Service Manual

This is achieved by configuring same VRRP group IDs to the extended L3 VLANs and VRRP stays active-active across all four
VLT nodes even though they are in two different VLT domains.
The following illustration shows a sample configuration with two data centers:
● Server racks, Rack 1 and Rack 2, are part of data centers DC1 and DC2, respectively.
● Rack 1 is connected to devices A1 and B1 in a Layer 2 network segment.
● Rack 2 is connected to devices A2 and B2 in a Layer 2 network segment.
● A VLT link aggregation group (LAG) is present between A1 and B1 as well as A2 and B2.
● A1 and B1 are connected to core routers, C1 and D1 with VLT routing enabled.
●
A2 and B2 are connected to core routers, C2 and D2, with VLT routing enabled.
● The core routers C1 and D1 in the local VLT domain are connected to the core routers C2 and D2 in the remote VLT Domain
using VLT links.
● The core routers C1 and D1 in local VLT Domain along with C2 and D2 in the remote VLT Domain are part of a Layer 3 cloud.
● The core routers C1, D1, C2, D2 are in a VRRP group with the same vrrp-group ID.
When a virtual machine running in Server Rack 1 migrates to Server Rack 2, L3 packets for that VM are routed through the
default gateway.
The following examples show sample configurations of the core routers.
NOTE: The following configuration assumes that all VLT-related settings are already present on the respective devices.
Sample configuration of C1:
vlt domain 10
peer-link port-channel 128
back-up destination 10.16.140.6
system-mac mac-address 00:00:aa:00:00:00
1082
Virtual Router Redundancy Protocol (VRRP)