8.2 Queueing | CIR vs minCIR ?

Hi,

Regarding this task, I observed that IOS didn't accept nested policing until I configured mincir command.

Actually, IOS complained about it :

 

Rack3R3(config-map-class)#service-policy output PMAP_VOIP
I/f Serial1/0.34 DLCI 304 class CMAP_VOIP requested bandwidth 200 (kbps), available only 128 (kbps)

Rack3R3(config-map-class)#frame-relay mincir 256000
Rack3R3(config-map-class)#service-policy output PMAP_VOIP
Rack3R3(config-map-class)#exit

Previously, frame-relay cir 256000 command was configured, so I think this is a good moment to ask you guys what's exactly the difference between frame-relay cir and frame-relay mincir commands ?

Thanks,

Ciprian.

Comments

  • Hi,

    I found the below info, but how does this relate to applying a service-policy to the map-class ?

    CIR is the commited information rate and should be = to BW contracted to
    the provider for that particular PVC. EIR is the excess information
    rate and should be = to max burst rate contracted to the provider. In
    general (but not always) CIR + EIR = AR (access rate) where AR is your
    Local Loop speed. But what is Cisco mincir?? well if you are a "good
    user" or "nice guy" you may set a mincir (minimum CIR) to reduce your
    throughput when you receive a congestion notice (BECN). Ex CIR=512
    mincir=384 means that you will reduce your traffic to 384K when there is
    congestion on providers network in order to help the network to deal
    with the congestion.

    Another aspect would be that when enabling traffic-shapping on the frame-relay interface, all VC get shaped by default to 56k. So basically, the router knows that on that DLCI it has 256k available when configuring under the map-class frame-relay just the command frame-relay cir 256000.

    Why on Earth is it complaining about ? [:D]

    Thanks,

    Ciprian.

  • The CIR is known as the target rate, or in other word, how many bits will be sent per second. Frame-relay minCIR is used in conjunction with adaptive shaping, so if you router starts to experience congestion (BECNs received), it will throtle down and reduce the rate of the transmission on that PVC until BECNs cease. The minimum rate that it will drop is controlled by the minCIR; it is half of the configure CIR by default.

  • thanks for your help !

    I think now I found the logic, I also found the following link :

    http://www.cisco.com/en/US/tech/tk652/tk698/technologies_configuration_example09186a0080094af9.shtml

    Note: The sum of the values for priority and
    bandwidth statements need to be less than or equal
    to minCIR for the PVC. Otherwise, the
    service-policy command cannot be assigned to the
    link. minCIR is half of CIR by
    default.

    So I think the whole idea is :

    - you have a BW from the provider of 256k, configured with frame-relay cir command

    - next you set the mincir, for congestion momments, to 256k also, WHICH also affects your reservations, prioritized or not, by adding all the values configured in all classes : priority and bandwidth values that is.

    - because by default mincir=cir/2, so the router can't do the reservations, so mincir has to be raised at a value higher than the above calculated sum (and I assume lower or equal to that cir, because that's what you get from the provider).

    Thanks,
    Ciprian.

     

  • Hi Ciprian,

    minCIR is the rate to which the router will throttle down at a minimum in the case a BECN was received from the frame-relay cloud. It defaults to half of the configured CIR.

    When configuring with MQC for the queuing, it always refers to the minCIR for the bandwidth allocation & you have to be careful when configuring FRTS with MQC.

    Hope this helps! 

Sign In or Register to comment.