Repository navigation
fix: detect unresponsive runtime connections and expose transport failures - #64
Merged
Merged
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem and behavior
A runtime connection can stop delivering events to Core while the daemon continues executing behind a half-open proxy connection. Core's underlying read/write errors were sent through an unset optional logging callback, leaving only
read_endedin diagnostics.Record structured transport failure classes and WebSocket close codes without peer-controlled text. Probe daemon WebSocket connections every 15 seconds and require inbound traffic or Pong within 60 seconds. Bound blocked inbound dispatch as well, so disconnect cleanup can stop the active Turn and retain ownership until Executor cleanup is confirmed. Healthy connections renew their deadline; this introduces no total Turn duration limit.
A Turn that fails only because its Runtime connection closed (
event_stream_incompletewithconnection_closed) is now eligible for automatic Turn retry under the existingOAC_TURN_AUTO_RETRIESpolicy, fingerprintruntime_disconnected: the same limit, backoff, cancellation check and idempotency key apply. Event window overflow and other subscription failures remain non-retryable. In production all five failed Turns in the last seven days were this class.This PR targets beta, the deployed production baseline. It does not change ingress timeouts.
Validation
Limits
This fixes diagnostic loss and bounds unresponsive WebSocket cleanup; it does not establish the initiating cause of the historical Core disconnect. Pong establishes WebSocket peer liveness, so a proxy that independently replies to probes can mask a backend failure. No live four-hour run or production rollout has been performed.