10-13-2006 07:49 AM - edited 03-13-2019 03:23 PM
I have been asked to set up qos/ priority queuing and have a question about the best way to handle it.
A traditional PBX is converting voice to ip packets and sending it across a WAN link to another PBX. The PBX is making the signaling as DSCP 48 and the voice as DSCP 40.
At any given time the PBX will allow 4 calls max across the WAN at roughly 48k per call.
The WAN links are T1, and we are told bandwidth utilization is low.
Here are my proposed router config additions:
class-map match-all voice
match ip dscp 40
match ip dscp 48
!
policy-map LLQ
class voice
priority percent 40
class class-default
fair-queue
interface Serial0/3/0
service-policy output LLQ
Are my match ip dscp statements correct?
Is this sufficient to give us basic QOS?
Thanks,
Andy
Solved! Go to Solution.
10-13-2006 08:04 AM
Hi
First it's more 'standard' to use DSCP 46 (ef) for voice streams, and 24 (CS3) for signalling... however this doesn't really matter as long as you prioritize what the PBX is marking.
Secondly, it's recommended that you provision the required amount of bandwidth for media streams in the policy using a 'priority' command, and then allocate some seperate bandwidth using the 'bandwidth' command for your signalling traffic.
This would require creating two classes - one for media, one for signalling...
Otherwise it looks OK.
Regards
Aaron
Please rate helpful posts...
10-13-2006 08:04 AM
Hi
First it's more 'standard' to use DSCP 46 (ef) for voice streams, and 24 (CS3) for signalling... however this doesn't really matter as long as you prioritize what the PBX is marking.
Secondly, it's recommended that you provision the required amount of bandwidth for media streams in the policy using a 'priority' command, and then allocate some seperate bandwidth using the 'bandwidth' command for your signalling traffic.
This would require creating two classes - one for media, one for signalling...
Otherwise it looks OK.
Regards
Aaron
Please rate helpful posts...
10-13-2006 08:31 AM
Thanks for the quick response Aaron. I'll set a second class up and seperate the signal and voice and use a priority cmd to set the voice.
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: