Application QoS over MPLS links

[deyster94]

My client asked me to do an review of ther QoS that they have on their MPLS routers at their HQ, DR site and remote locations.  They have QoS applied to the serial interfaces on all the routers.  It shows that depending on what the dscp setting is, they get a certain amount of bandwidth. 

Now on the gigabit ethernet interface, they have another QoS policy applied for certain applications.  This is in place in case the MPLS circuit gets saturated and allows their business applications to have priority on the MPLS circuit.  I looked at the QoS settings on a branch router and it looks fine.  They configured a class map with the ports that the applications use and then apply the dscp for those ports in the policy map. 

The part I am having a problem with is how to setup QoS on their HQ and DR site MPLS routers.  Users connect to both of these daily to access applications.  Both of these sites have 25 MB connections and the remote locations have anywhere from a fractional T1 (these are all being upgrade to a full T1 soon) to a 3 MB (2 bonded T1's) connection. 

What I need help with is how would I configure the return traffic from the HQ and DR site MPLS routers to the remote branches.

Here is an example of what they are trying to do:

Remote Branch connections to application on port 8080 that is located at HQ.  They have a policy map that says anything connecting on port 8080 has a dscp of af31.  This policy map is applied to the gigabit ethernet interface.  On the serial interface, anything with a dscp of af31 gets 60% of the bandwidth.

Let me know if you need anymore information.

TIA.

-Dan

[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 whatsoever (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

Unclear you're asking a specific question or what exactly your trying to learn.  "How should" would depend on your service/performance requirements.  In general, though . . .

If you had WAN p2p connections, you can tailor QoS policies for different traffic treatment on each interface.  Normally your major congestion points are WAN interfaces.  Policies may be identical everywhere, or they might differ between HQ and branch or they might differ per site.  Really depends what's the expected traffic that will be hitting a particular interface and your service requirements for that traffic.

When you jump into typical MPLS cloud connectivity, you can often treat traffic that egresses to the cloud similar to p2p egress but assuming you now have multipoint-to-point or asymmetric bandwidth flows, you need to also concern yourself with possible congestion from the cloud (i.e cloud to site).  Often the principle congestion points are the low bandwidth connections to/from branches.

Dealing with cloud egress is often accomplished by using some QoS treatment policy supported by your MPLS cloud vendor.  Often you can chose from different policies and they might be unique per cloud interface.  The way to get your traffic to be treated by the MPLS cloud policy, as expected, is marking your traffic to conform with what's supported by the MPLS cloud vendor's (selected) QoS policy.

For simplicity, any of your egress to the cloud policies might conform exactly to the MPLS cloud vendor's policy or it could be more sohisticated, although it at least needs to insure QoS markings are what the MPLS cloud's policy expects.

In other words, you pretty much can do much whatever you want with QoS policies on your interfaces (within the device's capabilities, of course), but you need to select the best MPLS cloud vendor QoS policy, and then mark your packets to take advantage of it.

[deyster94]

What the client is trying to accomplish is to have their business applications have priority on the MPLS circuits if the circuit is saturated.  The QoS configured on the remote routers is setup to match either IP address or ports that these application use.  My question is how to configure the QoS for the return traffic from the MPLS routers at their HQ and DR locations.

[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 whatsoever (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

deyster94 wrote:
What the client is trying to accomplish is to have their business applications have priority on the MPLS circuits if the circuit is saturated.  The QoS configured on the remote routers is setup to match either IP address or ports that these application use.  My question is how to configure the QoS for the return traffic from the MPLS routers at their HQ and DR locations. 

Two things you need to address.  First, you should prioritize this traffic, as needed, as it egresses the HQ or DR locations.  Second, you need to work with you MPLS vendor to see what QoS policies they might support.  Then you select one for prioritization needs of your traffic on MPLS/branch link (to branch).  Normally, you need to insure your traffic is marked to conform with the MPLS provider's QoS policy (that you've selected).

PS:

BTW, I've been assuming you're not running native MPLS and/or you don't have any direct control over "WAN" devices, other than you WAN edge.