Comment 10 for bug 1006553

juju destroy-environment will clean things out and clear out the upstart
jobs. the machine agent is a singleton on the host in a local provider
environment, regardless of the services deployed in containers.

On Fri, Jun 22, 2012 at 8:13 PM, Neal McBurnett <email address hidden>wrote:

> Thanks, Clint - makes sense, though I wonder how we'll configure
> services that need to restart on reboot.
>
> For the record, here are some more kern.log messages, which accompanied
> the strace that I talked about before, and give some more background on
> why the cpu calmed down.
>
> Jun 22 07:03:21 rf kernel: [74141.409402] init:
> juju-neal-local-machine-agent main process (1100) killed by TRAP signal
> Jun 22 07:03:21 rf kernel: [74141.429782] init:
> juju-neal-local-machine-agent main process ended, respawning
>
> Based on the respawn, I guess it may come back, so instructions geared
> to any newbies who run across this bug on how to most easily destroy the
> environment would help. I tried to dig up what I had done in my testing
> via the status command, but it doesn't help:
>
> $ juju status
> could not connect before timeout
> 2012-06-22 18:04:00,125 ERROR could not connect before timeout
>
> I can't run 'juju destroy-service' without knowing what service I
> started ;) So e.g. how can we figure that out?
>
> --
> You received this bug notification because you are subscribed to juju.
> https://bugs.launchpad.net/bugs/1006553
>
> Title:
> Juju uses 100% CPU after host reboot
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/juju/+bug/1006553/+subscriptions
>