You may also see these messages in the Linux log:
Linux log: READ CAPACITY failed Linux log: reservation conflict Linux log: attempt to access beyond end of device
Error occurs when the master database is corrupted or the clustered file system causes hangs. This problem may occur because the disk array is not properly configured for Veritas. To rectify:
Verify that the disk array firmware is up to date.
(Linux only) Verify that the disk initialization format in the Veritas Volume Manager is sliced.
Correct the disk array connection mode configuration settings so the logical unit numbers (LUNs) reserved for one host are visible to the other host, and verify the SCSI-3 fencing is properly enabled on the array. Generally, these settings are:
Set Host Mode Option = 19 (ON)
Set Host Mode Option = 254 (ON)
Set Host Mode Option = 186 (ON)
Set Host Mode Option = 60 (ON)
Host Mode = 0A
Path Switch Mode (Active/Active) = 1 (Enabled)
No_RSV_Conf Mode = 1 (Enabled)
Persistent RSV Cluster Mode = 1 (Enabled)
Unique Reserve Mode = 1 (Enabled)
See your Veritas documentation for settings specific to your operating system and hardware