Configuration Guide User guide

FastIron Configuration Guide 859
53-1002494-02
Layer 3 behavior with MCT
22
For MCT devices configured with VRRP or VRRP-E, track-port features can be enabled to track
the link status to the core devices so the VRRP or VRRP-E failover can be triggered.
It is not supported to configure several Layer 3 features on VE of the session VLAN. If already
configured, such a VLAN cannot be made the session VLAN.
It is not supported to configure UC/MC routing protocols and the IP follow feature on VEs of
member VLANs. If already configured, such a VLAN cannot be made a member VLAN.
IPv6 configurations are not supported on VEs of session and member VLANs,
Route-only ports cannot be used as CCEP and ICL ports.
Global route-only configuration is mutually exclusive with MCT cluster configuration.
It is not supported to use MCT management interface IPs for tunnel source.
It is not supported to configure static and policy based routes using VE on MCT session VLAN.
Configurations to redistribute connected routes will not advertise IPs on MCT management
interface.
IP addresses on the MCT management interface should not be used for BGP peers on
neighboring devices.
IP addresses on the MCT management interface should not be used for static configurations
on neighboring devices.
The track port feature should be used for VRRP switchover and controlling the validity of SPF
feature.
Up to 64 VRRP or VRRP-E instances are supported on an MCT cluster; however, with Jumbo
enabled, a maximum of 32 VRRP or VRRP-E instances is supported on an MCT cluster.
NOTE
Brocade recommends disabling ICMP redirect globally to avoid unintended CPU forwarding of traffic
when VRRP or VRRP-E is configured.
Layer 3 traffic forwarding behaviors
When one MCT device acts as a VRRP or VRRP-E master router and the peer device is VRRP or
VRRP-E backup, the following behavior will be seen:
Packets sent to VRRP-E virtual IP address will be Layer 2-switched to the VRRP-E master device
for routing.
The VRRP-E MAC address will be learned by the other MCT device that acts as backup router.
Both data traffic and VRRP-E control traffic received by the VRRP backup from MCT client will
need to travel through ICL unless the short-path forwarding feature is enabled.
When both MCT devices act as the VRRP or VRRP-E backup routers, the following behavior will be
seen:
Packets sent to VRRP-E virtual IP address will be Layer 2-switched to the VRRP-E master router
for routing.
The VRRP-E MAC address will be learned by both MCT devices acting as backup routers.
Both data traffic and VRRP-E control traffic will travel through the links connecting them to the
VRRP master.