[Bug]: Issue with stuck channels, or ghost channels.

[mcatatau]

Describe the bug

Sometimes dispatcharr gets stuck with streams, even without any clients connected.
Sometimes it is necessary to restart dispatcharr for the streams to be finalized.

How can we recreate this bug?

In Proxy Settings, set:
Buffering Timeout: 5
Buffering Speed: 1
Buffer Chunk TTL: 0
Channel Shutdown Delay: 0
Channel Initialization Grace Period: 0
In Stream Settings -> Default Stream Profile: ffmpeg/Streamlink

I redid the process several times, and this always happens.

Dispatcharr Version

v0.11.2-20251118160303

Docker Version

Docker version 27.5.1, build 9f9e405

What's in your Docker Compose file?

If not using Docker Compose just put not using.

Client Information

VLC 3.0.20 vetinari

[distributorofpain]

i have the same issue, its quite random though. I can channel surf and within a few minutes all the streams close properly when testing. I tried adding a delay of 5 seconds for channel shutdown(under settings). This helped with flipping back and forth between channels. I tested with plex, tivimate and jellyfin. Jellyfin does tend to hold channels longer, but most still released. My guess on this would be if the channel failed to actually stream and got stuck trying to initiate the stream. I see it much less as the versions have upgraded, so i gather there is some work to address errors as they are debugged.

I havent seen any yet with version 11.2 or 12, but my host server has been having issues, so its uptime has been crap as of late.

[slamanna212]

You are running a dev release, please try the running the latest stable release using the :main tag and let me know if the issue still exists

[distributorofpain]

i am up to version 13.1 now, i havent seen ghost channels in a little while. However, my server had a bad memory module that was causing it randomly reboot for a few weeks. I will keep an eye on it, but so far so good ( i went a week with it not rebooting and there were not stuck channels in 12.1 on either my dev or production VM). Both use stable version, not Dev. My Dev node is more for me playing with stuff, i havent tested dev versions as of yet.

[abiedster]

I have exactly the same issue as of today. I have to reboot it several times a day :(

[jcasimir]

Triage note: This issue appears to be mislabeled. It should have the question label removed and Bug added—this is a bug report about stuck/ghost channels requiring restart to clear (multiple users affected).

[Lawdie]

Also experiencing this issue intermittently. Killing the stream in the Status tab works for me however. Currently on v0.17.0. The video playback is triggered via Emby and is using ffmpeg

Would it be possible to add a "stream timeout" setting so if any stream runs for X minutes / hours it just ends and requires the client to restart the stream. (Or in the case of ghosts, not)