How are you identifying "voice traffic"?

I'll probably have to think of a hack to make what you are
asking for to work.


ea> On Thu, May 31, 2007 at 01:11:38PM -0400, Paul Stewart wrote:
> > Hi folks...
> > 
> > A while back (month or so) I posed a few questions about Policy Based 
> > Routing - thinking that was best way to conquer a new challenge... now 
> > I'm not so sure so looking for input.
> > 
> > Here's the layout and what I want to accomplish....
> > 
> > Customer premise has a Cisco 3662 router.  From the 3662 we have 2 
> > TI's leaving and 2 Ethernet connections leaving towards a 6509 back in 
> > our data center.  The 2 T1's go to a remote POP where they terminate 
> > on a Cisco 3640 router.  The Cisco 3640 router connects to a Cisco 
> > 7206VXR which in turn connects via TLS back to the same 6509 in our 
> > data center.  The ethernet connections leaving the customer site from 
> > the Cisco 3662 connect directly back to the 6509 with speeds of 6 Mb/s 
> > X 800Kb/s each.  The T1's are full
> > 1.544 Mb/s.
> > 
> > So, one router at customer premise that needs to connect back to one 
> > router in our data center using 4 paths.  The pair of T1's and the 
> > pair of ethernet ports should be "bonded" or load balanced.  
> > Traditionally this has been done via OSPF/CEF on our side of things.
> > 
> > We want all VOIP traffic passing between the customer site and our 
> > data center to travel via the T1 circuits and all Internet traffic to 
> > go via the ethernet connections.
> > 
> > I'm looking for the best routing protocol in this scenario that will 
> > allow me to use route-maps (or other alternatives) to identify source 
> > IP and destination IP subnets and apply priority.  At the same time if 
> > the "far end" of each connection is unavailable then I want the traffic to
> "fallover"
> > to the other connections as a backup automatically.
> > 
> > I had though at one point that OSPF would be ideal but I'm not aware 
> > of a way to apply a route-map to OSPF specifying only certain traffic 
> > prefers a certain path.  We do this all the time with BGP so I though 
> > maybe iBGP could be applied here but I have the feeling that there is a
> better solution....
> > 
> > Open to ideas and appreciate it...
> > 
> > Paul
> > 
> > _______________________________________________
> > cisco-nsp mailing list  cisco-nsp@puck.nether.net 
> > https://puck.nether.net/mailman/listinfo/cisco-nsp
> > archive at http://puck.nether.net/pipermail/cisco-nsp/
_______________________________________________
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Reply via email to