Got a client that today currently has two routers -- each with four IMUXed T1s -- at their location. The circuits connect into a frame relay network, and have PVCs that terminate at data centers. The routers are running DLSw.
At the data centers there are hub-type routers and, behind them, DLSw termination routers. Behind the DLSw routers are token-rings connected to TICs on 3745s. Each token-ring is isolated, and all TICs have the same MAC address. Each customer premise router peers to a different DLSw termination router to protect against NCP failure.
This setup works great. Outbound traffic from the client goes out whichever router answers first, and if one router tanks the only impact is to any sessions riding the DLSw circuits from that router.
But.
The customer is converting to Fast Ethernet using Catalyst switches. This presents a problem. Ethernet switches are not fond of seeing the same MAC address available through two ports. So what do I do? I can't change the MAC address of one of the TICs, because these support thousands of users.
Any help would be appreciated!
--chris
At the data centers there are hub-type routers and, behind them, DLSw termination routers. Behind the DLSw routers are token-rings connected to TICs on 3745s. Each token-ring is isolated, and all TICs have the same MAC address. Each customer premise router peers to a different DLSw termination router to protect against NCP failure.
This setup works great. Outbound traffic from the client goes out whichever router answers first, and if one router tanks the only impact is to any sessions riding the DLSw circuits from that router.
But.
The customer is converting to Fast Ethernet using Catalyst switches. This presents a problem. Ethernet switches are not fond of seeing the same MAC address available through two ports. So what do I do? I can't change the MAC address of one of the TICs, because these support thousands of users.
Any help would be appreciated!
--chris