User guide

Configuring OSPF OSPF Overview
OmniSwitch AOS Release 7 Advanced Routing Configuration Guide March 2015 page 1-13
Graceful Restart on Switches with Redundant CMMs
A chassis-based switch with two Chassis management Modules (CMMs) can support redundancy where if
the primary CMM fails or goes offline for any reason, the secondary CMM is instantly notified. The
secondary CMM automatically assumes the primary role. This switch between the primary and secondary
CMMs is known as takeover.
When a takeover occurs, which can be planned (e.g., the users performs the takeover) or unplanned (e.g.,
the primary CMM unexpectedly fails), an OSPF router must reestablish full adjacencies with all its
previously fully adjacent neighbors. This time period between the restart and the reestablishment of
adjacencies is termed graceful restart.
In the network illustration below, a helper router, Router Y, monitors the network for topology changes.
As long as there are none, it continues to advertise its LSAs as if the restarting router, Router X, had
remained in continuous OSPF operation (i.e., Router Y’s LSAs continue to list an adjacency to Router X
over network segment S, regardless of the adjacency’s current synchronization state).
OSPF Graceful Restart Helping and Restarting Router Example
If the restarting router, Router X, was the Designated Router (DR) on network segment S when the
helping relationship began, the helper neighbor, Router Y, maintains Router X as the DR until the helping
relationship is terminated. If there are multiple adjacencies with the restarting Router X, Router Y will act
as a helper on all other adjacencies.
Note. See “Configuring Redundant CMMs for Graceful Restart” on page 1-31 for more information on
configuring graceful restart.
Network Segment S
Restarting router X
Router A
Helping router Y
Router B
Router C