cancel
Showing results for 
Search instead for 
Did you mean: 

Brocade Ti24x default-gateway will not work if set to VRRP address on EXOS core

Brocade Ti24x default-gateway will not work if set to VRRP address on EXOS core

kjstech
New Contributor II
Hello,



We have two X690 core switches in our network doing layer 3, and they run VRRP between them. The VRRP address is the default gateway for every vlan with ip forwarding enabled. It works without a problem in our entire network except for the management of two Brocade TurboIron Ti24x switches.



I tried to log into these Brocade switches from my IT workstation and I could not. Perplexed I checked our cacti bandwidth monitoring system, and SNMP was working fine. The difference is that cacti is on the same subnet as the switches. So I SSH into the server and then SSH into the brocades and saw the default gateway was correct, 10.1.0.254. However since we changed our single Cisco 3750 core switch to a new dual core switch (x690's) and migrated that IP address to VRRP, the brocade mgmt can no longer route out. I had to change the default gateway to point directly to one of our core switch stacks, 10.1.0.251, and for our other brocade turbo iron I set that one to the opposite core switch stack, 10.1.0.252 (so at least one switch can be managed if a core goes down).



Does anyone know why this is? When ip default-gateway 10.1.0.254 is set, pings and traceroutes outside of the subnet fail.



I have not seen any other device behave this way. I saw the brocade switching forums are being migrated here, to the hub. That's kind of convienenent because the Brocade switches are talking to Exos switches.
3 REPLIES 3

kjstech
New Contributor II
No never found a solution.

That's weird. No silly things such as some static entries on the brocade side, even though being VRRP and certainly using the same VRID, you should have the same ARP resolution (that's the idea).

Drew_C
Valued Contributor III
Did you ever find a solution to this? We actually ran into the same problem when moving the labs to Extreme's offices. The ICX switch we were using rejects the ARP response from the core.
GTM-P2G8KFN