fix(workbench): bound JupyterLab cell-run retry timeout#534
Merged
Conversation
The type-and-run retry loop waited the full TIMEOUT_CODE_EXEC (30s) on every attempt, so a genuinely stuck cell took up to ~90s to reach the skip. Retries exist to recover a lost keystroke against an already-live kernel, which resolves in seconds — only the first attempt needs the full cold-kernel budget. Give attempt 1 TIMEOUT_CODE_EXEC and subsequent attempts the shorter TIMEOUT_DIALOG, bounding the worst-case failure path without changing the happy path. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
samcofer
enabled auto-merge (squash)
July 24, 2026 18:59
|
|
Preview Links
|
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Follow-up to #532. The JupyterLab type-and-run retry loop in
jupyterlab_executes_codewaited the fullTIMEOUT_CODE_EXEC(30s) on every attempt, so a genuinely stuck cell took up to ~90s to reach the skip.Retries exist to recover a lost keystroke against an already-live kernel, which resolves in seconds — only the first attempt needs the full cold-kernel budget. This gives attempt 1
TIMEOUT_CODE_EXECand subsequent attempts the shorterTIMEOUT_DIALOG(10s), bounding the worst-case failure path without changing the happy path.Verification
vip verify --filter launch_jupyterruns on the installed build:pwb.demo.soleng.posit.it(Workbench 2026.07, interactive) — 3/4 passed; the one failure was an upstream session-never-reached-Active-in-90s startup timeout, unrelated to this change.workbench-helm.local.cofer.me(Workbench 2026.06, headless, 5× serial) — 4/5 passed; the one skip was the accurate "UI-interaction timeout, not kernel death" message on a 140s-slow session start.Known follow-up
On very slow cold starts the shortened retry budget can still lose (helm run 5). A future improvement is to gate retries on the kernel-status indicator reaching
idlebefore re-running, so a still-connecting kernel waits rather than burning a short retry.🤖 Generated with Claude Code