Comment 18 for bug 1577519

Ok. Could you show me your build/deploy process? I was chatting with
beisner and he mentioned that landscape has an amulet like deploy
procedure for deploying the charms. I think that could help me figure
out how it ended up in a race condition.

On 05/31/2016 05:08 PM, Andreas Hasenack wrote:
> I don't have a repro. We deploy, and most of the time with ceph/ceph it
> happens. All the data we have is in the attached logs. We have logs for
> other services in the deployment if you want. Uncompressed in total it's
> about 2 gigabytes.
> On May 31, 2016 19:21, "Chris Holcombe" <email address hidden> wrote:
>
>> Andreas, what's the repro for this? I tried deploying 3x ceph-mon, 1x
>> ceph-osd and 1x ceph-radosgw and the radosgw waited until 2 more ceph-
>> osds were added the cluster went healthy. I'm curious how landscape is
>> deploying and how that is still hitting the bug.
>>
>> --
>> You received this bug notification because you are subscribed to
>> Landscape Project.
>> Matching subscriptions: default, everything
>> https://bugs.launchpad.net/bugs/1577519
>>
>> Title:
>> ceph-radosgw "Initialization timeout, failed to initialize"
>>
>> To manage notifications about this bug go to:
>> https://bugs.launchpad.net/landscape/+bug/1577519/+subscriptions
>>
>