Brocade Multi-Service IronWare Switching Configuration Guide (Supporting R05.6.00) - Manual (Page 756)

726 Multi-Service IronWare Switching Configuration Guide
53-1003036-02
Restart global timers
19
The process of syncing routes between a new MP and its LPs using the new timers are illustrated in
Figure and described in the following steps.
1. The MP switchover from active to redundant MP begins.
2. The system max-hold- timer starts.
3. The IGP/BGP restart process begins.
4. If the IGP/BGP restart process is completed before the system max-hold-timer expires, the
system max-hold-timer is cancelled and the protocols-converge-timer starts.
5. Once the protocols-converge-timer expires, the MP syncs up forwarding information with the
LPs.
6. If the system max-hold-timer expires before the IGP/BGP restart process is completed, the MP
syncs up forwarding information with the LPs at that time and the protocols-converge-timer is
never started.
FIGURE 190 MP to LP re-syncing process
Configuring the graceful-restart max-hold-timer
This timer defines the maximum hold time before a management module syncs up new forwarding
information to interface modules during the restart process. While the default value of 300
seconds will work in most cases, if a device is loaded with a very large number of routes and
OSPF/BGP peering adjacencies you might want to fine-tune your device’s performance by
increasing this value.
The value of this timer can be set using the command shown in the following.
Brocade(config)# graceful-restart max-hold-timer 500
Syntax: [no]graceful-restart max-hold-timer hold-time
The hold-time variable is the maximum number of seconds that a management routing module
waits before it syncs up new forwarding information to the interface modules during a restart. The
range for the hold time is 30 – 3600 seconds. The default time is 300 seconds.