RHEV-M error : nonOperationalReason=NETWORK_UNREACHABLE
We had to perform general power cut for our lab servers including the RHEV ones . Before this I set all hosts in the maintanance mode.
After the all servers are up and running I cannot get RHEV to normal - host servers are in non operational mode , fc domain is in non-responsive mode
/var/log/rhevm/rhevm.log when tried to activate one of the hosts:
2012-02-12 14:02:47,973 INFO [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] (pool-11-thread-19) START, SetVdsStatusVDSCommand(vdsId = bb17a744-0b9e-11e1-bed0-2c27d7287d08, status=NonOperational, nonOperationalReason=NETWORK_UNREACHABLE), log id: 78dafc0c
2012-02-12 14:02:47,978 INFO [org.ovirt.engine.core.bll.SetNonOperationalVdsCommand] (pool-11-thread-19) Host RH6LAB2 is set to Non-Operational, it is missing the following networks: apps,
2012-02-12 13:59:50,903 ERROR [org.ovirt.engine.core.vdsbroker.vdsbroker.VdsBrokerCommand] (QuartzScheduler_Worker-83) XML RPC error in command GetCapabilitiesVDS, the error was: java.util.concurrent.TimeoutException
Ping to it works
/Vlad .
Responses
Hi Vlad, to understand this you need to understand the logical networks in terms of cluster resource. Any logical network (LN) can be used by VMs, depending on the virtual NIC assignment. This means that every host a VM might migrate to (every hopst in the cluster that specific VM belongs to) should have that LN attached and operational. Because otherwise, a VM will migrate to it, and fail to use the LN it requires, and RHEV, after all, is there to serve VMs, and VM operations are the ultimate goal there.
This is why, if a host is missing anything it needs to provide the conditions for VMs to operate, it becomes non-operational.
Hope this makes sense
As for the host with the local domain, this is definitely a separate issue, can you open a new thread for that? Don't want to generate clutter, that will not be readable by other users who might find these threads useful in the future