Dusting off the Per-VLAN Default Gateway Solution

Hope the community can help here...

 

Have a very rigid greenfield application environment I'm building...

 

Essentially the three tiered application infrastructure is divided and segmented

 

I have 2 x LTM 10200 running 11.4 code, I wanted to create 3 partitions to act as virtual instances for each tier, and implement routing domains within each partition to segment individual applications. A bonus is that there is a 1:1 source/destination for the packet flows (Vlan A from Core Cisco will only go to Vlan X)

 

I dusted off the VLAN default gateway solution and everything seems fine, except my external VIPS are not pinging, all internal nodes are fine.

 

Here is the layout

 

"External VIP" vlan VLAN A (192.168.1.254 - HSRP SVI on core cisco nexus)

 

VLAN A (192.168.1.1 - floating self IP on F5)

 

VLAN X (172.16.100.1 - floating self-IP on F5)

 

VLAN X (172.16.100.254 - HSRP SVI on core cisco switch)

 

Issues: 0.0.0.0 virtual server X built to catch traffic to VLAN X, enabled on VLAN A (not working) 0.0.0.0 virtual server A built to catch traffic to VLAN A, enabled on VLAN X (works) **i built the listeners with allow all protocols)

 

I created a default gateway pool, however none of the externals can reach the nexts all the internals can, there are approximately 30 "pairs" of vlans

 

I tried to put in a route to the 192.168.1.254 gateway and the F5 sends a mesage saying that the route overlaps with the implied route from the 192.168.1.1 self-ip

 

All servers, ip's, vlans, trunks are in one of the newly created partitions

 

Anyone have any ideas?

 

Thanks!

 

MB