Docker container becomes unresponsive overnight

[wapkaplet]

Hi there,

I'm having an odd issue with my NextPVR docker container, running on Unraid. Overnight it becomes unresponsive. It still shows as running in the backend, but the webUI can't be reached anymore, and it doesn't respond to clients. A simple restart of the container resolves the problem. Here's what I see in the logs:

[30] LoadListingByAdvancedRules(channel=REMOVED)
[30] UpdateEPG@5
[30] Exported 2 recordings to '/config/recording-backup.xml'
[30] Exported 4 recurrings to '/config/recurring-backup.xml'
[30] No /config/scripts/PostUpdateEPG.sh file found
[30] Finished updating EPG
[14] PowerManagement AllowSleep (needs implementation)
[14] PowerManagement CancelWakeup (needs implementation)
[14] Requesting wake up for 'EPG-UPDATE' [EPG Update] at 01/07/2025 06:44
[14] PowerManagement ScheduleWakeup (needs implementation)
[13] Reloading recording schedule
[13] PowerManagement CancelRecordingWakeupRequests (needs implementation)
[13] [REMOVED] at 01/07/2025 19:30
[13] Requesting wake up for 'RecordingOID878' [REMOVED] at 01/07/2025 19:28
[13] PowerManagement ScheduleWakeup (needs implementation)

This happens every morning at 6:44am, which is when I have the guide update set. Seems like the container is going to sleep, which to my knowledge should not be happening. I'm sure this is more down to how I've setup the container than a NextPVR problem, but I'm stumped as to how to resolve it. Does anyone have any guidance on what I can do to correct this?

[mvallevand]

Full zipped logs please.

Martin

[wapkaplet]

Full zipped logs please.

Martin

Sent via PM.

[mvallevand]

Looking at your logs I see the XMLTV download fails at 6:44 so yes I can only assume your networking is broken.

Not sure why you are using a public IP, but maybe the IP is changing. I notice the IP on your web browser changes too, so your home router might need be reliable or you are using a VPN.

Martin

[wapkaplet]

Looking at your logs I see the XMLTV download fails at 6:44 so yes I can only assume your networking is broken.

Not sure why you are using  a public IP, but maybe the IP is changing.  I notice the IP on your web browser changes too, so your home router might need be reliable or you are using a VPN.

Martin

Thanks for looking! What's weird about that is that when I manually update the EPG it works as expected. The container itself is routed through a VPN. Not sure if that would have an effect on anything.

Anyway, I've just spun back up my Threadfin instance and connected NextPVR to that instead. Perhaps that will make things more stable. I'll see what happens tomorrow morning when the EPG update runs.

Appreciate the help!

[wapkaplet]

Just to follow-up on this, switching to NextPVR pointing to Theadfin for it's channels and EPG had no effect. The system was in the same state after trying to update the guide on schedule. I updated the NextPVR container so it's not routed through the VPN, and then set the guide update to a time when I could monitor it. It completed successfully and the container did not become unresponsive.

This is a little confusing to me, simply because when routed through the VPN, I can still manually update the EPG from the webUI. It doesn't seem like just routing it through a VPN should be enough to cause this problem. But I'd say definitively that's the root cause.

[wapkaplet]

For what it's worth, I figured this out. It had nothing to do with the EPG update, or the docker container itself. It did connect to the docker's route through a separate VPN container. There's a nightly docker backup running, and the problem was with the order in which it took down and brought back containers. If they came up in the wrong order, the NextPVR container would get confused as to its network connection. It would still be there, and still be running, but be unable to connect to anything.

Fixing the backup order fixed the overnight outages.

Thanks again for the help!