Product specifications

How Server Failover Works
14
When the server is online, the resource monitor regularly checks its availability and
automatically restarts the server or initiates a failover to the other node if a failure is detected.
The exact behavior can be configured using the Failover Cluster Manager. Because clients
connect to the virtual network name and IP address, which are also taken over by the failover
node, the impact on the availability of the server is minimal.
Network Failover Process
Avid supports a configuration that uses connections to two public networks (VLAN 10 and
VLAN 20) on a single switch. The cluster monitors both networks. If one fails, the cluster
application stays on line and can still be reached over the other network. If the switch fails, both
networks monitored by the cluster will fail simultaneously and the cluster application will go
offline.
For a high degree of protection against network outages, Avid supports a configuration that uses
two network switches, each connected to a shared primary network (VLAN 30) and protected by
a failover protocol. If one network switch fails, the virtual server remains online through the
other VLAN 30 network and switch.
These configurations are described in the next section.
Changes for Windows Server 2008
This document describes a cluster configuration that uses the cluster application supplied with
Windows Server 2008 R2 Enterprise. The cluster creation process is simpler than that used for
Windows Server 2003, and eliminates the need to rely on a primary network. Requirements for
the Microsoft cluster installation account have changed (see
“Requirements for Domain User
Accounts” on page 29
). Requirements for DNS entries have also changed (see “Active Directory
and DNS Requirements” on page 34
).
Installation of the Interplay Engine and Interplay Archive Engine now supports Windows Server
2008, but otherwise has not changed.