Nexus 5K Active/Active HSRP issue

[Steven Williams]

I am having an issue which i think is arp related. When I am running my 5548's in Active/Active HSRP, I see some hosts (via tracert) are hitting the individual IP of one as oppose to the other, so when I shut one down I have, guessing, half of my clients being timed out and down, because they are stuck on the one that is down and not routing over to the other one. Can this be fixed?                  

[Reza Sharifi]

Are the 5500 connected together using vPC?

In order for both 5Ks to forward in active/active mode, you need avPC peer link.

HTH

[Steven Williams]

Peer link is up and consistency checks out. Do they share the same arp tables?

[Reza Sharifi]

Yes, when the vPC is up, both switches can response and forward traffic.

You may want to add

add "vpc peer-gateway" to make sure both switches are responding.

HTH

[Ridma Ransara Jayaweera]

Hi

1) In HSRP implemented network when u use tracert it show the actual interface ip rather then VIP this is normal.

As you mention half of the clients goes to switch A and second half goes to switch B.

2)so if you shut down  SVI in one switch this is client ping may  get drops if their destination is in VPC.(becaus traffic goes through peer-link)**Nexus loop prevention rule***

Rid