If R2 and R3 are stubs or you are covering everything transiting those, I'd probably try Rob's second option. Match the input interface of BB1 and apply to the appropriate outbound interface and do your marking.
Bob -- Sent from my iPhone, please excuse any typos. On Jan 11, 2013, at 7:26 PM, Rob Pool <[email protected]> wrote: > Why not create a class map that matches an acl for each of the two > destination routers and add those class maps to a policy map applied inbound > on r1? > > I think another possible option would be to create a class map that matches > on the input interface, place that class map in a policy map to be applied to > each of the outbound interfaces. That way you won't miss any networks trying > to create an acl. > > Sent from my iPhone > > On Jan 11, 2013, at 4:08 PM, Joe S <[email protected]> wrote: > >> I've got what will probably sound like a stupid QOS question, but I can't >> come up with something I'm confident is the 'right' answer. >> >> Let's say I've got this setup >> >> R2 g0/0 -> g0/0 R1 g0/1 -> BB1 >> >> ...and then there's a serial connection from R1 to R3 >> >> And what I'm told is that any traffic coming from BB1 and destined for R2 >> networks should be marked with DSCP13 and traffic from BB1 networks >> destined for R3 should be set for DSCP EF. >> >> Because traffic from other directions could end up going through R1 to R2 >> and R3 I can't just setup an outgoing policies on the relevant interfaces >> to mark the DSCP. IIRC, last time I messed with this I couldn't tag the >> incoming traffic as originating from BB1 and then match that tag outgoing >> on the various interfaces to set the DSCP. >> >> The two ways it seems to me that this could work would be... >> >> 1. Mark the QOS on the stuff incoming from BB1 at an arbitrary value, match >> that arbitrary value on the way out, and then set the correct value >> >> or >> >> 2. Looking at the subnets coming from BB1, use those as your match clauses >> on the R1 outgoing interfaces and use those to set the DSCP. >> >> Now, number one up there seems risky because it's assuming you know all >> DSCP values and your arbitrary value could potentially step on them. Plus, >> you're setting values you weren't told to. Number two, which I'm guessing >> would be the "correct" answer seems to be a matter of making some huge ass >> assumptions. >> >> Are there more ways that anyone can suggest? What's the "correct" answer? >> _______________________________________________ >> For more information regarding industry leading CCIE Lab training, please >> visit www.ipexpert.com >> >> Are you a CCNP or CCIE and looking for a job? Check out >> www.PlatinumPlacement.com >> >> http://onlinestudylist.com/mailman/listinfo/ccie_rs > _______________________________________________ > For more information regarding industry leading CCIE Lab training, please > visit www.ipexpert.com > > Are you a CCNP or CCIE and looking for a job? Check out > www.PlatinumPlacement.com > > http://onlinestudylist.com/mailman/listinfo/ccie_rs _______________________________________________ For more information regarding industry leading CCIE Lab training, please visit www.ipexpert.com Are you a CCNP or CCIE and looking for a job? Check out www.PlatinumPlacement.com http://onlinestudylist.com/mailman/listinfo/ccie_rs
