Simon Deziel

@kadinsayani I'll try to have the `arm64` snap include a `lxd-agent` binary built for `armhf`. This should provide some usability back to that arch that's being phased out but folks...

Once implemented, this should allow (or make it easier) to drop the `nvidia-container` part of the LXD snap.

> @tomponline I've isolated the storage bucket backups functionality as discussed, though I'm seeing the pipeline fail due to large lxc binary size: `FAIL: lxc binary size exceeds 15MiB` `main`...

> @simondeziel I understand the current behavior is if you don't specify the IP as a positional argument, it tries to lookup the LXD server based on the given name...

> > While I'd keep the short name lookup logic, I'd probably use the address list as hints to send probes to, no? > > Do you think the client...

> > The `c.flagAcceptCert` makes me thinking. Is there ever a case where you would provide a token which doesn't contain the servers fingerprint? Because if not we could also...

`core24` was recently [updated](https://github.com/canonical/core-base/pull/372) to exclude private snap `/tmp` dir from being cleaned up every 30 days.

@hamistao as you described, I think that due the `/dev/disk/by-id` symlinks ending up as the same, the 2nd disk sharing the name prefix is the one taking over the `by-id`...

Let's close this bug as LXD has no real way to workaround this. The operator has to know to use different device prefixes if multipathd is to be used inside...

With the pastebin JSON data: ``` $ jq '.resources.cpu.sockets[] | .name' < /tmp/lxc-info.json "Intel(R) Xeon(R) CPU Max 9462" "Intel(R) Xeon(R) CPU Max 9462" "pc-q35-6.2" "pc-q35-6.2" "pc-q35-6.2" "pc-q35-6.2" "pc-q35-6.2" "pc-q35-6.2" "pc-q35-6.2"...