms) : 18512, 18894
.   : milestone, 18703,
profiling (18.4 ms) : 18215, 18585
.   : milestone, 18400,
tracing (17.634 ms) : 17456, 17812
.   : milestone, 17634,
Loading
  • baseline results
Variant Request duration [CI 0.99] Δ no_agent
no_agent 16.977 ms [16.811 ms, 17.143 ms] -
appsec 18.653 ms [18.461 ms, 18.845 ms] 1.676 ms (9.9%)
code_origins 17.905 ms [17.726 ms, 18.085 ms] 928.534 µs (5.5%)
iast 18.51 ms [18.325 ms, 18.695 ms] 1.533 ms (9.0%)
profiling 18.67 ms [18.483 ms, 18.857 ms] 1.693 ms (10.0%)
tracing 17.722 ms [17.546 ms, 17.898 ms] 744.95 µs (4.4%)
  • candidate results
Variant Request duration [CI 0.99] Δ no_agent
no_agent 17.928 ms [17.746 ms, 18.11 ms] -
appsec 18.519 ms [18.331 ms, 18.707 ms] 591.162 µs (3.3%)
code_origins 18.122 ms [17.943 ms, 18.302 ms] 194.151 µs (1.1%)
iast 18.703 ms [18.512 ms, 18.894 ms] 774.918 µs (4.3%)
profiling 18.4 ms [18.215 ms, 18.585 ms] 472.285 µs (2.6%)
tracing 17.634 ms [17.456 ms, 17.812 ms] -293.718 µs (-1.6%)
Request duration reports for insecure-bank 
gantt
    title insecure-bank - request duration [CI 0.99] : candidate=1.61.0-SNAPSHOT~4c54b6b488, baseline=1.61.0-SNAPSHOT~cc122288e5
    dateFormat X
    axisFormat %s
section baseline
no_agent (1.179 ms) : 1168, 1191
.   : milestone, 1179,
iast (3.277 ms) : 3229, 3325
.   : milestone, 3277,
iast_FULL (5.878 ms) : 5819, 5937
.   : milestone, 5878,
iast_GLOBAL (3.422 ms) : 3375, 3469
.   : milestone, 3422,
profiling (2.104 ms) : 2084, 2123
.   : milestone, 2104,
tracing (1.811 ms) : 1795, 1827
.   : milestone, 1811,
section candidate
no_agent (1.184 ms) : 1174, 1195
.   : milestone, 1184,
iast (3.195 ms) : 3153, 3237
.   : milestone, 3195,
iast_FULL (5.73 ms) : 5674, 5787
.   : milestone, 5730,
iast_GLOBAL (3.604 ms) : 3543, 3664
.   : milestone, 3604,
profiling (2.079 ms) : 2061, 2098
.   : milestone, 2079,
tracing (1.759 ms) : 1744, 1774
.   : milestone, 1759,
Loading
  • baseline results
Variant Request duration [CI 0.99] Δ no_agent
no_agent 1.179 ms [1.168 ms, 1.191 ms] -
iast 3.277 ms [3.229 ms, 3.325 ms] 2.098 ms (177.9%)
iast_FULL 5.878 ms [5.819 ms, 5.937 ms] 4.699 ms (398.5%)
iast_GLOBAL 3.422 ms [3.375 ms, 3.469 ms] 2.243 ms (190.2%)
profiling 2.104 ms [2.084 ms, 2.123 ms] 924.583 µs (78.4%)
tracing 1.811 ms [1.795 ms, 1.827 ms] 632.202 µs (53.6%)
  • candidate results
Variant Request duration [CI 0.99] Δ no_agent
no_agent 1.184 ms [1.174 ms, 1.195 ms] -
iast 3.195 ms [3.153 ms, 3.237 ms] 2.01 ms (169.7%)
iast_FULL 5.73 ms [5.674 ms, 5.787 ms] 4.546 ms (383.8%)
iast_GLOBAL 3.604 ms [3.543 ms, 3.664 ms] 2.419 ms (204.3%)
profiling 2.079 ms [2.061 ms, 2.098 ms] 895.01 µs (75.6%)
tracing 1.759 ms [1.744 ms, 1.774 ms] 574.546 µs (48.5%)

Dacapo

Parameters

Baseline Candidate
baseline_or_candidate baseline candidate
git_branch master charles.yu/djm-0000/stop-storing-stage-accumulators
git_commit_date 1773169596 1773179914
git_commit_sha cc12228 4c54b6b
release_version 1.61.0-SNAPSHOT~cc122288e5 1.61.0-SNAPSHOT~4c54b6b488
See matching parameters
Baseline Candidate
application biojava biojava
ci_job_date 1773181858 1773181858
ci_job_id 1494902086 1494902086
ci_pipeline_id 101755727 101755727
cpu_model Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz Intel(R) Xeon(R) Platinum 8259CL CPU @ 2.50GHz
kernel_version Linux runner-zfyrx7zua-project-304-concurrent-1-hxjqo39g 6.8.0-1031-aws #33~22.04.1-Ubuntu SMP Thu Jun 26 14:22:30 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux Linux runner-zfyrx7zua-project-304-concurrent-1-hxjqo39g 6.8.0-1031-aws #33~22.04.1-Ubuntu SMP Thu Jun 26 14:22:30 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux

