Installation guide

• HP-UX – MCSG 11.17
• HPIA – MCSG 11.18
• IBM AIX – HACMP 5.4
• Sun Solaris – VCS4.0, SunCluster 3.2
• Linux-AMD – VCS4.1
• Win2003 – Cluster Manager 5.2
There are two ways to upgrade a cluster subsystem:
• Perform a major upgrade, which involves cluster downtime, then shut down and restart all
servers. You must:
a) Run sp_companion suspend, as described in Using Sybase Failover in a High
Availability System .
b) Take offline the resource groups of primary and secondary companions on both nodes.
Ensure that the companion server and corresponding resource groups are not
automatically brought online until the cluster system upgrade is complete.
c) Upgrade the cluster subsystem following the instructions from the cluster system
vendor. You may find options to migrate the current resource groups to the new cluster
version. If such an option is not available (or if the resource groups get deleted or
corrupted), re-create the resource groups and configure them appropriately after you
have upgraded the cluster system.
d) Bring the resource groups online. This should bring primary and secondary
companions online on their respective nodes.
e) Run sp_companion resume as described in Using Sybase Failover in a High
Availability System .
• Perform a minor upgrade, to avoid cluster downtime. Nodes are failed over to other nodes
and upgraded one at a time. For example, assuming ASE1 is the primary companion on
node N1, and ASE2 is secondary companion running on node N2:
a) Upgrade the primary companion:
1.
Relocate the primary resource group to N2 or shut down ASE1. This causes ASE1
to fail over from N1 to N2.
2.
Upgrade the cluster subsystem on N1 following the upgrade instructions provided
by the vendor.
3.
Fail back ASE1 from N2 to N1. See the appropriate cluster chapter in Using Sybase
Failover in a High Availability System for more details on Adaptive Server
failback.
b) Upgrade the secondary companion. If you are using a symmetric configuration, follow
the steps in "upgrade the primary companion," above, for ASE2.
If you are using an asymmetric configuration:
1.
Take the secondary resource group offline and verify ASE2 is shut down. ASE2 is
unavailable during this upgrade.
CHAPTER 10: Adaptive Server Upgrades
Installation Guide 99