Docker issues - Unable to start any container, says they do not exist

[Demon1337]

Version: 6.12.6

Hardware: HP Proliant DL380 Gen9

Network Card: 10G PCIE

GPU: Tesla M40

 

I have recently had issues with my server. The USB stick just stopped booting into UEFI mode. It was an old stick so it having issues is expected. I loaded a fresh OS onto a new USB drive and things were back going. Fresh OS, Fresh USB. 

 

24 hours later, the server just locks up and wont let me reach any of the containers except with ping commands. VMs were the same story. Tried login at the box itself. No dice. Gives an error saying:

 

-bash: /usr/share/bash-completion/bash_completion: input/output error

 

I reserve that I need to reboot it. So I do that but unplugging for a while and coming back to it. Server boots and most things look fine. It was a bad reboot, but things work. I start looking into issues related to 6.12.6 and find something about MACVLAN. I don't see anything enable for me for that, so I just go about installing the typical plugins. 

1. Fix common problems

2. GPU driver

3. Nerd Tools

 

Fix Common problems picks up MACVLAN as an error. Ok, I set out digging where. I properly shutdown the array and bridging is enabled, which means MACVLAN should be disabled. I spin back up the array and look at the docker containers individually. One of them is set to bridge instead of BR0 (I IP all of my services) I delete that container as it wasn't needed anyways. Suddenly I look at all my containers are off. Powering on gets an error saying it doesn't exist. Logs say this for every container:

Namecheap-DDNS: Error response from daemon: network dd7e9e2d8f64eceeee342e267c8acb612a734b51ffde7059005b70b0251a6e9f not found

 

Listing networks shows:

NETWORK ID     NAME      DRIVER    SCOPE
40dd8737aab6   br0       ipvlan    local
214e8ee49a31   bridge    bridge    local
1f1179c5678d   host      host      local
a013fcade590   none      null      local

Listing the containers...completely empty. 

I would think I could just create that network but not sure exactly how that would be done, or how it would respond to the network. No idea if it just picks up my containers. 

 

I have no clue exactly what is going on at this point and am at wits end with the whole thing. I have has zero issues out of Unraid since March out of this box. No errors or anything. The old USB stick ran for something like 4 years without issues. I am just frustrated that something obvious seems broken and I can't quite put my finger on what went wrong.  If anyone has any clue exactly whats up here, I am all ears. I fear if this keeps up I might have to change what OS I use because this is really starting to become a large pain point. 

unraid-diagnostics-20240101-1759.zip

Edited January 2 by Demon1337

[JorgeB]

Try recreating the docker image and restoring the containers from CA.

[Demon1337]

13 hours ago, JorgeB said:

Try recreating the docker image and restoring the containers from CA.

I have finally fixed it. I found that the templates from the old USB drive were not there and without them moved over the containers would not boot. I had attempted to load the same containers I had and when I did the first one all the data was deleted in the process. Thankfully it was the test container. 

 

For anyone that has this issue in the future:

 

You will need the templates from the containers. They exist at: 

/boot/config/plugins/dockerMan/templates-user

 

Attached is a version of the template in case you need it to remake yours. You can also get the template from another container loaded fresh. The important part is the network information. In my case br0 and the IP. but you should try to get as much information as you can. 

 

my-PlexMediaServer.xml