Update charm-dev to juju 3.1

[sed-i]

The juju snap now has version 3.1 as stable so it might be a good time to update the charm-dev blueprint.

• Introduce explicit version pin to reduce brittleness of the blueprint: juju 3.1/stable and microk8s 1.27-strict/stable.
• Pick up some goodies from @Abuelodelanada's cloud-init.
• Pick up the credentials fix from #33

[sed-i]

Something's quite odd:

If I first try to bootstrap lxd, and then microk8s, then bootstrapping lxd fails but then bootstrapping microk8s succeeds.

update.go:85: cannot change mount namespace according to change mount (/run/user/1000/doc/by-app/snap.juju /run/user/1000/doc none bind,rw,x-snapd.ignore-missing 0 0): cannot inspect "/run/user/1000/doc": lstat /run/user/1000/doc: permission denied

If, on the other hand, I first try to bootstrap microk8s, and then lxd, then bootrapping microk8s fails but then bootstrapping lxd succeeds.

Terminated
ERROR No controllers registered.

Perhaps related: https://bugs.launchpad.net/juju/+bug/2003582

[sed-i]

This is working now locally.

[sed-i]

I tested this manually several times and it now works well consistently. Mind taking another look @zorzigio @Abuelodelanada @townsend2010 ?

[taurus-forever]

@townsend2010 is it possible to approve this PR?

The next week, we are planning to demo Data charms using multipass and share Lab possibility for visitors to try multipass launch charm-dev + juju deploy ..., the broken charm-dev is a show stopped for us.

As a workaround, we will use --cloud-init with local blueprints, but "Magic of simplicity" will gone. Tnx in advance!

[townsend2010]

Hi @taurus-forever!

I still have an open question here. I would really like for @sed-i to help me better understand why this 30 seconds sleep is necessary. My concern stems that it seems arbitrary and only leads to an even longer bootstrap of the instance (even worse user experience) and I'd like to determine if it's not necessary, can be set to a much lower timeout, or some type of wait that signals things are ready, I would be much more happy about that.

That said, given the demo next week, this sounds pretty critical, so I guess we can merge it, but would at least like some acknowledgement to my concern.

[sed-i]

I missed your reply back then, @townsend2010. Let me try this again now, maybe the issue is gone for me too.

[sed-i]

@townsend2010 the sleep is gone. @taurus-forever I changed the juju channel to 3.1/stable because it will be supported until 25.04.

[townsend2010]

Hey @sed-i,

Thanks for this! I will see how the integration tests go, give it a quick spin, and then most likely approved and merge if nothing seems wrong.

[sed-i]

Hi @townsend2010, I tested this locally after switching to 3.1 and it looks fine. Wdyt?

[townsend2010]

bors merge

[bors[bot]]

Build succeeded!

The publicly hosted instance of bors-ng is deprecated and will go away soon.

If you want to self-host your own instance, instructions are here. For more help, visit the forum.

If you want to switch to GitHub's built-in merge queue, visit their help page.

• Bors