Migrating X.25 over ACC to X.25 over PSI

In the above X.25/ACC HA setup, the two ACC adapters on different systems (System A and

system B) connect to a dual interface panel. Once the system A (Primary node) fails, Serviceguard

package reconfigures System B (stand by node) into a running state.

For more details, refer to the section ACC Service Guard Configuration Notes of X.25/ACC

Installation and Configuration Guide available at:

http://docs.hp.com/en/netcom.html#Advanced%20Communications%20Controller

Note: X.25/ACC HA does not support local failure and Hunt group facility.

X.25/PSI HA Summary

The HA features include Local failover and Remote Failover.

Local Failover

A primary and a standby X.25 PSI adapter connected to an X.25 switch that supports the Hunt

group facility are used to make the X.25/PSI interface highly available on a system. The two serial

lines (a primary PSI adapter connected to one of the serial port of the switch and a standby PSI

adapter connected to the other serial port of the switch) should be configured in the same Hunt

group on the X.25 switch side. When the primary adapter fails, the X.25 link goes down and all

the incoming calls from the calling DTE are directed to the standby adapter in the host system as a

part of Hunt group operation.

Remote Failover

When X.25/PSI is integrated with the Serviceguard package, the X.25/PSI interface becomes

highly available in a cluster of nodes. The X.25/PSI interface migrates along with the ServiceGuard

package to the standby node if the X.25 PSI adapter on the primary node fails. This feature

requires that the HP Serviceguard product (B3936AA) is available and configured on all the nodes

in the cluster. Also, the two serial lines (a PSI adapter of the primary node connected to one of the

serial port of the switch and a PSI adapter of the standby node connected to the other serial port of

the switch) should be configured in the same Hunt group on the X.25 switch side.

Serviceguard uses the x25ifstate command to monitor the X.25 status. If the monitor detects that the

X.25 link it is monitoring is down, it exits. Serviceguard detects that the monitor service belonging

to the X.25 HA package has disappeared, stops the package, and starts the same package on the

second system configured in the X.25 HA cluster. All the incoming calls from the calling DTE are

directed to the adapter in the standby node in the cluster as a part of Hunt group operation.

Figure 7, shows a typical X.25/PSI Remote Failover setup.