Skip to content

Improve troubleshoot analyzers: select/release inference, cache cross-diagnosis, richer reports and file discovery - #17

Merged
mouxinqq merged 2 commits into
developfrom
codex/modify-troubleshoot-for-skills-alignment-xpkhs2
Apr 13, 2026
Merged

mouxinqq merged 2 commits into
developfrom
codex/modify-troubleshoot-for-skills-alignment-xpkhs2

Conversation

@mouxinqq

Copy link
Copy Markdown
Owner

Motivation

  • Improve accuracy and usability of the troubleshooting scripts by better correlating select/release events, inferring worker types for releases/tokens, surfacing cross-diagnostics for cache, and generating richer detail files for reports.

Description

  • Enhance log_parser.match_select_release and related helpers to normalize worker URL keys, infer release/token worker types from nearby selects, use worker-FIFO matching, and produce richer per-worker/type summaries and profiles.
  • Add counter last-state analysis and new regexes (select count / release worker count) in load.py, plus loosened diagnostics thresholds and additional diagnostics messages.
  • Add cross-diagnosis logic for cache (session stickiness × hitRatio × fallback × evictions) and wire it into cache analyzer/reporting; split summary/detail outputs and add many detail files (cache_diagnosis, cache_*, load_counter_state, errors_topn, latency_diagnoses, etc.).
  • Improve errors analyzer with IMPACT_RULES, classify_impact, URL extraction, and surface YAML-missing info; trace analyzer now invokes select/release checks and reports unmatched selects/releases.
  • Health analyzer: include register events in timelines and compute detection latency metric; latency and load report formatters include metric notes and link to detail outputs.
  • Fix chart.render_table alignment/separator generation.
  • Improve troubleshoot.py log-file discovery (Path-based, fuzzy matching), add more detailed report assembly and save many detail files, and add user-friendly messages for missing files.
  • Add helper functions and self-test case in log_parser.py to validate token-release inference behavior.

Testing

  • Ran python3 scripts/log_parser.py self-test which includes select/release token-type inference checks and the added sample case, and the self-tests passed.
  • Performed smoke runs of the troubleshoot orchestration against sample logs to verify that reports and new detail files (detail/...) are produced without exceptions.
  • No unit test framework changes; all automated inline self-tests and script-level smoke checks completed successfully.

Codex Task

@codeinputmachine

Copy link
Copy Markdown

Hey @mouxinqq, Noticed this PR has a merge conflict. I'm building an online editor where you can fix it directly in the browser, no local setup needed. Free to give it a shot https://codeinput.com/products/merge-conflicts Would love your feedback.

@mouxinqq
mouxinqq merged commit bc2d4f2 into develop Apr 13, 2026
1 of 2 checks passed

This branch had an error being deployed

1 failed deployment
Metax_ci b45b91cf Deployed Apr 13, 2026 by mouxinqq via Trigger Jenkins for PR #26
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants