DX Unified Infrastructure Management

Hi,

In our environment, there are some servers which are on cluster and when we have configured CDM on those ,we are getting alerts of some disk not available when these nodes are shifting on clusters.

Like "Q: drive is not available" and this is because the node is active on the other server on cluster.

So how can we overcome this alerts of disk not available on cluster servers with the help of CDM probe.

Thanks in advance.

Have a look to the cluster probe, it works in conjuntion with cdm probe and if the server that you want to monitor has a supported HA cluster manager such as Windows cluster, Veritas cluster, etc, you should be able to monitor clustered disks and avoid those kind of alarms.

Thanks for the reply.
In the CDM probe configuration,do we not have any way to avoid this alerts and configure the clustered disks??

Do we have to use cluster probe for that??

Set fixed default of disk = no in cdm raw configure in disk section ...auto discover of disks must not be enabled in  cluster servers

If auto discover of disks is not enabled, how do new disks added to the cluster get monitored?  I presume they must be manually added to the monitoring profile, if my thinking based on the above statement is correct.

Yes , you must manually enable the monitoring .

Thanks, that's what I was hoping you would say as it matches up with our experience as well.  Was this by design,  oversight or a bug?

Generally cdm doesn't not enable the disks automatically for monitoring ,we need to do by custom package .In case you have cluster probe deployed it will work along with cdm . So fixed default must be no  .

Someone can correct if i have missed any thing 

Not sure I understood the response.  Yes we use cdm probe in conjunction with cluster probe. Which still leads one to believe that new disks added to the cluster should automatically be added to the monitoring profile.  Unless I am not understanding why one would not want that?  From an administrative perspective every manual task is another burden upon the already long list of administrative tasks. 

Not the confuse this thread, but another issue monitoring clusters is when the cluster fails from say node A to node B, we get alarms for disks from node A that we must manually clear.  I would think that if the cluster logic indicates that node B is now active that disk alarms from node A should not be set or at the very least they should automatically be cleared upon successful health status of disks on node B.

When using the cdm probe in conjunction with the cluster probe, the shared disk profiles are set up initially through the cdm probe and then the cluster probe is configured to assure that the shared cluster disk profiles are successfully activated on the new cluster node on a failover.  The fixed disk default section of the cdm probe cannot be active because this forces the cdm probe to detect any new disk added to the node for the cdm probe to start monitoring, but the cluster probe will not be aware of this change.  You will get the disk alarms from the failed node when the disk resources are transferred to another node in the cluster in this scenario.

Hope this clarifies this for you.