QoS Policy question, default class set to af43

[wilson_1234_2]

I have a question about a QoS policy with traffic prioritized in a way I am unfamiliar with. The classes are marked with access-lists, but the class-default is set to dscp af43.

I am wondering if the fair-queue setting would affect how the traffic would be prioritized with this policy.

Would all other traffic be higher priority than Signal af41 and Data Af33 (aside from the Voice)?

policy-map QoS-For-VoIP

class QoS-Voip

  priority percent 60

  set dscp ef

class QoS-Signal

  bandwidth percent 4

  set dscp af41

class Data

  bandwidth percent 5

  set dscp af33

class class-default

  fair-queue

  set dscp af43

!

We have a lot of replication traffic that is in the Data queue, but there is some traffic in the Class-default queue that does not seem to be getting priority over the Data traffic.

[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

If the remote side is also a DS3, and there are no logical caps, we wouldn't need to concern ourselves with MPLS vendor's QoS if there was no other traffic also going to the remote site from other sites. If other site traffic to the same remote is possible, that's when we very much want to understand (exactly) the behavior of the MPLS vendor's QoS. (You might also want to determine what other QoS options the MPLS vendor might support, if any.)

So what we know so far, MPLS vendor's QoS supports 4 markings, but we don't know how they treat them when there's congestion and there's traffic with more than one marking.  Also we now know, two of your class marking in your existing policy, the AF33 and AF43 will be remarked to BE and treated alike.  (This likely accounts for some of your observed behavior.)

Your IOS is pre-HQF and FQ has some "interesting" behavior how it interacts with other defined classes.  We really need to fully undertand your MPLS vendor's QoS, but in the meantime, you might try:

policy-map QoS-For-VoIP

class QoS-Voip

  priority percent 60

  set dscp ef

class QoS-Signal

  priority percent 4

  set dscp af41

class class-default

  fair-queue

  set dscp 0

or

policy-map QoS-For-VoIP

class QoS-Voip

  priority percent 60

  set dscp ef

class QoS-Signal

  prority percent 4

  set dscp af41

class Data

  bandwidth percent 1

  set dscp 0

class class-default

  fair-queue

  set dscp 0

[wilson_1234_2]

Thanks for the replies, and just trying to understand here,

So, I am not sure I understand why I should not be concerend with the provider QoS, is this because within the cloud, they are not congested, so QoS would not be in play within the cloud?

Also, with the provider marking AF11, why didn't you use that in your example?

Also, by removing the exising queues that are getting remarked to 0, why are your two examples different than leaving it as it is?

"Understanding the vendor QoS", do you mean behavior when there is congestion on my DS3, or within the MPLS cloud?

[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

Thanks for the replies, and just trying to understand here,

If you don't understand, continue to post questions.

So, I am not sure I understand why I should not be concerend with the provider QoS, is this because within the cloud, they are not congested, so QoS would not be in play within the cloud?

You should be concerned with the provider's QoS.  However, so far, you only know they accept 4 markings, not how they treat them.  When you find out that, then we can reassess your QoS policy.

Also, with the provider marking AF11, why didn't you use that in your example?

Because we don't now how they treat it.  As I've noted, sometimes CS1 and AF1x are deprioritized relative even to BE.  Again, obtain all the details and we can re-assess.

Also, by removing the exising queues that are getting remarked to 0, why are your two examples different than leaving it as it is?

Well in both I moved your signally traffic to LLQ, priority percent 4 vs. bandwidth percent 4.  This because as I noted, FQ in pre-HQF has some interesting behavior.  I also in the example where I retained the data class, reduced its bandwidth setting from 5 to 1.  Again, in this example, I'm trying to allow for how FQ works in pre-HQF.  I also provided an example where we allow your data traffic to also use FQ.  Not sure which will work better, but as you're trying to deprioritize the data class, either of the provided examples may work better for you.

Once again, when you find out exactly what the MPLS vendor's QoS does, we can reassess.

"Understanding the vendor QoS", do you mean behavior when there is congestion on my DS3, or within the MPLS cloud?

I mean what the MPLS vendor does with congestion leaving their cloud.  (Congestion within their cloud is often unusual.)

[wilson_1234_2]

Thanks Joseph, I really appreciate the time you have taken.

I missed that you changed it to Priority.

I will get more information and perhaps have some follow up questions

[wilson_1234_2]

Joseph,

I am afraid I need to ask again about this because I do not follow:

Your comment:

"If the remote side is also a DS3, and there are no logical caps, we wouldn't need to concern ourselves with MPLS vendor's QoS if there was no other traffic also going to the remote site from other sites."

This was where I asked why I shouldn't be concerned. Suppose there is only the two sites using the MPLS cloud, are you saying that I shouldn't be concerned because there typically would be no congestion within the cloud and the traffic would be prioritized as it leaves the CE router?

Would this traffic maintain it's priority across the MPLS infrastructure due to the MPLS not being congested and no QoS being engaged?

Per the provider, the AF11 does not get de-prioritized, it maintains the marking during congestion.
The AF11 class is provided for data application requiring some priority, like SQL.

[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

"If the remote side is also a DS3, and there are no logical caps, we wouldn't need to concern ourselves with MPLS vendor's QoS if there was no other traffic also going to the remote site from other sites."
This was where I asked why I shouldn't be concerned. Suppose there is only the two sites using the MPLS cloud, are you saying that I shouldn't be concerned because there typically would be no congestion within the cloud and the traffic would be prioritized as it leaves the CE router?

Exactly!

Would this traffic maintain it's priority across the MPLS infrastructure due to the MPLS not being congested and no QoS being engaged?

Also correct, assuming that there is no internal congestion within the MPLS cloud.  Usually the bottlenecks are getting into and out of the cloud.

Per the provider, the AF11 does not get de-prioritized, it maintains the marking during congestion. 
The AF11 class is provided for data application requiring some priority, like SQL.

Yea, provider needs to be a little more specific than "some priority"; ditto for other markings.

However, assuming that AF11 has priority over BE, you might try policies like:

policy-map QoS-For-VoIP

class QoS-Voip

  priority percent 60

  set dscp ef

class QoS-Signal

  priority percent 4

  set dscp af41

class Data

  bandwidth percent 1

  set dscp 0

class class-default

  fair-queue

  set dscp af11

or

policy-map QoS-For-VoIP

class QoS-Voip

  priority percent 60

  set dscp ef

class QoS-Signal

  priority percent 4

  set dscp af41

class Data

  bandwidth remaining percent 1

  set dscp 0

class class-default

  bandwidth remaining percent 99

  set dscp af11