Comment 21 for bug 1751842

I'm not saying make it part of the juju status output, except maybe to
summarize - i.e. just show health status there.

You could capture the rest in the unit log.

On Fri, Mar 2, 2018 at 11:32 AM, Chris MacNaughton
<email address hidden> wrote:
> It's a bit big for juju status output:
>
> $ juju ssh ceph-mon/0 sudo ceph status
> sudo: unable to resolve host juju-adc4f7-icey-0
> cluster dd921caa-1d6c-11e8-893f-fa163e127e47
> health HEALTH_OK
> monmap e1: 3 mons at {juju-adc4f7-icey-0=10.5.0.24:6789/0,juju-adc4f7-icey-1=10.5.0.16:6789/0,juju-adc4f7-icey-2=10.5.0.12:6789/0}
> election epoch 10, quorum 0,1,2 juju-adc4f7-icey-2,juju-adc4f7-icey-1,juju-adc4f7-icey-0
> osdmap e13: 3 osds: 3 up, 3 in
> flags sortbitwise,require_jewel_osds
> pgmap v1021: 64 pgs, 1 pools, 0 bytes data, 0 objects
> 7964 MB used, 36463 MB / 44475 MB avail
> 64 active+clean
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1751842
>
> Title:
> Image creation failed: Error finding address for
> http://xx.xx.xx.xx:9292/v2/images/x-x-x-x-x/file: Unable to establish
> connection to http://xx.xx.xx.xx:9292/v2/images/x-x-x-x-x/file
>
> Status in OpenStack ceph-mon charm:
> New
> Status in OpenStack ceph-osd charm:
> New
> Status in OpenStack glance charm:
> Invalid
>
> Bug description:
> We have hit several failures of this pattern. This is using HA glance
> with ceph as the back end.
>
> Sometimes this occurs right after we finish the openstack deploy and are uploading images:
> http://paste.ubuntu.com/p/hm7K7FfYsB/
>
> At least once it's happened during a rally test.
>
> Here is the bundle and overlay:
> http://paste.ubuntu.com/p/Rhf4FN2kJ3/
>
> Despite the error message, I think it can connect, because glance-api.log has these log messages:
> http://paste.ubuntu.com/p/zRrR8QczXN/
>
> 10.244.40.215 is the ip of the glance 0 unit, and the above log comes
> from the glance 1 unit. I think it's haproxy from glance 0 forwarding
> the connection to glance 1.
>
> I think this is probably a timeout problem; maybe haproxy needs to be
> adjusted?
>
> This is also suspicious, regarding upload speed:
> http://paste.ubuntu.com/p/g5f6fhyc2M/
>
> Some example failures from rally:
> http://paste.ubuntu.com/p/ZPbFD4WNRZ/
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/charm-ceph-mon/+bug/1751842/+subscriptions