Motorola 6806800C44B 用户手册

In the active application:

1.

Invoke the HA state handling callback function

2.

 Increment the counter and write it to the local checkpoint

3.

Start the AMF initiated health check (as a result, the health check callbacks are dispatched 
by AMF periodically)

4.

Stop responding to the health check after certain number of health checks

5.

Send an error report with "component failover" as the recommended recovery

In the standby application:

1.

Read the local checkpoint and update the counter value when standby assignment 
happens

2.

Each update to the local checkpoint by the active results in a callback to the standby

3.

Start tracking the protection group associated with the assigned CSI

4.

Start and stop passive monitoring of the component

When the active application sends an error report, the standby application receives the active 
assignment

The new active application resumes incrementing the counter value

The new active application receives the protection group callback and stops tracking this 
protection group.

The previous active application is terminated

A new application is instantiated (as a part of repair)

 The new active component then unregisters and finalizes with AMF

The configuration for the sample application is captured in the System Description File. It 
comprises of the following entities:

z

A service group (SG) that comprises 2 service units (SU) in a 2N-redundancy model
Each SU contains a single component.

z

 A single service instance (SI) is configured to be assigned to the SG

z

 The two SUs come up in the payload nodes (safNode=PL_2_3 and safNode=PL_2_4 
respectively).

The sample application also provides scripts (available in 

/opt/motorola/ncs/dev/source/avsv 

directory on the development host) to control the 

component life cycle. These are:

z

 comp_inst.sh script (to instantiate the sample application) 

z

comp_term.sh script (to terminate the sample application).