QOS, DSCP and WMM

  • 1
  • Idea
  • Updated 3 years ago
I kept telling myself to post this up when I have 5 mins. I hope this proves useful. I raised a few support calls telling me to use a marker map and a classifier map. It does not appear to need a marker map at all. The documentation was not that helpful. We use DSCP markings for voice ef (46) and video af 41 (34).

Classifier Map in use shown below



Client to Inbound behaviour

Client marking voice and video using window qos mappings as mentioned above.  Voice ef (46) and af 41 (34)   The Wlan driver uses DSCP and then puts both Voice and Video into the into the video queue when transmitting to the AP as per WMM std. Original Client added DSCP value is correctly maintained onward into the WAN/LAN.

AP to Client

Voice dscp 46 gets to AP via ethernet interface and classfied as voice due to classifier map. It then gets put into Voice Queue.  As a side note DSCP appears to get remarked to 48.

Video af41 get to AP and classfied and then put into Video Queue I suspect this will also have its dscp value remarked also. I just cant find my testing captures at the moment. 

 Just be aware that the DSCP values change client bound it you have a voice team looking at incidents.

I hope this helps someone.

Ben
Photo of Ben Moore

Ben Moore

  • 16 Posts
  • 0 Reply Likes

Posted 3 years ago

  • 1

There are no replies.