Repository navigation
Fail sessions that run far past their time budget - #55
Open
josephschorr wants to merge 2 commits into
Open
josephschorr wants to merge 2 commits into
josephschorr wants to merge 2 commits into
Conversation
The operator never read status.runDuration, so a runner wedged in a call that never returned could hold its session in Running indefinitely, with reported run-time climbing far past the budget and the runner's own between-turn budget check never reached. reconcileRunDuration is a backstop for that case. A Running session whose runner-reported run-time exceeds the effective maxDuration by more than a 10 minute grace is failed with BudgetExceeded via a new signed lifecycle event (RunDurationExceeded), and the ordinary Failed reap then tears down its runner. The grace keeps the operator from racing a healthy runner, which enforces the cap itself. This also makes the admission webhook's description of runDuration accurate.
|
The latest updates on your projects. Learn more about Vercel for GitHub. 1 Skipped Deployment
|
The backstop reads runner-written status.runDuration, so a runner that under-reports or freezes it is never caught. Say so where the backstop and the field are documented, and drop an unrelated reindent.
This branch has not been deployed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A session could stay in Running indefinitely if its runner got stuck — for example, waiting on a model provider that stopped responding — because nothing outside the runner enforced the class's
maxDurationbudget.The operator now acts as a backstop: a running session whose recorded run-time goes more than 10 minutes past its
maxDurationis marked Failed (BudgetExceeded) with a message explaining why, and its runner is cleaned up like any other failed session. Healthy sessions are unaffected; the runner still enforces the budget itself.