cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1691
Views
5
Helpful
2
Replies

Redundancy and HSRP Dual Uplinks

BoJack Horseman
Level 1
Level 1

Hello all,

currently my company network is colocated at a Data Center providing us two uplink fiber cables with HSRP enables routers behind them, so that of the uplinks one is active and another one is passive. To make it work we did put a switch on top of the racks receiving both the cables, and the result is that everything is working as intended. 

Now, we'd like to provide redundancy for this switch (a Cisco 2960-X) by adding another one. The solution we came up with would be to add another identical switch, connect the two with a port-channeled trunk, and then make NIC bonding (active-backup mode) on the linux servers below, so that we can cross connect these servers to the two independent switches (one cable to the first, the other to the second).

My question is: if the HSRP change the router's state from active to passive so that the previously inactive fiber uplink becomes the new active, would the switch be able to cause the backup interface (the one connected to the switch who was connected to the inactive fiber) on the servers to come up as well?

Will this solution works? Or if the uplinks change from active to passive the servers would keep trying to reach the passive fiber thus causing a down state of the network?

Thanks a lot for any advice and answers.

1 Accepted Solution

Accepted Solutions

chrihussey
VIP Alumni
VIP Alumni

If I understand you correctly the final goal would be to have your servers do NIC bonding and have the active link go to one switch and the backup to the other switch. The switches are inter-connected with a port-channel and each switch has a fiber uplink to the Data Center HSRP routers.

If this is all a single layer two domain I don't see why it would not work.

1- If the primary NIC failed on a server, the backup NIC would go to its connected switch, then through the port channel to other switch then to the active HSRP router.

2- If the active HSRP router/link failed the primary server NIC should route through its connected switch, across the port channel, to the other switch and then to the other (now active) HSRP router.

3- Finally, if done correctly and the HSRP routers are communicating through your switches, if a switch failed, it should all stay up and work across the other switch.

Hope this makes sense and helps.

View solution in original post

2 Replies 2

chrihussey
VIP Alumni
VIP Alumni

If I understand you correctly the final goal would be to have your servers do NIC bonding and have the active link go to one switch and the backup to the other switch. The switches are inter-connected with a port-channel and each switch has a fiber uplink to the Data Center HSRP routers.

If this is all a single layer two domain I don't see why it would not work.

1- If the primary NIC failed on a server, the backup NIC would go to its connected switch, then through the port channel to other switch then to the active HSRP router.

2- If the active HSRP router/link failed the primary server NIC should route through its connected switch, across the port channel, to the other switch and then to the other (now active) HSRP router.

3- Finally, if done correctly and the HSRP routers are communicating through your switches, if a switch failed, it should all stay up and work across the other switch.

Hope this makes sense and helps.

Thank you very much chirussey,

I confirm that it worked perfectly! Thanks a lot for your advice and help!

Review Cisco Networking for a $25 gift card