Repository navigation
docs: operator runbook for maintenance scheduling - #342
Merged
Merged
Conversation
Eleven topic documents describe individual mechanisms but none answers the operator's questions in order: is backlog real, is anything scheduled, is the running thing working, why is nothing scheduled, why does compaction never commit, what to do when an executor died holding ownership. This runbook does, naming the exact endpoint, metric or etcd key for each step and what the answer means. The leading rule is that progress alone decides whether an execution is stuck: leases, heartbeats, Running state, retry countdowns, Ready pods and elapsed time are not evidence either way. Recovery is automatic and fenced; the runbook says when to wait and when to escalate, and never to delete ownership keys by hand. Numbers (15 s retry loop, 15-failure attention threshold, 120 s fence, 30 s lease TTL, 3-failure cooldown, 256/4096 read thresholds) were checked against source. Known gaps are listed so they are not debugged as faults. Refs lance-format#333. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
17 of 18 tasks
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Part of #333 (P0.6, backlog O2/O3).
What
docs/operations/maintenance-runbook.md: symptom-driven guide for WAL merge / compaction / index / repair scheduling. Sections: where to look (table of question → endpoint/metric/key), backlog growing, deciding whether a running execution is stuck, demand but nothing scheduled (seven admission reasons, cheapest first), compaction never commits on a hot table, executor died holding ownership, draining, known gaps.Leading rule: progress decides whether something is stuck; leases, heartbeats, Running state, retry countdowns, Ready pods and elapsed time do not. Recovery is fenced and automatic; never delete ownership keys by hand.
Uses the metrics from #339 and the typed
/merge-progressfrom #341. All numeric claims (15 s retry loop, 15-failure attention, 120 s fence, 30 s lease TTL, 3-failure cooldown, 256/4096 thresholds) verified against source.Docs only;
typosclean.🤖 Generated with Claude Code