I was under the same impression that "IP RTP PRIORITY" was a "one-liner" which got you out from under having to do alot of fancy Queing if all you needed was the ability to prioritize voice or video. Can you change the bandwidth with the "bandwidth" configuration command to raise the RTP ceiling? The quieng question is a good one, because, allthough queing is supposed to be applied to a phisical interface, you can apply quing in a frame-relay class-map, and different maps can be applied to different interfaces.
Anthony Pace ""Michael Williams"" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > I'm wanting to use the IP RTP Priority command but I'm running into a couple > of situations that aren't covered by any documentation I can find and wanted > to see if anyone else knew or had used this and could offer some suggestions: > > BTW, I'm doing this on a 4700M+ with IOS 12.1(9) with a OC-3 ATM module and > two FastEthernet modules. > > 1) When using the 'ip rtp priority' command you must configure a range of > UDP ports (no problem) but then you *must* specify a bandwidth restriction, > and the range is from 0-2000 Kbps... No where in the docs does it specify > that 2000 Kbps is the maximum you can use, but it seems I'm limited. > > 2) It appears you can only configure this command on a physical interface, > not a subinterface. Example. When trying to use this command on an ATM PVC > (subinterface) it lets you enter the command, but then I get the following > error: > > Router(config-subif)#ip rtp priority 2300 63 2000 > IP RTP: Not enough bandwidth: available 0 needed 2000 > > This ATM subinterface is configured with a 'bandwidth 40000', so I can't > figure out why is says 'available 0'. It does the same on the FastEthernet > subinterfaces (for an ISL trunk). > > If this is the case, then this sux azz because in our WAN core we have an > OC-12, and from the looks of it, I can only give up to 2Mbps priority to RTP > traffic (to the actual OC-12 interface). Since we're looking at running > (the equivalent) of 10 T1 voice trunks through our core, we would want up to > 15Mbps reserved for RTP (not including overhead). The only solution I can > think of is to use the G.729a/b compression to reduce the voice traffic 8:1 > so that 10 voice trunks would only require 1.5Mbps, which isn't a bad > solution, but I can't believe Cisco would be so shortsighted as to only > allow you to reserve 2Mbps of bandwidth for RTP traffic no matter what the > bandwidth of the interface (i.e. OC-3 or OC-12). > > Otherwise, it looks like I'll have to go with PQ or something. Does using > LLQ (which is really just a single Priority queue on top of CBWFQ) have a > similar restriction? My goal in using IP RTP Priority was that I could > configure all interfaces in the voice path with a single command instead of > configuring Priority queues or Classes (with CBWFQ in LLQ) on every > router/interface. > > As a side question: When configuring subinterfaces, can you only apply a > queueing method to the physical interface or can you apply different > queueing methods to subinterfaces? Does the queueing method on the physical > interface "override" the queueing configured on the subinterfaces? > ARRRRGGGGHHHHHH!!! > > Any input or suggestions? > > TIA, > Mike W. Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=42611&t=42555 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