Summary

Found 1 performance improvements and 0 performance regressions! Performance is the same for 11 metrics, 0 unstable metrics.

scenario Δ mean execution_time candidate mean execution_time baseline mean execution_time
scenario:dacapo:tomcat:appsec better
[-1.458ms; -1.110ms] or [-38.130%; -29.021%]		 2.540ms 3.825ms
Execution time for tomcat 
gantt
    title tomcat - execution time [CI 0.99] : candidate=1.61.0-SNAPSHOT~4c54b6b488, baseline=1.61.0-SNAPSHOT~cc122288e5
    dateFormat X
    axisFormat %s
section baseline
no_agent (1.479 ms) : 1467, 1491
.   : milestone, 1479,
appsec (3.825 ms) : 3603, 4046
.   : milestone, 3825,
iast (2.27 ms) : 2200, 2340
.   : milestone, 2270,
iast_GLOBAL (2.305 ms) : 2235, 2375
.   : milestone, 2305,
profiling (2.096 ms) : 2040, 2152
.   : milestone, 2096,
tracing (2.071 ms) : 2016, 2125
.   : milestone, 2071,
section candidate
no_agent (1.479 ms) : 1467, 1490
.   : milestone, 1479,
appsec (2.54 ms) : 2484, 2597
.   : milestone, 2540,
iast (2.278 ms) : 2207, 2349
.   : milestone, 2278,
iast_GLOBAL (2.309 ms) : 2238, 2380
.   : milestone, 2309,
profiling (2.11 ms) : 2053, 2167
.   : milestone, 2110,
tracing (2.082 ms) : 2027, 2136
.   : milestone, 2082,
Loading
  • baseline results
Variant Execution Time [CI 0.99] Δ no_agent
no_agent 1.479 ms [1.467 ms, 1.491 ms] -
appsec 3.825 ms [3.603 ms, 4.046 ms] 2.346 ms (158.6%)
iast 2.27 ms [2.2 ms, 2.34 ms] 791.053 µs (53.5%)
iast_GLOBAL 2.305 ms [2.235 ms, 2.375 ms] 826.138 µs (55.9%)
profiling 2.096 ms [2.04 ms, 2.152 ms] 617.137 µs (41.7%)
tracing 2.071 ms [2.016 ms, 2.125 ms] 591.885 µs (40.0%)
  • candidate results
Variant Execution Time [CI 0.99] Δ no_agent
no_agent 1.479 ms [1.467 ms, 1.49 ms] -
appsec 2.54 ms [2.484 ms, 2.597 ms] 1.061 ms (71.8%)
iast 2.278 ms [2.207 ms, 2.349 ms] 799.081 µs (54.0%)
iast_GLOBAL 2.309 ms [2.238 ms, 2.38 ms] 830.098 µs (56.1%)
profiling 2.11 ms [2.053 ms, 2.167 ms] 631.169 µs (42.7%)
tracing 2.082 ms [2.027 ms, 2.136 ms] 602.829 µs (40.8%)
Execution time for biojava 
gantt
    title biojava - execution time [CI 0.99] : candidate=1.61.0-SNAPSHOT~4c54b6b488, baseline=1.61.0-SNAPSHOT~cc122288e5
    dateFormat X
    axisFormat %s
section baseline
no_agent (15.64 s) : 15640000, 15640000
.   : milestone, 15640000,
appsec (14.909 s) : 14909000, 14909000
.   : milestone, 14909000,
iast (18.227 s) : 18227000, 18227000
.   : milestone, 18227000,
iast_GLOBAL (17.529 s) : 17529000, 17529000
.   : milestone, 17529000,
profiling (14.967 s) : 14967000, 14967000
.   : milestone, 14967000,
tracing (15.086 s) : 15086000, 15086000
.   : milestone, 15086000,
section candidate
no_agent (14.984 s) : 14984000, 14984000
.   : milestone, 14984000,
appsec (14.837 s) : 14837000, 14837000
.   : milestone, 14837000,
iast (18.206 s) : 18206000, 18206000
.   : milestone, 18206000,
iast_GLOBAL (18.032 s) : 18032000, 18032000
.   : milestone, 18032000,
profiling (15.502 s) : 15502000, 15502000
.   : milestone, 15502000,
tracing (15.091 s) : 15091000, 15091000
.   : milestone, 15091000,
Loading
  • baseline results
Variant Execution Time [CI 0.99] Δ no_agent
no_agent 15.64 s [15.64 s, 15.64 s] -
appsec 14.909 s [14.909 s, 14.909 s] -731.0 ms (-4.7%)
iast 18.227 s [18.227 s, 18.227 s] 2.587 s (16.5%)
iast_GLOBAL 17.529 s [17.529 s, 17.529 s] 1.889 s (12.1%)
profiling 14.967 s [14.967 s, 14.967 s] -673.0 ms (-4.3%)
tracing 15.086 s [15.086 s, 15.086 s] -554.0 ms (-3.5%)
  • candidate results
Variant Execution Time [CI 0.99] Δ no_agent
no_agent 14.984 s [14.984 s, 14.984 s] -
appsec 14.837 s [14.837 s, 14.837 s] -147.0 ms (-1.0%)
iast 18.206 s [18.206 s, 18.206 s] 3.222 s (21.5%)
iast_GLOBAL 18.032 s [18.032 s, 18.032 s] 3.048 s (20.3%)
profiling 15.502 s [15.502 s, 15.502 s] 518.0 ms (3.5%)
tracing 15.091 s [15.091 s, 15.091 s] 107.0 ms (0.7%)

