ATM Configuration and Troubleshooting Guide

Configuring EMS High Availability over ATM
Types of Recovery
Chapter 696
The EMS HA-ATM software installed on systems A and B allows you to
define that the interface elb0 is the secondary for primary interface
ela0, and elb1 is the secondary of primary ela1. MC/ServiceGuard
must also be installed on both systems; it allows you to define system B
as the adoptive node in the package. Local recovery manages transfer of
IP addresses from one ELAN interface to another within the same
system, by using a daemon called hatmmon that monitors primary ELAN
interfaces. If local recovery is not possible, hatmmon notifies
MC/ServiceGuard, through EMS, that the ELAN interface resource is
down. Then, MC/ServiceGuard starts remote recovery of the package
from one system to another over the Local Area Network (LAN or FDDI
or Token Ring).
Note that, although Figure 6-1 shows an adoptive system with the same
number of adapters as the original system, it is possible to have different
numbers of adapters in each system.
NOTE Although a package runs on only one node within a cluster at a time, the
monitors for a resource run on all cluster nodes for that package. This
means that local recovery actions take place on all of those cluster nodes,
regardless of the package’s status.