On Wed, Sep 21, 2016 at 9:19 PM, Federico Alberto Sayd <fsayd@uncu.edu.ar> wrote:

Hello:

Since I upgraded to 4.0 I have troubles with the storage of my oVirt setup. I have 2 IBM 3512 arrays connected to 8 ovirt hosts through FC.
When a host is rebooted or when a new port get online in the FC switchs although the port isn't connected to storage or oVirt hosts; oVirt reports I/O problems and pause some vm's. I have to manually resume the vm's. It seem that any change (even non disruptive changes) to FC switchs and HBAs causes that ovirt detects I/O problems.

How I can debug this issue. How can I know if It is related to vdsm/ovirt, mulitpathd..?

We need to inspect engine log, vdsm log, supervdsm.log, sanlock log, and /var/log/messages.

If ovirt detect io problems it is probably because they are problems.

Can you file a bug and attach the logs?

Nir