Confirm the dropped-connection retry (PR #134) against a live source #138
Labels
No labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set
Reference
cmoriarty/sports-breaker#138
Loading…
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
PR #134 sends an upstream request once more when the server closes a kept-alive connection before answering (
httpx.RemoteProtocolError). It was written from one event and is covered by its unit test only.The event (2026-11-10 01:09:26Z): a thetvapp channel failed on "no new segment for 17 s; last error: playlist: upstream unreachable: RemoteProtocolError". The error was the channel's last one, not proven to be why it stalled.
To check
RemoteProtocolErroras its last error.Done when production logs show the retry carrying a channel through, or show it is not what failed that channel.
The retry now logs
upstream: <host> dropped a kept-alive connection; asked again. Merged in PR #140 (25fa577), deployed 2026-11-10 05:21Z, archived in4952376. No followed game was live at the deploy, so none of it has been seen in production yet.Left open for the check itself: look for that line in the container log beside a channel that kept playing, and for any channel that still fails with
RemoteProtocolErroras its last error.Seen in production on 2026-11-10, so closing. The retry is PR #134 and its log line is PR #140 (
25fa577).upstream: a.hatemamere3.space dropped a kept-alive connection; asked again, the only such line that evening.RemoteProtocolErrorappeared in the log around it, and no channel failed with it.Not proven: the log names only the host, so it does not show whether that request was the channel's or a stream check's. A new ticket can follow if a channel fails with
RemoteProtocolErroras its last error.