Fix CSP memory overhead sizing with host off-heap limit - #2138
Merged
Merged
Conversation
Use a single effective predicate so CSP platforms retain budget-aware overhead while non-CSP host off-heap sizing and explicit values remain unchanged. Signed-off-by: Partho Sarthi <psarthi@nvidia.com>
Document why CSPs must stay on budget-aware overhead sizing and why the protected overhead floor differs between the normal and specialized paths. Signed-off-by: Partho Sarthi <psarthi@nvidia.com>
parthosa
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
There was a discrepancy in how AutoTuner handled
spark.rapids.memory.host.offHeapLimit.enabledon CSP platforms: it skipped the specialized sizing calculation but finalized memory as if that calculation had run. This left only the JVM minimum instead of budget-aware executor overhead. The change uses one effective predicate for both decisions while preserving explicit overhead values and existing non-CSP behavior.Fixes #2135.
Test plan
QualificationAutoTunerSuite: 55 tests passed.ProfilingAutoTunerSuiteV2: 71 tests passed.Scope
This change does not modify
maxBytesInFlighthandling.Post-deploy Monitoring & Validation
No additional operational monitoring is required. The change affects offline AutoTuner recommendations and is covered by the Qualification and Profiling suites above.