TMS zl Management and Configuration Guide ST.1.1.100226

8-8
High Availability
Configuring High Availability
Configuring High Availability
Before you configure HA, review this summary of HA behavior and function-
ality:
■ Only one HA cluster (two modules) is supported in a single switch chassis.
■ You cannot install HA cluster members that are members of different
clusters in the same switch chassis.
■ All cluster members must be running the same software version. Modules
that are configured for the same HA cluster but that have different
software versions will ignore each other.
■ All cluster members must be in routing mode.
■ If you use the IDS/IPS signatures, it is highly recommended that you have
an IDS/IPS subscription for each module. (See “IDS/IPS and HA” on page
8-7.)
■ The order in which you configure and reboot the members of an HA
cluster is extremely important. (See “Boot Order” on page 8-4.)
■ If cluster members are in different switch chassis, follow these guidelines
(See “Linking Inter-Chassis Clusters” on page 8-5):
• You must ensure that the same TMS VLANs are configured on both
host switches; otherwise, the configuration will not properly fail over.
• You should have redundant connections between the host switches,
and STP or RSTP should be implemented to prevent loops.
• The switch ports that connect the host switches should be tagged
members of the HA VLAN.
• It is highly recommended that other network devices not be placed
between the switch chassis that contain the cluster members to
reduce the risk of link failure.
■ Not all connections fail over from master to participant. (See Table 8-1 on
page 8-6.)
■ ProCurve strongly recommends that you change the default HA VLAN
(VLAN 1) to a dedicated VLAN that does not carry general data traffic.
■ ProCurve recommends that you not associate the HA VLAN with a zone
or give the host switch an IP address on the HA VLAN.
■ ProCurve also strongly recommends that you back up the configuration
of the master before you begin configuring the HA cluster.