Fix wrong interval for OOM protection

[Jackie-Jiang]

• Track sleep time separately for query resource tracking and workload resource tracking
• Fix the super frequent workload resource tracking (change default interval from 1ms to 100ms)
• Do not track workload resource when WorkloadBudgetManager is disabled

Behavior Change

• Default interval for workload resource tracking is increased from 1ms to 100ms (most likely a typo in the original PR [Query Resource Isolation] WorkloadBudgetManager and Host enforcement #15798 )

[Jackie-Jiang]

@vvivekiyer Can you please verify if the default 1ms workload tracking interval is a typo? It doesn't make much sense to track a 1 minute granularity budget with 1ms interval

[Copilot]

Pull request overview

This PR fixes a critical performance issue with workload resource tracking interval and refactors the resource aggregation logic to track query and workload resources independently.

Changes:

• Changed default workload sleep time from 1ms to 100ms (fixing likely typo from PR #15798)
• Refactored WatcherTask to track separate sleep times for query vs workload resource aggregation
• Made workload resource aggregator creation conditional on WorkloadBudgetManager being enabled

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 4 comments.

File	Description
pinot-spi/src/main/java/org/apache/pinot/spi/utils/CommonConstants.java	Changed DEFAULT_WORKLOAD_SLEEP_TIME_MS from 1ms to 100ms; minor line formatting
pinot-spi/src/main/java/org/apache/pinot/spi/accounting/WorkloadBudgetManager.java	Renamed _isEnabled to _enabled; added public isEnabled() method
pinot-core/src/main/java/org/apache/pinot/core/accounting/WorkloadResourceAggregator.java	Updated constructor to take WorkloadBudgetManager parameter; added assertion check
pinot-core/src/main/java/org/apache/pinot/core/accounting/ResourceUsageAccountantFactory.java	Major refactoring: separate timing logic for query/workload aggregation, conditional workload aggregator creation, removed EnumMap dispatch pattern

[codecov-commenter]

Codecov Report

❌ Patch coverage is 1.69492% with 58 lines in your changes missing coverage. Please review.
✅ Project coverage is 63.20%. Comparing base (739043c) to head (faf67b2).
⚠️ Report is 5 commits behind head on master.

Files with missing lines	Patch %	Lines
...ore/accounting/ResourceUsageAccountantFactory.java	0.00%	44 Missing ⚠️
...he/pinot/spi/accounting/WorkloadBudgetManager.java	8.33%	6 Missing and 5 partials ⚠️
...ot/core/accounting/WorkloadResourceAggregator.java	0.00%	3 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##             master   #17753      +/-   ##
============================================
- Coverage     63.22%   63.20%   -0.03%     
  Complexity     1454     1454              
============================================
  Files          3176     3176              
  Lines        191002   191038      +36     
  Branches      29204    29216      +12     
============================================
- Hits         120763   120740      -23     
- Misses        60818    60876      +58     
- Partials       9421     9422       +1     

Flag	Coverage Δ
custom-integration1	100.00% <ø> (ø)
integration	100.00% <ø> (ø)
integration1	100.00% <ø> (ø)
integration2	0.00% <ø> (ø)
java-11	63.15% <1.69%> (-0.03%)	⬇️
java-21	63.18% <1.69%> (-0.01%)	⬇️
temurin	63.20% <1.69%> (-0.03%)	⬇️
unittests	63.19% <1.69%> (-0.03%)	⬇️
unittests1	55.58% <1.69%> (+0.02%)	⬆️
unittests2	34.06% <0.00%> (-0.05%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 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.