test(amber): cover workflow lifecycle cleanup

[aglinxinyuan]

What changes were proposed in this PR?

Adds WorkflowLifecycleManagerSpec for the lifecycle manager's externally observable cleanup behavior.

Scenario	Observable contract
Multiple users	Cleanup waits until the final user disconnects.
Workflow state	A RUNNING update cancels cleanup; a terminal update schedules it again.
Repeated terminal updates	A later update replaces the older deadline.

The suite uses an isolated Pekko actor system and restores AmberRuntime's shared actor-system reference after the suite, so other Amber suites retain their runtime setup.

The assertions were mutation-tested against temporary production changes, all reverted before commit:

Temporary mutation	Result
userCount == 0 to userCount < 0	1 test failed
status == RUNNING to status != RUNNING	2 tests failed
Remove cancellation of an active cleanup deadline	1 test failed

Any related issues, documentation, discussions?

Closes #7150

How was this PR tested?

"C:/Program Files (x86)/sbt/bin/sbt" -java-home "C:/Users/linxi/.jdks/jbr-17.0.14" "WorkflowExecutionService/testOnly org.apache.texera.web.WorkflowLifecycleManagerSpec"

Tests: succeeded 3, failed 0

"C:/Program Files (x86)/sbt/bin/sbt" -java-home "C:/Users/linxi/.jdks/jbr-17.0.14" "WorkflowExecutionService/Test/scalafmtCheck" "WorkflowExecutionService/Test/scalafix --check"

Both checks completed successfully.

Was this PR authored or co-authored using generative AI tooling?

Generated-by: Codex (GPT-5)

[github-actions]

Automated Reviewer Suggestions

Based on the git blame history of the changed files, we recommend the following reviewers:

• No candidates found from git blame history.

[github-actions]

⚠️ Benchmark changes need a look

🟢 0 better · 🔴 6 worse · ⚪ 9 noise (<±5%) · 0 without baseline

Compared against main 91ed98d benchmarked on this same runner, so the delta is largely free of cross-runner hardware noise. The "7d avg" column still reflects the gh-pages dashboard. Treat <±5% as noise unless repeated.

Dashboard · Run

	config	throughput	MB/s	latency	max Δ latest / 7d
🔴	bs=10 sw=10 sl=64	565	0.345	17,086/25,570/25,570 us	🔴 +11.7% / 🔴 +57.9%
🔴	bs=100 sw=10 sl=64	1,190	0.727	80,445/111,486/111,486 us	🔴 +13.5% / 🟢 +23.4%
🔴	bs=1000 sw=10 sl=64	1,416	0.864	698,985/817,941/817,941 us	🔴 +9.6% / 🟢 +42.7%

Baseline details

Latest main 91ed98d from same runner

config	metric	PR	latest main	7d avg	Δ latest	Δ 7d
bs=10 sw=10 sl=64	throughput	565 tuples/sec	577 tuples/sec	765.25 tuples/sec	-2.1%	-26.2%
bs=10 sw=10 sl=64	MB/s	0.345 MB/s	0.352 MB/s	0.467 MB/s	-2.0%	-26.1%
bs=10 sw=10 sl=64	p50	17,086 us	16,633 us	12,703 us	+2.7%	+34.5%
bs=10 sw=10 sl=64	p95	25,570 us	22,893 us	16,199 us	+11.7%	+57.9%
bs=10 sw=10 sl=64	p99	25,570 us	22,893 us	19,557 us	+11.7%	+30.7%
bs=100 sw=10 sl=64	throughput	1,190 tuples/sec	1,243 tuples/sec	965.22 tuples/sec	-4.3%	+23.3%
bs=100 sw=10 sl=64	MB/s	0.727 MB/s	0.759 MB/s	0.589 MB/s	-4.2%	+23.4%
bs=100 sw=10 sl=64	p50	80,445 us	77,795 us	103,830 us	+3.4%	-22.5%
bs=100 sw=10 sl=64	p95	111,486 us	98,235 us	110,778 us	+13.5%	+0.6%
bs=100 sw=10 sl=64	p99	111,486 us	98,235 us	118,713 us	+13.5%	-6.1%
bs=1000 sw=10 sl=64	throughput	1,416 tuples/sec	1,446 tuples/sec	992.33 tuples/sec	-2.1%	+42.7%
bs=1000 sw=10 sl=64	MB/s	0.864 MB/s	0.883 MB/s	0.606 MB/s	-2.2%	+42.7%
bs=1000 sw=10 sl=64	p50	698,985 us	691,759 us	1,016,959 us	+1.0%	-31.3%
bs=1000 sw=10 sl=64	p95	817,941 us	746,141 us	1,062,501 us	+9.6%	-23.0%
bs=1000 sw=10 sl=64	p99	817,941 us	746,141 us	1,082,611 us	+9.6%	-24.4%

Raw CSV

config_idx,batch_size,schema_width,string_len,num_batches,total_ms,total_tuples,total_bytes,tuples_per_sec,mb_per_sec,lat_p50_us,lat_p95_us,lat_p99_us
0,10,10,64,20,354.22,200,128000,565,0.345,17085.53,25569.89,25569.89
1,100,10,64,20,1680.00,2000,1280000,1190,0.727,80445.13,111485.97,111485.97
2,1000,10,64,20,14122.59,20000,12800000,1416,0.864,698984.87,817941.21,817941.21

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 80.53%. Comparing base (91ed98d) to head (784191b).
⚠️ Report is 12 commits behind head on main.

Additional details and impacted files

@@             Coverage Diff              @@
##               main    #7151      +/-   ##
============================================
+ Coverage     79.58%   80.53%   +0.94%     
- Complexity     3837     3900      +63     
============================================
  Files          1160     1160              
  Lines         46188    46188              
  Branches       5145     5145              
============================================
+ Hits          36761    37199     +438     
+ Misses         7795     7311     -484     
- Partials       1632     1678      +46     

Flag	Coverage Δ		*Carryforward flag
access-control-service	70.00% <ø> (ø)		Carriedforward from 4f24bb6
agent-service	77.42% <ø> (ø)		Carriedforward from 4f24bb6
amber	75.67% <ø> (+2.57%)	⬆️
computing-unit-managing-service	20.49% <ø> (ø)		Carriedforward from 4f24bb6
config-service	65.97% <ø> (ø)		Carriedforward from 4f24bb6
file-service	66.80% <ø> (ø)		Carriedforward from 4f24bb6
frontend	83.08% <ø> (ø)		Carriedforward from 4f24bb6
notebook-migration-service	78.89% <ø> (ø)		Carriedforward from 4f24bb6
pyamber	97.31% <ø> (ø)		Carriedforward from 4f24bb6
workflow-compiling-service	26.31% <ø> (ø)		Carriedforward from 4f24bb6

*This pull request uses carry forward flags. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[mengw15]

LGTM

[Copilot]

Pull request overview

Adds a new Amber unit test suite to validate WorkflowLifecycleManager’s externally observable cleanup contract (user-count gating, cancellation while RUNNING, and deadline refresh on repeated terminal updates), using a suite-local Pekko ActorSystem wired into AmberRuntime for scheduler-driven assertions.

Changes:

• Introduces WorkflowLifecycleManagerSpec with three scenarios covering user-count tracking and cleanup scheduling behavior.
• Uses ExecutionStateStore state updates to drive the lifecycle manager’s state-change subscription path.
• Temporarily swaps AmberRuntime’s global actor-system reference for deterministic scheduling during the suite, then restores it.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.