Brocade BigIron RX Series Configuration Guide - Manual (Page 523)

BigIron RX Series Configuration Guide 445
53-1002484-04
VSRP and MRP signaling
15
If a VSRP failover from master to backup occurs, VSRP needs to inform MRP of the topology
change; otherwise, data from the host continues along the obsolete learned path and never reach
the VSRP-linked device, as shown in
Figure 82 .
FIGURE 82 VSRP on MRP rings that failed over
To ensure that MRP is informed of the topology change and to achieve convergence rapidly, there is
a signaling process that controls the interaction between VSRP and MRP. When a VSRP node fails,
a new VSRP master is selected. The new VSRP master finds all MRP instances impacted by the
failover. Then each MRP instance does the following:
• The MRP node sends out an MRP PDU with the mac-flush flag set three times on the MRP ring.
• The MRP node that receives this MRP PDU empties all the MAC address entries from its
interfaces that participate on the MRP ring.
• The MRP node then forwards the MRP PDU with the mac-flush flag set to the next MRP node
that is in forwarding state.
The process continues until the Master MRP node’s secondary (blocking) interface blocks the
packet. Once the MAC address entries have been flushed, the MAC table can be rebuilt for the new
path from the host to the VSRP-linked device (
Figure 83 ).
X
X
Path 1 Path 2
MRP
Member
MRP
Master
MRP
Member
MRP
Member
MRP
MRP Member
VSRP Backup
VSRP
Host
MRP Member
VSRP Master
MRP
Member
MRP
Member
MRP
MRP Master
VSRP Backup
VSRP
MRP Member
VSRP Master
Device 1 Device 1
Host