fix(cachewd): decouple scheduler context from signal context#312
Merged
Conversation
joshfriend
force-pushed
the
jfriend/fix-scheduler-shutdown-context
branch
from
efc4687 to
692f2e3
Compare
joshfriend
force-pushed
the
jfriend/fix-scheduler-shutdown-context
branch
from
692f2e3 to
8a6ea50
Compare
The graceful shutdown PR (#307) passes the signal-notified context to the scheduler, so SIGTERM immediately kills all workers even though server.Shutdown keeps HTTP handlers alive for up to 150s. In-flight jobs (snapshots, repacks, fetches) get context-cancelled mid-execution. Give the scheduler its own context via context.WithoutCancel (same pattern already used for the HTTP server BaseContext) and cancel it after server.Shutdown completes. This lets workers finish current jobs during the drain window. The scheduler drain is bounded to 10s so long-running jobs don't block process exit past the pod's terminationGracePeriodSeconds.
joshfriend
force-pushed
the
jfriend/fix-scheduler-shutdown-context
branch
from
8a6ea50 to
e913611
Compare
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 692f2e3fc7
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
worstell
approved these changes
May 15, 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.
The graceful shutdown PR (#307) passes the signal-notified context to the scheduler, so SIGTERM immediately kills all workers even though
server.Shutdownkeeps HTTP handlers alive for up to 150s. In-flight jobs (snapshots, repacks, fetches) get context-cancelled mid-execution.Gives the scheduler its own context via
context.WithoutCancel(same pattern already used for the HTTP serverBaseContext) and cancels it afterserver.Shutdowncompletes. Workers finish their current jobs during the drain window instead of dying instantly. The drain is bounded to 10s so long-running jobs don't block process exit past the pod'sterminationGracePeriodSeconds.Production evidence from today's rolling deploys:
Worker terminatedlog lines per podtar failed: signal: killed,context canceledcash-serversnapshot killed 50s in,ios-registermirror snapshot killed 5m31s inShutdown order after this change:
server.Shutdowndrains in-flight HTTP requests (up to 150s)