PostHog / PostHog/posthog

Scout reaper: run during planning so a wedged lane recovers within a tick, not a full interval

Open
#65,033 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
39.9k
Forks
3.4k
Avg merge
6h 51m
Merged PRs (30d)
232

Description

Background

PR #65028 restored _self_heal_stale_runs to reap orphaned scout runs. The reaper sits on the dispatch path (inside arun_signals_scout), running just before the single-flight guard.

Gap

Because the reap only runs when a child workflow is actually dispatched, an orphaned lane isn't checked until the config is next due — not within the cutoff or the 30-min coordinator tick. Compounding this, _stamp_dispatched_runs advances last_run_at even when the dispatched child later skips on the running guard, and _overdue_seconds then suppresses the lane for its full run_interval_minutes.

Net effect: recovery is bounded by run_interval_minutes, not by STALE_RUN_CUTOFF_S.

  • Default 3h scouts: recovers within a couple of hours — fine.
  • Daily general scout: up to ~a day of extra suppression after a crash (one missed daily run).
  • Pathological max-interval configs: longer.

This is still a bounded recovery (the permanent wedge — the original incident — is fixed), but the daily default lane this PR is meant to protect can still miss a run.

Proposed fix (either or both)

  1. Run the reaper during planning (_collect_planned_runs in scout_coordinator.py) so any orphan is reaped within a tick regardless of due-ness.
  2. Do not advance last_run_at when a dispatch resolves to a skip on the running guard, so a wedged lane stays due and re-dispatches next tick instead of being suppressed for a full interval.

Source

Raised by the codex reviewer (P2) on PR #65028: https://github.com/PostHog/posthog/pull/65028#discussion_r3446970733
Also listed as an intended fast-follow in that PR's agent context.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start in scout_coordinator.py at _collect_planned_runs, then trace arun_signals_scout, _stamp_dispatched_runs, and _overdue_seconds. Compare the two proposed fixes and verify how planning, the running guard, and last_run_at interact. Done means an orphaned lane is reaped within a coordinator tick and a skipped dispatch does not suppress the lane for a full interval.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
backend
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
48/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.