failed to destroy model - state changing too quickly

[james beedy]

the title says it all ..... http://paste.ubuntu.com/25442668/

[John A Meinel (jameinel)]

state changing too quickly means we were trying to apply a transaction and
got a failure more that ~5 times trying to apply it. It's possible that
there is something else wrong that is blocking the change that isn't
concurrent changes to the database. It would seem we aren't giving a lot of
other information about the failure to the client. Have you checked machine
logs to see if there is something interesting going on there?

John
=:->

On Sep 1, 2017 08:01, "james beedy" <email address hidden> wrote:

> Public bug reported:
>
> the title says it all ..... http://paste.ubuntu.com/25442668/
>
> ** Affects: juju
> Importance: Undecided
> Status: New
>
> ** Description changed:
>
> - $ juju status
> - Model Controller Cloud/Region Version SLA
> - elasticsearch-dev jujucharms.com aws/us-west-2 2.2.2 unsupported
> -
> - App Version Status Scale Charm Store Rev OS Notes
> -
> - Unit Workload Agent Machine Public address Ports Message
> -
> - Machine State DNS Inst id Series AZ Message
> -
> -
> - $ juju destroy-model elasticsearch-dev
> - WARNING! This command will destroy the "elasticsearch-dev" model.
> - This includes all machines, applications, data and other resources.
> -
> - Continue [y/N]? y
> - Destroying model
> - ERROR failed to destroy model "elasticsearch-dev"
> - ERROR cannot destroy model: failed to destroy model: state changing too
> quickly; try again soon
> + the title says it all ..... http://paste.ubuntu.com/25442668/
>
> --
> You received this bug notification because you are subscribed to juju.
> Matching subscriptions: juju bugs
> https://bugs.launchpad.net/bugs/1714409
>
> Title:
> failed to destroy model - state changing too quickly
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/juju/+bug/1714409/+subscriptions
>

[Paul Gear (paulgear)]

I'm seeing this on the default (totally empty) model after successfully upgrading both the controller and model to 2.2.3: https://pastebin.canonical.com/198370/

[Paul Gear (paulgear)]

Occurred on my next upgrade, this time on AWS: https://pastebin.canonical.com/198375/

[Anastasia (anastasia-macmood)]

@Paul Gear (paulgear), @james beedy,

Do you have logs to attach?

The 2 commits that went into 2.2 and that deal directly with model destruction are:
https://github.com/juju/juju/commit/efe6e6a67ceeb5f984d47455079e3f2ac2832dfc and https://github.com/juju/juju/commit/ba3cc6808c416146b28b6ef8a7609477df0845bf.

[Paul Gear (paulgear)]

Logs from controllers relating to both attempts listed above are at https://private-fileshare.canonical.com/~paulgear/lp1714409/