Spectralink 84-Series Series Wireless Telephones Administration Guide
1725-86984-000_P.docx
September 2016
227
With failover, all servers (primary and secondary) share the same registration data. In this
scenario, secondary servers support all the features that the primary server supports.
With re-register on failover disabled
—the default behavior—when a handset’s registration
request is diverted to a secondary server, the handset
doesn’t have to register with the
secondary server.
A potential issue with the default failover behavior is that some servers or intermediate SIP-
aware devices may limit a handset
’s functionality if the server hasn’t successfully processed the
handset
’s registration request. With re-register on failover enabled, when a handset’s SIP
request is diverted to a secondary server, the handset will first register with the secondary
server.
Optional behaviors enhance redundancy features. These behaviors include re-registration and
recovery behaviors, as well as a behavior that controls how existing calls
—calls that are
established before a server fails
—are treated.
Re-registration behavior
The handset must complete a new registration with the failover
server before communications can take place between the handset and the failover server.
Recovery behavior
This behavior requires handsets to communicate with the server that
processed the last successful transaction, rather than always with the primary server. If this
behavior is configured, you must set up rules to determine when the primary server is tried
again (for example, whenever the handset has a new request, or after a specific period of
time). The secondary server will remain operational while the handset is trying to re-register
with the primary server
(‘failback’).
Behavior for existing calls
This behavior controls the handling of calls established
through the failed server after failover occurs. When this behavior is enabled, handsets
won’t communicate with failed servers that recover until failback succeeds. This helps avoid
situations in which large numbers of handsets toggle rapidly between servers when there is
an intermittent failure.
The following diagrams show how a network uses the re-registration on failover behavior. In the
diagrams, primary and secondary re-registration on failover-aware Session Border Controllers
(RRoFO-aware SBCs)
are set up so that if a handset
can’t communicate with the primary
RRoFO-aware SBC, the handset can attempt to register, and then communicate, with a
secondary RRoFO-aware SBC.
In the following diagram, handsets are communicating with a primary RRoFO-aware SBC that is
just about to fail.