LLQ Priority , behaviour of bulit in policer

[HUBERT RESCH]

Hi,

we are discussing about the behaviour of the built-in policer of LLQ.

If we configure following service-policy on a WAN-interface in the case there is no other traffic, is traffic which matches the criteria of class QOS-Class-Realtime allowed to exceed the 20% of the total available bandwidth, or is exceeding traffic always dropped independent if the line is congested or not.

policy-map QOS-Policy-WAN

class QOS-Class-Network-Control

class QOS-Class-Realtime

    priority percent 20

class QOS-Class-Platinum

    bandwidth remaining percent 45

class QOS-Class-Gold

    bandwidth remaining percent 10

class QOS-Class-Silver

    bandwidth remaining percent 25

class QOS-Class-Bronze

    bandwidth remaining percent 10

class QOS-Class-Best-Effort

Thx

Hubert

[Marwan ALshawi]

Hi hubert

The LLQ builtin policed will not take effect until the interface get congested

In other words If the interface is NOT congested the Traffic of the LLQ class cam exceed the LLQ allocated bandwidth

Hope this help

Sent from Cisco Technical Support iPhone App

[HUBERT RESCH]

Thx that what I assumed. If I remember correctly there they noticed in some QOS-guide that you should not configure more than 30% of available BW for the LLQ , I cant find this statement, do you know what could be the reason for this ?

[HUBERT RESCH]

Thx that what I assumed. If I remember correctly there they noticed in some QOS-guide that you should not configure more than 30% of available BW for the LLQ , I cant find this statement, do you know what could be the reason for this ?

[ajay chauhan]

Not sure if there is such limitation for that but MQC will only allow 75% of the interface bandwidth to be allocated by default and to remove this restriction command -Max-reserved-bandwidth 100 should be used under interface.

Thanks

Ajay

[HUBERT RESCH]

This restriction does not longer exist with 15.0 based images

[Marwan ALshawi]

Because the llq will always serviced first in the case of interface congestion it's best practice to allocate around 30 or 33% for VoIP if enough to avoid having the llq class take over other types of traffic

However it is not must if it dosent meet your needs

Hope this help

If helpful rate

[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

NB: That's correct in that Cisco documents this as the reason for their recommendation.  However, keeping your LLQ to under about 1/3 of link capacity also tends to insure adequate bandwidth to avoid queuing delays against itself.

[gfcisco31]

Hubert, checking your config:

policy-map QOS-Policy-WAN

class QOS-Class-Network-Control

class QOS-Class-Realtime

    priority percent 20

class QOS-Class-Platinum

    bandwidth remaining percent 45

class QOS-Class-Gold

    bandwidth remaining percent 10

class QOS-Class-Silver

    bandwidth remaining percent 25

class QOS-Class-Bronze

    bandwidth remaining percent 10

class QOS-Class-Best-Effort

and... not having a restriction (75% of int bandw) your class REALTIME can use as much bandwidth it needs, IF there's no other classes competing for the bandwidth and we have no congestion.

In case of congestion, LLQ classes will have a MAX  while the others will have a MIN bandwidth guaranteed.

LLQ have strict priority, which means that packets will be fwd as soon as it hits the router. In case the queue is FULL (congestion)  the router will try to transmit, BUT it wont have strict priority

I always configure a policing policy map under the real time classes, to make sure when it exceeds the MAX packet will be dropped.

I hope this helps..

Please, rate useful posts.