Skip to content

refactor: split runtime_trace cfg impls and report trace periodically - #546

Draft
gnolong wants to merge 2 commits into
mainfrom
fix/runtime-trace-periodic-report
Draft

refactor: split runtime_trace cfg impls and report trace periodically#546
gnolong wants to merge 2 commits into
mainfrom
fix/runtime-trace-periodic-report

Conversation

@gnolong

@gnolong gnolong commented Jul 22, 2026

Copy link
Copy Markdown
Contributor

No description provided.

gnolong added 2 commits July 22, 2026 03:14
Address post-merge review comments on #530:

- Split runtime_trace.rs into per-feature modules (traced/noop) selected
  at module level, so each method is defined once per implementation
  instead of interleaving feature/not(feature) variants in one file.
- Add RuntimeTraceMonitor to the periodic monitor flush loop: the runtime
  trace summary is now dumped every pipeline.checkpoint_interval_secs,
  plus a final dump on shutdown, making it usable for long-running CDC
  tasks. The finish-time-only dump in start_task is removed since the
  shutdown flush covers it.
- Add a structured snapshot_global() API (always marker-aggregated) and
  export runtime_trace_* per-marker gauges via Prometheus when both
  metrics and tracing features are enabled.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant