Administrator Guides EN Owner's manual

41-001391-00 Rev 03 – 04.2012 11-1
Chapter 11
Fail-Over
Overview
The aim of a High-Availability fail-over server configuration is to ensure the highest possible quality-of-service by provid-
ing redundant systems with redundant services that may be employed whenever the primary system becomes unservice-
able. In this version of the Application Server a 1:1 redundancy is implemented (if configured). This means that there is
one primary server, referred to as the Master Server, as well as a default secondary or backup server, referred to as the
Slave Server. These two servers communicate via a proprietary encrypted heartbeat protocol via the network interface as
well as via an optional serial cable connection.
The sysmon service is responsible for starting the SIP service and other necessary services at startup, as well as monitoring
the health of the platform and these services continually. This is true for both single node and high availability pair servers.
In a Fail-Over configuration, the sysmon service also implements the heartbeat protocol and manages the fail-over state
machine. Settings to manage this should be made during the initial Network Configuration. This includes the heartbeat
interval, heartbeat multicast group and maximum missed heartbeats.
When an network error or vital service error is experienced, the sysmon service will activate the secondary (Slave) server to
become 'active'. This means network traffic will be directed to the slave server now and this switch over process should be
invisible to external devices.
Network Requirements
• The HA servers must be connected to the same subnetwork, preferably to the same Ethernet switch.
• Network connectivity between the servers is required during the Network Configuration phase because SSH keys are
exchanged over the network during this process. This should have happened during the initial network installation
phase.
Note:
HA servers cannot be connected via a gateway server.