-
|
I run native nextcloud but have the recording backend as a docker container from: ghcr.io/nextcloud-releases/aio-talk:latest This has worked great in the past. with the latest update, nextcloud reports Error 503 and cannot connect to the talk high performance backend container. When I restore the old container image it works again. The error from what I can remember in the logs that it is unable to resolv my NC domain and cannot connect janus and nats. How can I see what has changed in between images so that I can troublehsoot? |
Beta Was this translation helpful? Give feedback.
Replies: 9 comments
-
|
Hi, please post the complete logs here. |
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
|
And this is the log when it is working fine, previous image: and when you use the talk app you get things like: |
Beta Was this translation helpful? Give feedback.
-
Indeed this doesnt look right. |
Beta Was this translation helpful? Give feedback.
-
|
Can you post the whole compose file you used foe the talk container here? |
Beta Was this translation helpful? Give feedback.
-
|
I think I fixed it by adding the following environment variable:
|
Beta Was this translation helpful? Give feedback.
-
|
So does it work now? |
Beta Was this translation helpful? Give feedback.
-
|
Yes, I made post about this in janus github, I guess it has to do with janus 1.4.1 |
Beta Was this translation helpful? Give feedback.
-
|
I’m seeing what appears to be the same or a very similar issue after updating Nextcloud AIO to 33.0.3. Environment / context:
Container status showed: The health check failed immediately on port 8081: The Talk container logs repeatedly showed Janus failing to start because of the TURN hostname: However, DNS inside the same container worked: The generated Janus config was: As a temporary workaround, I changed only the Janus TURN server value to the resolved private IP: docker exec -it nextcloud-aio-talk sh -c '
cp /conf/janus.jcfg /conf/janus.jcfg.bak
sed -i "s|turn_server = \".*\"|turn_server = \"<REDACTED_PRIVATE_IP>\"|" /conf/janus.jcfg
'After that, Janus started and the health check passed: Docker then showed: I also tested rolling Docker back from 29.5.2 to 29.3.0 because Docker was updated around the same time, but the Janus/TURN hostname failure remained after the rollback. So this does not appear to be fixed by changing Docker versions. This looks like a regression or behavior change in the updated AIO Talk / Janus image where Janus rejects the TURN hostname even though normal Is there a supported persistent way in AIO to set |
Beta Was this translation helpful? Give feedback.
I think I fixed it by adding the following environment variable:
- TURN_DOMAIN=talkhpb.vom-bruch.com