Docker & Unraid Connect loses internet/connection daily; rest of Unraid is fine

[rvs007]

I've been having this problem for weeks now and it's getting tiresome doing the workaround on a daily basis.  So I'm hoping someone can help me figure out what's happening.

 

On a daily basis (ie. once a day), my Docker containers would lose internet access.  All my containers run fine, but after awhile they can't seem to connect to the internet (usually after 1 day).  These are the symptoms I observe when this happens:

 

- Docker containers lose internet access

- Download of appfeed fails in the CA Apps tab

- All Docker containers show "not available" when checking for updates in the Docker tab

- Unraid Connect Error with message: "NETWORK:  queryA ECONNREFUSED mothership.unraid.netCLOUD: Socket closed"

 

I've searched the forum and others have experienced this and to fix the problem, the workaround posted has been:

 

1. Turn off Docker in the Settings and turn it back on (Docker container internet access gets restored after that)

2. Use 'unraid-api restart' in Terminal (this restores the connection for Unraid Connect)

 

What I haven't been able to figure out is how to fix this permanently so that this doesn't happen on a daily basis and I have to go through the above 2 steps daily. 

 

I've posted my latest diagnostics file (logs taken after the problem last happened and after I restored access with the 2 steps mentioned above).  

 

Can someone kindly review and see what might be leading to the problem on a daily basis?  I've posted the latest diagnostics logs.  

 

ps - I've also tried rebooting the server, but that didn't seem to make the problem go away.

 

Thanks in advance.  

 

homeserver2-diagnostics-20240526-2353.zip

[JorgeB]

Syslog is full of ffmpeg call traces, if you know which container uses that, try leaving it disable to see if you still see the same issue.

[rvs007]

On 5/27/2024 at 3:50 AM, JorgeB said:

Syslog is full of ffmpeg call traces, if you know which container uses that, try leaving it disable to see if you still see the same issue.

Thanks for that suggestion.  It looks like it's the docker container running Scrypted as I have temporarily disabled it and the problem hasn't returned for 2 days.  Will continue to monitor for the next few days.