I don't think this is entirely true, but it might depend on the software release. We're using 12.2(3) and you definitely have to turn on FRF to get the dual FIFO queue; FRTS alone doesn't do it. We have several routers using FRTS with no fragmentation and they only have a single FIFO queue. When I did some testing with this, adding FRF created the dual FIFO queue but then our voice calls sounded worse, even though we weren't actually fragmenting packets!
Weird. Oh well, we've canned our VoIP project anyway. At the moment it just isn't feasible, and no, it doesn't have a feasible successor, either. :-) John >>> "Steven A. Ridder" 1/28/02 2:20:01 PM >>> Fragmenting above a serialization problematic size doesn't create the dual-FIFO queue as certain CCO pages say. It's the frame-relay traffic-shaping command that does. ""Steven A. Ridder"" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > It seems way off. You don't frag a packet above the MTU. As for the subnet > mask as an IP, I can't imagine the router taking it. Why can't he put the > real IP in there? It's been a while since I've seen a customer do FRF.11 C, > though. I'd do his config right and also add the map classes to both > routers. Furthermore, unless he has about 60 or so calls going across, I'd > reduce the reserved BW from 720K to a more reasonable number. > > > ""Erich Kuehn"" wrote in message > news:[EMAIL PROTECTED]... > > I have a customer that we provide frame-relay service to. He is trying to > do > > VoIP over Frame-Relay, while I have quite a bit of experience with IP and > > Frame-Relay, once you put voice into it I get lost. > > > > His problem is that if his circuit goes down at location A, Once Loc B > comes > > back up (to my frame-relay switch) Loc A will not come back up. The only > way > > to force Loc A back up is to shut the interface on the frame-switch to > which > > Loc A connect and then open it back up. Strange I know. He is running > > similar routers at both locations (3660's on 12.2xt code) and the config > are > > nearly identical. With exception > > > > At Loc B under the serial subinterface he has > > > > frame-relay inte.5 255.255.255.252 > > > > Never seen this can anyone explain??? > > > > He also has this in his config at both Loc A and B > > > > Map-class Frame-relay VoIP_FR > > frame-relay fragment 1600 > > frame-relay ip rtp priotity 16384 16383 720 > > no frame-relay adaptive-shaping > > frame-relay fair-queue > > > > At Loc A he makes reference to the map-class under the serial subif at Loc > B > > he does not. > > > > Anyone with some input?? > > > > Thanks > > > > > > Erich Kuehn > > Network Engineer > > Backbone Communications Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=33506&t=32968 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
