[libvirt] W/ commit 9e159b5: error from service: CreateMachine: Machine 'qemu-foo' already exists

[Kashyap Chamarthy]

On Mon, Sep 29, 2014 at 08:17:11AM +0200, Guido Günther wrote:
> On Mon, Sep 29, 2014 at 07:12:29AM +0530, Kashyap Chamarthy wrote:
> > Applied it:
> > 
> >     $ git log | head -5
> >     commit da0a33c3f286ff69b910bc2a84e9ec3a844e3ce8
> >     Author: Guido Günther <agx at sigxcpu.org>
> >     Date:   Thu Sep 25 14:30:42 2014 +0200
> >     
> >         qemu: use systemd's TerminateMachine to kill all processes
> > 
> > And tested (by building RPMs, restarting libvirtd and attempting to
> > start the guest again, still no luck:
> > 
> >     $ sudo systemctl restart libvirtd
> >     $ sudo systemctl daemon-reload
> 
> Did you reboot in between? 

No, I didn't. I realized rebooting might help solve the issue as this
was related to systemd, but didn't wanted to, as it's my main laptop,
and have several other things that I'm processing.

> The patch fixes the _stopping_ of the vm. Once the vm is in the state
> you describe the patch won't help.

Ah.

> 
> An alternative to rebooting is to remove related processes manually, see:
> 
>  http://bugs.debian.org/761521

I don't see x2d<$vmname>.scope under machine.slice (mostly as the
guests are all off, but I can't start them anyhow):

  $ tree /sys/fs/cgroup/cpuset/machine.slice
  /sys/fs/cgroup/cpuset/machine.slice
  ├── cgroup.clone_children
  ├── cgroup.procs
  ├── cpuset.cpu_exclusive
  ├── cpuset.cpus
  ├── cpuset.mem_exclusive
  ├── cpuset.mem_hardwall
  ├── cpuset.memory_migrate
  ├── cpuset.memory_pressure
  ├── cpuset.memory_spread_page
  ├── cpuset.memory_spread_slab
  ├── cpuset.mems
  ├── cpuset.sched_load_balance
  ├── cpuset.sched_relax_domain_level
  ├── notify_on_release
  └── tasks

And, the 'tasks' file is empty:

  $ cat /sys/fs/cgroup/cpuset/machine.slice/tasks

Is there anything else that I can clean up manually here?

Thanks for the response, Guido.

--
/kashyap