ATM Configuration and Troubleshooting Guide

Configuring EMS High Availability over ATM
Types of Failures Managed by EMS HA-ATM
Chapter 6104
NOTE For HP-UX 10.20, when hatmmon is started, it causes an immediate
transfer to the secondary adapters if the primary ELAN interface is not
operational. Also, if IP addresses are already associated with an
operational secondary ELAN interface, hatmmon leaves those addresses
on the secondary interface.
For HP-UX 11.X, if the primary ELAN interface is not operational when
hatmmon starts, the daemon cannot transfer the primary IP addresses to
a secondary ELAN interface. Instead, hatmmon logs an error for IP
address transferring and notifies EMS that the resource is down. You do
not need to restart hatmmon, but the primary ELAN interface must be
operational before any future local recovery can occur.
Failures Managed by Remote Recovery
As mentioned earlier, remote recovery manages primary system failures,
failures of all ATM adapters in a system, and failures of all ATM
switches or cables connected to the adapters in the primary system. All
of these failures are covered in the description of system failure below.
System Failure
Remote recovery for a system failure requires a specific kind of
configuration. The configuration, and what happens if a system fails, are
discussed below.
Necessary Configuration We recommend you provide backup
adapters within the same system where possible. Then, configure the
system as part of an MC/ServiceGuard High Availability cluster.
What Happens in Case of Failure Applications are started on
another system if one of these failures occurs:
• No backup adapters are configured in the primary system.
• All the backup adapters in the primary system fail.
• The system fails.
• All links fail.
• All switches fail.