QoS for voice data and voice signalling

[LucaSalvatore_2]

What is the reccomended bandwidth allowance for voice signalling?

On our WAN routers we generally have  15% of the bandwidth allocated to anything marked with ef which is all the voice data.

But what is the best approach for signalling?

Should the config use the priority command or bandwidth command in the policy map (for signalling)?

And if we allocate 15% using the priority command to voice data, what percentage should signallying get?

thanks,

[sean_evershed]

Have you seen the Enterprise QoS Solution Reference Network Design Guide that can be found here?

http://www.cisco.com/en/US/docs/solutions/Enterprise/WAN_and_MAN/QoS_SRND/WANQoS.html

Here is a sample from the three class voice and data model.

As you can see the recommendation is to use priority for voice and bandwidth percent for signalling.

This document recommends 5% for signalling. This depends however on the type of phones you are using and what other business requirements you may have.

 class-map match-all Voice

  match ip dscp ef            ! IP Phones mark Voice to EF

 class-map match-any Call Signaling

  match ip dscp cs3           ! Future Call-Signaling marking

  match ip dscp af31          ! IP Phones mark Call-Signaling to AF31 

!

policy-map WAN-EDGE

  class Voice

   priority percent 33        ! Maximum recommended LLQ value

   compress header ip rtp     ! Optional: Enables Class-Based cRTP

  class Call Signaling

   bandwidth percent 5        ! BW guarantee for Call-Signaling

  class class-default

   fair-queue                 ! All other data gets fair-queuing

!

Please rate posts that are helpful.

Thanks