L2 Marking for VPLS WAN - DSCP to COS mapping - Advice needed.

[craig.cartlidge1]

Hi, 

I need some advice about a L2 VPLS WAN service we have purchased, we have agreed 3 queues with a 25% priority queue as the highest.

We need to Mark traffic as it exits onto the WAN with 802.1p markings as the carrier does not inspect DSCP, I am guessing we need to apply a policy inbound and outbound for each interface connecting to the WAN to remark the traffic?

Would something like the below suffice? Obviously more tailored to our organisation 

class-map match-all BRANCH-MISSION-CRITICAL

match access-group name MISSION-CRITICAL-SERVERS

!

policy-map BRANCH-LAN-EDGE-IN

class BRANCH-MISSION-CRITICAL

set ip cos 5 ! marking for Mission-Critical traffic

!

...

!

interface FastEthernet0/0

no ip address

speed auto

duplex auto

!

interface FastEthernet0/0.60

description WAN_CONNECTION 10.1.60.1

encapsulation dot1Q 60

ip address 10.1.60.1 255.255.255.0

service-policy output BRANCH-WAN-EDGE-OUT 

service-policy input BRANCH-WAN-EDGE-IN 

!

...

!

ip access-list extended MISSION-CRITICAL-SERVERS

permit ip any 10.200.200.0 0.0.0.255 ! MC Data Server-Farm Subnet

What about the internal LAN subnets? Will I need to apply the remarking policy there too for ingress and egress? wont all this marking and remarking have a large CPU overhead?

Thanks 

[Joseph W. Doherty]

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages wha2tsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

You didn't mention your actual devices.

To be honest, I've never had to mark L2 CoS.  However, as the tagged frame is generated on egress, I would assume you would need to set the L2 CoS marking on interface egress (i.e. your "WAN" connected interface using BRANCH-WAN-EDGE-OUT).

Unclear why you would need BRANCH-WAN-EDGE-IN, on your "WAN" interface, unless your provider strips L3 ToS.

What about the internal LAN subnets? Will I need to apply the remarking policy there too for ingress and egress?

Whether you need "LAN" port policies, ingress or egress, really depends on your overall QoS needs and policies.

wont all this marking and remarking have a large CPU overhead?

Depends on the device and your policies.  For software processed QoS, it does add CPU, but it's usually not an excessive amount.  (For my polices, it seems to increase CPU load by about 5% - but "your mileage may vary".)