lxd.migrate
Command ‘lxd.migrate’ is available in ‘/snap/bin/lxd.migrate’
The command could not be located because ‘/snap/bin’ is not included in the PATH environment variable.
lxd.migrate: command not found
Ok, so your PATH is somehow restricted, is it because you had to install snapd?
Normally /snap/bin is in the default PATH on systems with snapd installed.
Anyway, for the migration you can run /snap/bin/lxd.migrate directly, but you’ll need your PATH sorted after so you can get to the snap version of the lxc command.
ok, it seems i am back online. created a container with our profile, and it seems to start, as well as all the old containers. that’s a relief. thank you very much for your time. But i am a great supporter for containers, and i hope to get better with it, so the researchers can easily do their stuff.
thank again!
ps now that i am on snap, do the updates come automagically?
Correct, it will auto-update.
You’re currently on the latest track which rolls every month or so to the next LXD release. If you’d rather be on the 4.0 LTS track and only get bugfix point releases, you can do this with snap refresh lxd --channel=4.0.
Note that downgrades aren’t possible, so you can currently switch to 4.0 because latest is also 4.0, but once 4.1 is in latest you will not be able to stick to 4.0 without a reinstall.
You can also set a snap maintenance window so that the automatic updates are applied within a fix time window, that can be quite useful especially on such a shared system.
yeah i read about it, so i can do it with lxd now instead of zfs?
jef
I’m not sure what you mean there or the relation with zfs?
i was mixing up with snapshots on zfs, but i think you meant updates via snap.
Yeah, I was talking about update of the LXD daemon itself.
If you’re talking about snapshots of your containers, the new LXD does have built-in snapshot scheduling and expiry which you can use and then not have to worry much about that until you need to restore one 
thanks again Stephane, still have to read/learn alot.
but my collegues are already very optimistic about using the containers after what i am showed them. and now deep learning in containers is on the program …
jef