I was working with a certain bundle… did some fiddling… but starting to see a message I don’t know what to do with
juju debug-log --verbose
machine-79: 16:23:21 ERROR juju.worker.dependency "broker-tracker" manifold worker returned unexpected error: cannot create instance broker: creating LXD container manager: Get "http://unix.socket/1.0": dial unix /var/snap/lxd/common/lxd/unix.socket: connect: connection refused
is this LXD complaining? my charm? thanks for taking a look.
This is Juju saying that on machine-79 it is failing to talk to LXD. Were you trying to deploy an application to an lxd container on machine 79? I believe we look both at /var/snap/lxd as well as /var/lxd (depending on how LXD is installed).
It would be helpful to know what your intent was, and what Ubuntu and Juju versions are involved (eg juju 2.9 on ubuntu focal, etc).
thanks John,
I had a large bundle… deployed it just fine… nothing seemed out of the ordinary beyond some minor charm related bugs… but this error almost sounded like a deeper problem.
I’m on juju v2.9.2 via Ubuntu 20.04 LTS installation (focal) … What I can do is… I am in midst of debugging this bundle… once it get ready and set to redeploy… I’ll keep an eye out if this occurs again and check back… otherwise I guess chalk it up to my charm code doing something odd?
I’m getting this same error message:
machine-0: 22:17:10 ERROR juju.worker.dependency "broker-tracker" manifold worker returned unexpected error: cannot create instance broker: creating LXD container manager: Get "http://unix.socket/1.0": dial unix /var/lib/lxd/unix.socket: connect: no such file or directory
The error message is posted roughly every 2 minutes.
The machine is the one and only worker node in a manual cloud created on Linode.
I didn’t try and deploy to LXD, or at least, not intentionally. It only contains a single Caddy application unit, and Caddy Domain subordinate. Both charms are created by me using the operator framework, and I don’t think there’s anything in them that should try and use LXD.
I vaguely remember forgetting to tack on --to 0 on my first attempt to deploy, getting an error, slapping my extensive forehead, and then deploying successfully to the machine.
juju 2.9.31 Controller and worker are both running Ubuntu 22.04. I have not installed LXD on either machine.
If there’s anything else I can do to help gather info, please let me know.