Add hybrid profile support, resolution matching, and system metrics#55
Merged
izzet merged 12 commits intollnl:developfrom Mar 21, 2026
Merged
Conversation
- Add comm and device layers to AI Logging preset with derived_metrics and layer_deps; broaden compute and data_loader conditions - Retain min/max stat columns (time_min/max, size_min/max, offset_min/max) through profile standardization and coalescing - Implement profile bucket expansion in Analyzer for finer-than-bucket analysis granularity with uniform and weighted distribution strategies - Move granularity validation and expansion from DFTracer reader to Analyzer so resolution matching is format-agnostic - Add profile_distribution and profile_time_granularity to AnalyzerConfig - Add 15 new tests covering edge cases from production data: dur without dur_sum, missing fhash, non-POSIX categories, stat column preservation, coalescing min/max semantics, expansion, profile-only traces, and custom profile_time_granularity
… adding missing size_bin columns
…ndling of NA values
…rt-resolution-match
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## develop #55 +/- ##
===========================================
- Coverage 62.52% 58.45% -4.08%
===========================================
Files 26 27 +1
Lines 2231 2903 +672
===========================================
+ Hits 1395 1697 +302
- Misses 836 1206 +370 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
…stem metrics tests
hariharan-devarajan
approved these changes
Mar 20, 2026
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
ph="C"counter events into a canonical profile schema, enabling analysis of selectively-aggregated traces alongside normal duration tracescat="sys"counter events (CPU utilization, memory stats) into per-time_rangesystem metrics and join them onto flat views for I/O bottleneck correlationKey changes
dftracer.py: NewTYPE_PROFILE(6) andTYPE_SYSTEM(7) event types with_standardize_profile_partitionand_standardize_system_partitionfor canonical schema mappinganalyzer.py:_expand_profile_buckets,_validate_and_expand_profiles,_reconcile_hlm,_compute_hybrid_hlmfor the full resolution matching pipeline; system metrics joined in_process_flat_viewtypes.py:ReadTraceResultextended withprofiles,profile_time_granularity, andsystem_metricsconfig.py: New config optionsprofile_distribution(uniform/weighted) andprofile_time_granularityTest plan
tests/test_hybrid_profiles.py: 21 tests covering profile parsing, coalescing, expansion (uniform + weighted), per-layer reconciliation, stat column preservation, edge casestests/test_system_metrics.py: 16 tests covering system event parsing, standardization, mixed traces, variable time granularity