Can VRPP be used for DCI

[Jklein]

Ok we're designing a network using a MX450, MS 425, MS 350. I'm curious can the Warm Spare VRRP protocol be used for DCI on a WAN consisting of VPLS?

Is there a way to design the switch setup to do fail over at the port level? So if a single port or server goes down it can be routed automatically without moving the whole switch?

[MilesMeraki]

Hello @Jklein, I'm a little confused by your post. In theory if your using HA for MX redundancy and VRRP for MS Redundancy you could use multiple switchport to multiple NIC's in the servers to provide redundancy. - I hope this makes sense.

[PhilipDAth · Accepted solution]

I'm confused by your question.  With regard to servers, if you dual connect them to two switches and configured the NICs on the server as Active/Standby then a single port failure will cause the server to start using the standby NIC in the other server.

[Jklein]

Philip what I mean is let's say that you have a cloud setup. You have a primary facility in Philadelphia your secondary facility is Chicago.

Let's say that customers 1 2 3 4 all have their own servers. They are on ports A B C D on a MS 350. Customer A server fails can we send customer A over to the Chicago facilities MS 350 without having to fail over the whole switch?

In other words can you do Warm Spare on a port level? Or just on the whole switch level?

[PhilipDAth]

You would have to put each customer into their own subnet  so each customer had their own VRRP gateway.  Would you also need some kind of layer 2 circit between the two sites that trunked all the VLANs.

[Jklein]

We actually are since with VPLS like MPLS you can do a spoke and hug setup. So that wouldn't be an issue we can actually guarantee data travels across the carriers fiber back bone anyway. That's the only way you could do warm spare for DCI otherwise the Switches fail over constantly.

[MilesMeraki]

You can't do warm spare at a switchport level with MS switches, only at a Switch level.