On a clustered setup, split-brain might cause the disks to go into a fail state

In a cluster environment, Split-brain might cause disks and volumes to go into a failing state.

During internal testing it is observed with HP MSAP2000 array that after split- brain, disks are going into a failing state. This is due to some of the SCSI reservation commands taking longer time than expected. (2076136)

Workaround: Reactivate the volumes and disks from the VEA console manually and then online the ServiceGroup in case of a VCS setup and online the ApplicationGroup for Microsoft failover cluster from the clustered GUI console.

To avoid this in future, increase the value of registration time in the registry. Create a DWORD key with name RegistrationTimer' and value should be calculated as below: "If SCSI-3 is enabled from SFW {3 seconds for each disk in the disk group; number of disks in the disk group (DG) => sleeping reservation for the going to be online DG}". Value should be specified in Milliseconds. Default value of this key is 7000 (7Secs).