Building network automation solutions

9 module online course

Start now!

Can We Trust Server DSCP Marking?

A reader of my blog sent me this question:

Do you think we can trust DSCP marking on servers (whether on DC or elsewhere - Windows or Linux )?

As they say “not as far as you can throw them”.

Does that mean that the network should do application recognition and marking on the ingress network node? Absolutely not, although the switch- and router vendors adore the idea of solving all problems on their boxes.

We have to get used to the fact that we should run our networks like a utility service… and with every service comes Service Level Agreement.

In case of differentiated quality of service, you have to base your QoS implementation on a service definition: the user can generate this much traffic of this traffic class. After having that service definition (or SLA if you wish) in place, stop worrying about traffic marking. Configure ingress policing, shaping, or remarking to prevent SLA abuse, and let your customers mark their own traffic.

The mechanism to use depends on what’s available in your first-hop switching platform and how aggressive you want to be. Remarking is often better (and always more polite) than policing.

Finally, make sure you have a system to monitor the per-port QoS counters, and have a nice graph explaining why the users' high-priority traffic is dropped for the moment they come screaming…

1 comments:

  1. If its the same company then, there is no way to be able to do that

    Specially when you are marking on DC for the traffic to be prioritized on the WAN.

    Generally, WAN Link will have less BW than inside your DC

    For example, if you police for D1 Class at 100Mbps on every server interface, and you have 10000 Servers and your WAN is 1Gbps, then there is no way to protect the WAN with that, how then can we trust DSCP from DC in this case ?

    "No trust" Unless coming from specific source is less risky :)

Add comment
Sidebar