Acme Packet 4500
 

PLA-Support>Sun Systems>CommsGBU>Session Delivery Network>SN-SND: Acme Service Provider
 

In this Document

Goal

Solution

Applies to:

Goal

This document describes how to use the gw-heartbeat functionality to monitor the status of the gateway.

Solution

In a scenario where the active SD still has positive link state on its interface(s) but a link between the next-hop switch and its router/gateway has been severed, there is no way the SBC can determine that the gateway is not reachable anymore because the ARP entry in the arp cache is still valid for this gateway. Heartbeat mechanism is a means by which the SD can determine its layer 2 health by sending periodic ARP messages to its provisioned IP gateway(s). Failure to receive a response to a number of these messages will cause the SBC's health to decrement (a configurable value) and trigger a switchover.

If the standby SBC's path to the gateway is working, it is determined by both members in the HA to be the healthier system, and should be responsible for processing calls. Recommended settings for enabling gateway heartbeats under network-interface are:

gw-heartbeat
state enabled
heartbeat 10
retry-count 3
retry-timeout 1
health-score 30

The configuration above will send heartbeat messages every ten seconds. Should any request fail to receive a response, it will retry three times, each with a timeout of one second. Should the gateway fail to respond to the four ARP requests, the system will decrement its health by 30 points enough to cause a switchover in most circumstances. It is also advisable to increment the health-score value by one with each new heartbeat configuration (30, 31, 32, etc.). In the event of an unreachable gateway, a unique decremented value will be easily identifiable (e.g., if the health drops from 100 to 69, it is due to unreachability on the gateway configured with a health-score of 31).

This solution has no attachment