Comment 6 for bug 2019920

> That's true, but it would be helpful to check the running version of lxd on each node during/after upgrades. We feel it may help catch issues, since upgrading a multi-node lxd cluster must be done one at a time. Unless of course I'm misinterpreting the upgrade process (involving juju run lxd/N upgrade).

As you're running the upgrade process via Juju, doesn't `juju status` already show you the LXD version you're running? Replicating that info within AMS might also be not always possible as we need to rely on what LXD tells us and if it doesn't respond during a failed upgrade (LXDs REST API wont respond if it can't get the cluster online) the `amc node ls` output would not really help you.