Thanks for your suggestions Matthew. I was running source specific multicast in the core, so RP's are not required. I like your suggestion regarding hypervisors and may use that on some other topologies that give me grief. Here I wasn't loosing devices and hence config. I think periodic hello's on any of the many protocoks were being missed due to go slows and hence ospf would go down, followed by ldp, then bgp, then mdt data, followed by ospf up etc etc. Could probably tune this out if convinced it was the cause, but wasn't. Thanks again
Matthew Holden <[email protected]> wrote: >It has been a couple years since I built one, but I think the missing link >was a static RP config between the PE's outside the VPN > >Also, if you're having stability problems consider separating the PE's into >a separate instance of the hypervisor from the CE's and the P. That way if >one config or chassis type is unstable, when it goes down it doesn't take >all the unsaved configs with it. > >On Tue, Feb 21, 2012 at 4:57 PM, Mark Beynon <[email protected]> wrote: > >> Afternoon all, would very much appreciate your help with this. I have >> knocked up several adhoc topologies (gns3) for multicast over layer 3 vpn, >> and all went as I would expect until I decided to try anycast RP. This >> didn't work. Fearing it was my lack of understanding, I reduced the >> complexity to non layer 3 vpn and implemented MSDP. No problems. Cleared >> that down, any cast config, no problem. Back to layer 3 vpn anycast but >> can't get it working. It could be the instability of the gns3 causing >> issues as topology often flaps, but I don't think so. >> >> My latest topology (I tried a few), consisted of five routers.. >> >> R4 --- R2 --- R1 --- R3 --- R5 >> >> R2, R3 = PE >> R1 = P >> R4, R5 = CE >> >> Layer 3 vpn config with mdt default using SSM in core. >> >> OSPF between CE and PE's. >> >> To start with, I set up R2 as BSR from lo1, and RP from lo2. >> >> All routers successfully learn BSR and RP and multicast works. >> >> Then I add lo2 to R3, with same IP address as R2, and configure it as RP >> candidate. Now, local RP's on either side see their local multicst joins >> etc, and multicast works on either side of the core, but not across core >> (as expected) due to unknown sources. >> >> So I configure MSDP between the RP's, connecting from lo1 interfaces. >> Msdp sa-cache correctly shows active sources, but multicast cross core >> stops working. For completeness I configure another BSR candidate on R3 >> from lo1 but it makes no difference. >> >> Am I missing some complication from the layer 3 vpn? >> Is this a valid configuration? >> >> My ldp, ospf, bgp and mdt do become quite unstable in the core due gns3 >> resource I think, so this might account for it, but I would expect to see >> some sign of life. Mdt/msdp does come up so at that point I would expect >> success... but nothing. >> >> No config attached so this is all to check my suggested config would be >> valid. Worked fine in the non layer 3 vpn config, so I assume it is... >> >> Thanks for your time if you have been kind enough to read all this! >> _______________________________________________ >> 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 >> > > > >-- >H.G. Wells once said that civilization was in a race between education and >catastrophe _______________________________________________ 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
