[ovirt-devel] VDSM crashed while adding host to newly created cluster

[Christopher Pereira]

On 26-03-2015 19:24, Christopher Pereira wrote:
> Is it normal that 1) the engine stops VDSM during activation and 2) 
> that VDSM stops or restarts the storage during shutdown?

Well, no wonder that the engine-storage disappears since the gluster 
process is a child-process of the vdsmd.service

    CGroup: /system.slice/vdsmd.service
            ├─14895 /usr/bin/python /usr/share/vdsm/vdsm
            ├─14964 /usr/libexec/ioprocess --read-pipe-fd 43 
--write-pipe-fd 42 --max-threads 10 --ma
            ├─15836 /usr/libexec/ioprocess --read-pipe-fd 49 
--write-pipe-fd 48 --max-threads 10 --ma
*           ├─15911 /usr/sbin/glusterfs 
--volfile-server=h2.imatronix.com --volfile-id=engine /rhev/d*
            └─15922 /usr/libexec/ioprocess --read-pipe-fd 61 
--write-pipe-fd 60 --max-threads 10 --ma

A workarround is to mount the engine-storage manually, so that 
restarting vdsmd on the host doesn't kill the engine VM.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ovirt.org/pipermail/devel/attachments/20150326/195a50d6/attachment-0001.html>