cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
497
Views
0
Helpful
3
Replies

Voice control signaling traffic from a Cisco phone

johnsos
Level 1
Level 1

Silly question I guess but here it goes. What is the purpose for having the RTP and control signaling traffic marked with different markings and then placing the traffic into different queues? What not treat all traffic coming from the phone the same? Help me understand the reason I would place the different traffic types into the separate queues. Thank you for your help in advance. P.S. is there any document that explains why this is done?

1 Accepted Solution

Accepted Solutions

Chester Rieman
Cisco Employee
Cisco Employee

Well , the RTP(Voice) traffic is much, much more susceptible to jitter and other hiccups that are a minor concern to data protocols. The Idea in j=using seperate queues is to allow the RTP traffic alone to get the pq - this allows the queue to flow smoothly with consistant packet sizes and a constant data stream. The signaling traffic is kept seperate because it really does not need to be in the pq and if it were there it would cause inconsistancies in the types and sizes of packets going through the queue.

http://cisco.com/en/US/products/sw/iosswrel/ps1839/products_feature_guide09186a0080087afb.html

Basically just trying to avoid anything that might cause congestion for the RTP stream

Hope this helps

View solution in original post

3 Replies 3

johnsos
Level 1
Level 1

I posted this same message in the Unified Communications and Video connection as well.

Steven,

Delay on the TCP setup of a phone call is usually more acceptable than loss or delay on a Real Time media stream. Loss of a signaling packet might not be noticeable to the end-user but robotic and choppy voice are, again as a general acceptance level. The RTP paths are put into a LLQ which essentially works as a FIFO queue. The more data types put into that queue the more potential for delay or loss on that media stream.

With that being said, it's still recommended that Signaling receive guarenteed bandwidth. Though this is recommened to use CBQFQ to achieve this. This will insure that LLQ will always be processed and that Signaling will be serviced in a weight fair round robin fashion.

Thanks

Fred

Chester Rieman
Cisco Employee
Cisco Employee

Well , the RTP(Voice) traffic is much, much more susceptible to jitter and other hiccups that are a minor concern to data protocols. The Idea in j=using seperate queues is to allow the RTP traffic alone to get the pq - this allows the queue to flow smoothly with consistant packet sizes and a constant data stream. The signaling traffic is kept seperate because it really does not need to be in the pq and if it were there it would cause inconsistancies in the types and sizes of packets going through the queue.

http://cisco.com/en/US/products/sw/iosswrel/ps1839/products_feature_guide09186a0080087afb.html

Basically just trying to avoid anything that might cause congestion for the RTP stream

Hope this helps

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: