Devcontainers fail with `Cannot connect to the Docker daemon at unix:///var/run/docker.sock`

I don’t know what is going on.
Seems that an update in the latest tag of bluefin-dx-nvidia must have broken docker + devcontainers.

❯ docker ps -a
Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?

ujust check-local-overrides doesn’t show anything outside of ca-certificates and some fstab /cryptab stuff.

Is anybody else getting the same error as I am?
Or is my bluefin install broken? If so, how do I fix it?

I just upgraded to latest bluefin-dx-nvidia (38.20240330.0 (2024-03-30T16:50:47Z):
image

I issued the same Docker command docker ps -a without any errors:
image

But our Docker installs might not be the same: I changed my original Docker configuration from needing sudo to a rootless Docker.

I have no problem with Docker with bluefin-dx-nvidia-38.20240330.0, but my LXC containers cannot start now.

I think it is related to starting some services at boot.

This is an opportunity to test the resilience of Bluefin. I will try to go back to previous OS version, and let you know how it went.

OK.
Back from the boot of bluefin-dx-nvidia-38.20240329.0.

I had the same error when trying to start LXC containers:

But this time I was able o track where the problem was. And the Bluefin upgrade is not to blame. There is a service that does not automatically start and is essential for
starting LXC containers. Possibly a similar service in Docker is not started automatically at boot.

To fix it, I run this command:

sudo systemctl start libvirtd.service

and I am now back in business!

Couple of screenshots of the status before (throwing errors) and after:

Before

After

Hope this helps.

Note

I read somewhere - related to SilverBlue - that Docker and virtualization services are not started by default in order to keep the OS snappy. Let’s remember that Bluefin-DX is aimed at developing not serving applications. I we want those services starting at boot, we have to explicitly set them to do so.

Had to do this as sudo :

ip link del docker0
rm -rf /var/docker/network/*
mkdir -p /var/docker/network/files
systemctl start docker

Was this caused by the 38.20240331 upgrade?

This topic was automatically closed 12 hours after the last reply. New replies are no longer allowed.