HP A5500 HI Switch Port Level Redundancy

[Pankaj4HP]

Hi Team,

I have two HP A5500 HI Series switch configured using IRF link & hence 1 is master & another is slave.

Now, I need to know about port level redundancy that whether it will work or not.

I have two links for a node. One link is on switch1 which is connected to handover router1 & another is on Switch2 whih is connected to handover router2. 

At my HP Switch a vlan is configured & an IP is assigned to that VLAN. Then that VLAN is assigned on both the switch ports where links are connected. 

At the other end, vrrp is running on both handover routers.

At my end i have a ststic route pointing to the VRRP virtual IP of handover roueters as gateway.

Currently connectivity is working through my primary switch. But i need to know that if primary link goes down, will traffic pass through secondary link? 

Due to business criticality, i can't test this redundancy in production environment. I am asking this question, bacause only VLAN is configured & assigned on both ports, then I'm not sure whether traffic will failover to secondary or not.

Please share your expert coments or let me know if I need to prvide more details. Thanks..

Regards.

Pankaj

[Linkk]

Hi,

I will recap what I understand, please correct me if I'm wrong:

You have an IRF with two HPE A5500HI switches. You have two routers configured with VRRP. 
Switch 1 is connected to router 1 and switch 2 is connected to router 2. 

Are the router connected to each other in any other way? With this configuration, you would not have redundancy if any of the IRF members fail. VRRP works in master/standby and the other router would not become master if the HP switch 1 fails. Also you network can not connect to the VRRP master if the only uplink is down and there is no other way.

To get redundancy, a good way would be to make a link-aggregation between the IRF and the two routers. So that router 1 has one LA to both members (one link to switch 1 and one link to switch 2) and router 2 has the same configuration with an LA to both members (one link to switch 1 and one link to switch 2). 

[KSHKND]

l would suggest the usage of interface tracking on the routers, that way if one of the link / switch fails the Active router would go to initialize state and your backup router should start passing traffic.