@charlesmyu charlesmyu added the type: enhancement Enhancements and improvements label Feb 20, 2026
@charlesmyu charlesmyu force-pushed the charles.yu/djm-0000/fix-spark-plan-metrics branch from e413d1d to 89df516 Compare February 23, 2026 21:52
@charlesmyu charlesmyu force-pushed the charles.yu/djm-0000/stop-storing-stage-accumulators branch from 9594586 to cb6648d Compare February 23, 2026 21:52
@charlesmyu charlesmyu force-pushed the charles.yu/djm-0000/fix-spark-plan-metrics branch from 89df516 to 8651527 Compare February 24, 2026 20:06
@charlesmyu charlesmyu force-pushed the charles.yu/djm-0000/stop-storing-stage-accumulators branch from cb6648d to 780cd6f Compare February 24, 2026 20:06
@charlesmyu Graphite App
Copy link
Contributor Author

charlesmyu commented Feb 24, 2026 •
edited
Loading

This stack of pull requests is managed by Graphite. Learn more about stacking.

@charlesmyu charlesmyu force-pushed the charles.yu/djm-0000/stop-storing-stage-accumulators branch from 780cd6f to a43ff79 Compare March 4, 2026 15:37
@charlesmyu charlesmyu force-pushed the charles.yu/djm-0000/fix-spark-plan-metrics branch from 8651527 to 7e4b7de Compare March 4, 2026 15:37
@charlesmyu charlesmyu marked this pull request as ready for review March 4, 2026 15:49
@charlesmyu charlesmyu requested a review from a team as a code owner March 4, 2026 15:49
pawel-big-lebowski
pawel-big-lebowski approved these changes Mar 5, 2026
View reviewed changes
Copy link
Contributor

@pawel-big-lebowski pawel-big-lebowski left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It’s always better to keep the primary types in memory. Great enhancement.

Base automatically changed from charles.yu/djm-0000/fix-spark-plan-metrics to master March 10, 2026 19:06
@gh-worker-dd-mergequeue-cf854d gh-worker-dd-mergequeue-cf854d bot requested a review from a team as a code owner March 10, 2026 19:06
@charlesmyu charlesmyu force-pushed the charles.yu/djm-0000/stop-storing-stage-accumulators branch from a43ff79 to b1b3e4c Compare March 10, 2026 19:17
mhlidd
mhlidd approved these changes Mar 10, 2026
View reviewed changes
Copy link
Contributor

@mhlidd mhlidd left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM from a Java POV

...k-common/src/main/java/datadog/trace/instrumentation/spark/AbstractDatadogSparkListener.java Outdated
// stage ID -> accumulator ID? Put this behind some FF
private final Map<Long, SparkSQLUtils.AccumulatorWithStage> accumulators =
new RemoveEldestHashMap<>(MAX_ACCUMULATOR_SIZE);
private final Map<Long, Integer> acc2stage = new HashMap<>();
Copy link
Contributor

@mhlidd mhlidd Mar 10, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
private final Map<Long, Integer> acc2stage = new HashMap<>();
private final Map<Long, Integer> accumulatorToStageID = new HashMap<>();

nit: I just find this easier to understand

Copy link
Contributor Author

@charlesmyu charlesmyu Mar 10, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

85b414f (this PR)

...tion/spark/spark-common/src/main/java/datadog/trace/instrumentation/spark/SparkSQLUtils.java Outdated

private static Set<Integer> stageIdsForPlan(
SparkPlanInfo info, Map<Long, AccumulatorWithStage> accumulators) {
private static Set<Integer> stageIdsForPlan(SparkPlanInfo info, Map<Long, Integer> accumulators) {
Copy link
Contributor

@mhlidd mhlidd Mar 10, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
private static Set<Integer> stageIdsForPlan(SparkPlanInfo info, Map<Long, Integer> accumulators) {
private static Set<Integer> stageIdsForPlan(SparkPlanInfo info, Map<Long, Integer> accumulatorToStageID) {

nit: Same as above

@charlesmyu charlesmyu enabled auto-merge March 10, 2026 21:59
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@pawel-big-lebowski pawel-big-lebowski pawel-big-lebowski approved these changes

@mhlidd mhlidd mhlidd approved these changes

Assignees

No one assigned

Labels

inst: apache spark Apache Spark instrumentation type: enhancement Enhancements and improvements

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@charlesmyu @pawel-big-lebowski @mhlidd