Not sure if this is an Unraid/Docker problem or something else I did, but here goes...

 

I recently rebooted my server and found that almost none of my Docker containers had started up as they should, and when I tried starting them manually I got an error saying 'no such container'. Long story short, I got those containers working by disabling IPv6 in the Docker settings for the custom network, which I previously had enabled. I use a custom network in order for my web services to get their own internal IP address, to ultimately be accessed externally.

 

I was hoping I could keep IPv6 enabled so I could then assign IPv6 addresses to my containers when I wish to enable external IPv6 access in the future. If it makes any difference as to why this might have happened, the only thing that changed here recently is I changed ISP and got a static IPv6 prefix from them (I had a dynamic address from the previous ISP).

 

Any ideas on how to resolve this?

OK I got it working in the end but not the way I expected to. I have an OPNsense router and if you set the IPv6 mode to track the WAN interface, there's an option to set a prefix ID. I think this was originally 0, I can't remember now. I changed it to 1, rebooted the router, then tried re-enabling IPv6 in Docker once again. Now the containers startup as normal and I can now set my own IPv6 address for each one. I'm guessing that whatever prefix ID Docker was trying to assign to the custom interface clashed with my router's settings, and therefore refused to create the custom interface until I either disabled IPv6 or changed the router settings.

Thought I would post this update in case anyone else finds themselves in the same position.