01-16-2017 08:03 AM
We have two Brocade ICX 6610 connected in a stack to 2 Cisco ASR 1000 Routers, which are running High Availability via Application Redundancy mode. The default route for the switch is via the Cisco routers.
Management connectivity to the switch stack is intermittent, which is caused by the ARP entry for the Virtual IP of the Cisco devices moving between the 2 Brocade switches. When we lose management connectivity, we can see the ARP entry on the secondary switch even though the Cisco ASR's have not failed over and the first unit is still the active partner.
The switches are in Layer 3 mode with an SVI assigned to a VLAN which is in common with the ASR's.
The most bizzare thing is that we have this same setup at multiple sites, with it working on some sites and failing on others.
01-24-2017 08:29 PM
I would have to see the configureation to know more.
You refer to the 6610's as switches and mention the two Cisco ASR 1000 Routers are running HA via App Redundancy Mode.
I honestly have no idea how you set this up, but I presume it uses something like the Cisco HSRP for the virutal gateways.
Anyway, I presume you have two links perhaps LACP on the brocade side and PAgP on the Cisco side. check that.
Sounds like you have some crazy ARP flood, which is common in a loop. Not saying this is the cause, but it is something I can think of off the top of my head.