Can any one recommend a veritas cluster scsi timeout values for HDS VSP array, there was an incident due to hard failure of HD on VSP which may have caused scsi timeout value to exceeed and oracle to report file system corrupted but no real corruption reported on the HDS side.
any one experiance a similar incident
OS is Solaris 9
do see indications of timeouts
for requests from hosts to drive that spared out and we have seen instances
where an OS can interpret these timeouts as corruption, even though corruption
did not actually occur."