Veritas Storage Foundation 5.1 SP1 for Oracle RAC Administrator"s Guide (5900-1512, April 2011)

Table 4-1
Failure scenarios and recovery strategy considerations
DescriptionCP server issue
After you configure server-based I/O fencing, you may need
to replace the CP servers.
As an administrator, you can perform a planned replacement
of a CP server with either another CP server or a SCSI-3 disk
without incurring application downtime on the SF Oracle
RAC cluster.
Note: If multiple clusters are sharing the CP server, all of
the clusters have to initiate online replacement
independently of each other.
CP server planned
replacement
After setting up I/O fencing using CP servers, you may need
to refresh a CP server registration. As an administrator, you
can perform a planned refreshing of registrations on a CP
server without incurring application downtime on the SF
Oracle RAC cluster: See “Refreshing registration keys on
the coordination points for server-based fencing”
on page 141.
Note: Registration refreshing on a CP server would be
performed in case the CP server agent alerts on loss of such
registrations on the CP server database.
CP server planned
registration refreshing
The applications and high availability of applications on
the SF Oracle RAC clusters sharing a CP server are immune
to the failure of the CP server process or an unresponsive
CP server process, unless a split-brain occurs after a failure
or unresponsive process and before the CP server process
can be restarted.
CP server process failure
The applications and high availability of applications on
the SF Oracle RAC clusters sharing a CP server are immune
to failure of a CP server, unless a split-brain occurs after
the failure and before the CP server can be restarted.
CP server failure
Prevention and recovery strategies
Failure scenarios and recovery strategies for CP server setup
242