VAPI-3542 fix(signaling): disable auto-reconnect before closing on disconnect - #13
Merged
Merged
Conversation
The rpc-websockets client is constructed with `reconnect: true` and `max_reconnects: 0` (unlimited), and the library only skips reconnecting when the close code is exactly 1000. Any other code reconnects forever — including 1001/StatusGoingAway, which pv-gateway sends on every close where it wants the device to come back (heartbeat timeout, and Handle's teardown backstop). _disconnect() removed all listeners and closed the socket but never disabled reconnect, so an explicit disconnect could leave a reconnect loop running behind a client that no longer has an "open" handler. The reconnected socket therefore never calls setMediaPreferences, never creates peer connections and never answers the heartbeat — an inert connection on the gateway with no liveness check on either side, which accumulates for as long as the process lives. Reconnect stays enabled for the lifetime of a live connection; it is only disabled in _disconnect(), where the caller has asked to tear this client down for good. Failing to disable it no longer prevents the close. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
✅ Snyk checks have passed. No issues have been found so far.
💻 Catch issues earlier using the plugins for VS Code, JetBrains IDEs, Visual Studio, and Eclipse. |
smoghe-bw
changed the title
cguthrie-bw
approved these changes
Aug 4, 2026
4 tasks
smoghe-bw
added a commit
that referenced
this pull request
Aug 20, 2026
…urns 409 (#17) * fix(signaling): stop auto-reconnecting when the gateway returns 409 The client is constructed with `reconnect: true, max_reconnects: 0`, and the error handler only ever treated 403 as terminal. A 409 — the gateway rejecting the connection because another device already holds the endpoint — was therefore treated as retryable, so the SDK reconnected several times a second against a gateway that was correctly telling it the connection could not be established. Gateway logs for one endpoint show 21 such rejections in 90 seconds across three hosts. Worse, each retry that did get a socket before being rejected left an inert connection behind, which held the endpoint and caused the next round of rejections — the same self-sustaining loop described in the _disconnect comment added by #13. Collect the terminal handshake failures into one table, disable reconnect and surface the error for both, and return so a fatal error no longer falls through into the generic error log. Retry policy is left to the app. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(signaling): tear down prior client before reconnecting connect() overwrote this.ws without closing the previous JsonRpcClient, leaving its unlimited auto-reconnect loop running in the background. That orphaned client's "open" handler calls setMediaPreferences() via this.ws, which by then points at the new client, so the orphan's own socket never sends anything and just idles until the gateway reaps it — producing repeated "new websocket connection" / "never called setMediaPreferences" storms against the same endpoint. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> --------- Co-authored-by: smoghe-bw <smoghe-bw> Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
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
3 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 rpc-websockets client is constructed with
reconnect: trueandmax_reconnects: 0(unlimited), and the library only skips reconnecting when the close code is exactly 1000. Any other code reconnects forever — including 1001/StatusGoingAway, which pv-gateway sends on every close where it wants the device to come back (heartbeat timeout, and Handle's teardown backstop)._disconnect() removed all listeners and closed the socket but never disabled reconnect, so an explicit disconnect could leave a reconnect loop running behind a client that no longer has an "open" handler. The reconnected socket therefore never calls setMediaPreferences, never creates peer connections and never answers the heartbeat — an inert connection on the gateway with no liveness check on either side, which accumulates for as long as the process lives.
Reconnect stays enabled for the lifetime of a live connection; it is only disabled in _disconnect(), where the caller has asked to tear this client down for good. Failing to disable it no longer prevents the close.