percona-cluster service in unknown state
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
pyjuju |
New
|
Undecided
|
Unassigned |
Bug Description
Percona-cluster stopped syncing in my cloud without anything to point to the reason why. Mysql wasn't starting on any of the lxc containers after restarting them.
This was working on Friday...
I decided that since this is a lab, I would destroy the service and redeploy. Right now the service is stuck in an unknown state and it won't go away.
All of the units have been removed. All of the lxc containers are gone.
juju-1 status --format yaml
...
percona-cluster:
charm: local:trusty/
exposed: false
life: dying
service-status:
current: unknown
message: Waiting for agent initialization to finish
since: 07 Dec 2016 20:27:11-05:00
relations:
shared-db:
- keystone
...
any way to force remove the service? I've seen recommendations to remove the machine with a --force, but the machines are gone.
Not sure what else I can try.
$ juju-1 destroy-service percona-cluster --debug
2016-12-12 21:11:28 INFO juju.cmd supercommand.go:37 running juju [1.25.6-
2016-12-12 21:11:28 DEBUG juju.api api.go:154 trying cached API connection settings - endpoints [10.9.190.22:17070]
2016-12-12 21:11:28 INFO juju.api api.go:266 connecting to API addresses: [10.9.190.22:17070]
2016-12-12 21:11:28 INFO juju.api apiclient.go:262 dialing "wss://
2016-12-12 21:11:28 INFO juju.api apiclient.go:194 connection established to "wss://
2016-12-12 21:11:29 DEBUG juju.api api.go:476 API hostnames unchanged - not resolving
2016-12-12 21:11:29 DEBUG juju.api api.go:506 cacheChangedAPI
2016-12-12 21:11:29 INFO juju.cmd supercommand.go:435 command finished