Ospf graceful restart – Brocade Multi-Service IronWare Routing Configuration Guide (Supporting R05.6.00) User Manual
Page 191
Multi-Service IronWare Routing Configuration Guide
163
53-1003033-02
OSPF graceful restart
2. Compare the networks that have the same network address, to determine which network is
more specific. The more specific network is the one that has more contiguous one bits in its
network mask. For example, network 10.0.0.0 255.255.0.0 is more specific than network
10.0.0.0 255.0.0.0, because the first network has 16 ones bits (255.255.0.0) whereas the
second network has only 8 ones bits (255.0.0.0).
•
For the less specific network, use the networks address as the ID.
•
For the more specific network, use the network’s broadcast address as the ID. The
broadcast address is the network address, with all ones bits in the host portion of the
address. For example, the broadcast address for network 10.0.0.0 255.255.0.0 is
10.0.255.255.
If this comparison results in a change to the ID of an LSA that has already been generated, the
router generates a new LSA to replace the previous one. For example, if the router has already
generated an LSA for network with ID 10.0.0.0 for network 10.0.0.0 255.255.255.0, the router
must generate a new LSA for the network, if the router needs to generate an LSA for network
10.0.0.0 255.255.0.0 or 10.0.0.0 255.0.0.0.
OSPF graceful restart
The OSPF Graceful Restart feature provides support for high-availability routing. With this feature
enabled, disruptions in forwarding are minimized and route flapping diminished to provide
continuous service during times when a router experiences a restart.
With OSPF graceful restart enabled, a restarting router sends special LSAs to its neighbors called
grace LSAs. These LSAs are sent to neighbors either before a planned OSPF restart or immediately
after an unplanned restart. The grace LSA specifies a grace period for the neighbors of the
restarting router to continue using the existing routes to and through the router after a restart. The
restarting router comes up, it continues to use its existing OSPF routes as if nothing has occurred.
In the background, the router re-acquires its neighbors prior to the restart and recalculates its
OSPF routes and replaces them with new routes as necessary. Once the grace period has passed,
the adjacent routers return to normal operation.
OSPF Graceful Restart can be enabled in the following configurations:
•
Configuring OSPF Graceful Restart for the Global Instance – In this configuration all OSPF
neighbors other than those used by VRFs are made subject to the Graceful Restart capability.
The restart timer set globally does not apply to Graceful Restart on a configured VRF.
•
Configuring OSPF Graceful Restart per VRF – In this configuration all OSPF neighbors for the
specified VRF are made subject to the Graceful Restart capability. The restart timer set for a
specific VRF only applies to that VRF.
NOTE
If a 32-slot Brocade NetIron XMR or Brocade MLX series system running a version 03.6.00 or later
application image is configured for OSPF graceful restart and intended to be used in switchover or
hitless upgrade, the OSPF dead-interval needs to be changed to 60 seconds on OSPF interfaces to
ensure that the graceful restart process succeeds without a timeout. Instructions for changing the
OSPF dead-interval are provided in