[cisco-voip] QoS help

Erick Wellnitz ewellnitzvoip at gmail.com
Mon Mar 19 15:30:34 EDT 2012


I'm going to eliminate our CM based software MTPs with n IOS software MTP.

So, if I configure a trunk with the new MRGL, the trunk will use this to
allocate the MTP, correct?

On Mon, Mar 19, 2012 at 1:20 PM, Wes Sisk <wsisk at cisco.com> wrote:

> stated alternatively, does any phone colocated with ccm have any issues
> with vq when using software media resources such as MTP or CFB?  stream
> statistics from the phone (ii, ??, qrt, or cmr records) would help quantify
> any anomalies in the RTP stream.  If your CMR records show consistent poor
> audio at that remote site then you have a much stronger pointer to the
> culprit.
>
>
>  On Mar 19, 2012, at 11:56 AM, Erick Wellnitz wrote:
>
> CPU is bleow 20% on all nodes.  Network utilization isn't out of the
> ordinary either.
>
>
>
>
> On Mon, Mar 19, 2012 at 10:26 AM, Wes Sisk <wsisk at cisco.com> wrote:
>
>> generally you should deplete mtp resources before voice quality impact.
>> software MTP utilizes network I/O and CPU utilization. we've rarely seen VQ
>> issues caused by sw mtp/cfg but it's not unheard of. if you're using sw
>> media resources then verify the network interface for the CUCM server.
>>
>>  On Mar 16, 2012, at 3:03 PM, Erick Wellnitz wrote:
>>
>> This is unreal.  They tell me it's fixed and tested...but two hours later
>> we get more complaints.
>>
>> These guys have me starting to doubt myself.  CM software MTP wouldn't be
>> causing QoS-like issues, would it?  What about if an SIP trunk with "MTP
>> required" checked and we run out of media resources?
>>
>> On Fri, Mar 16, 2012 at 9:48 AM, Erick Wellnitz <ewellnitzvoip at gmail.com>wrote:
>>
>>> There was an incorrect QoS setting on a different switch on the remote
>>> end.  Problem solved.
>>>
>>> I'll save my gripes about groups not talking to each other for another
>>> day.  ;)
>>>
>>> On Thu, Mar 15, 2012 at 4:10 PM, Erick Wellnitz <ewellnitzvoip at gmail.com
>>> > wrote:
>>>
>>>> I'm not a LAN/Ethernet QoS guru by any means.  Our network group has
>>>> control over QoS and I'm trying to get to the bottom of an issue we've seen
>>>> routing calls over a sip trunk to a location using a 1gig Ethernet
>>>> connection on a DWDM ring.  No firewall is present.  The configuration
>>>> doesn't seem quite right to me for some reason.
>>>>
>>>> Basically what we are doing is using an alternate PSTN gateway for toll
>>>> free traffic until we get additional capacity added to handle the calls.
>>>> Users report garbled audio and an 'underwater' quality of the audio.
>>>> Getting information from the phone on call quality (jitter, etc) may not
>>>> help because the calls complained about are to an external conference
>>>> bridge via tol lfree number.
>>>>
>>>> If this config checks out, I plan to ask the network team if they can
>>>> capture some data for this particular link.
>>>>
>>>>
>>>> Here is the local end QoS config:
>>>>
>>>>
>>>> mls ip cef load-sharing full
>>>>
>>>> mls netflow interface
>>>>
>>>> mls qos map cos-dscp 0 8 16 24 32 46 48 56
>>>>
>>>> mls qos
>>>>
>>>> mls cef error action reset
>>>>
>>>>
>>>> interface GigabitEthernet3/43
>>>>
>>>> description GIG 3/43
>>>>
>>>> ip address X.X.X.X X.X.X.X
>>>>
>>>> ip flow ingress
>>>>
>>>> ip pim sparse-dense-mode
>>>>
>>>> ip summary-address eigrp 1 X.X.X  X X.X.X.X
>>>>
>>>> ip summary-address eigrp 1 X.X.X  X X.X.X.X
>>>>
>>>> ip summary-address eigrp 1 X.X.X  X X.X.X.X
>>>>
>>>> load-interval 30
>>>>
>>>> wrr-queue bandwidth 30 40 30
>>>>
>>>> wrr-queue queue-limit 40 30 15
>>>>
>>>> wrr-queue threshold 2 60 80 100 100 100 100 100 100
>>>>
>>>> wrr-queue threshold 3 60 80 100 100 100 100 100 100
>>>>
>>>> wrr-queue random-detect min-threshold 1 40 60 80 80 80 80 80 80
>>>>
>>>> wrr-queue random-detect max-threshold 1 70 80 100 100 100 100 100 100
>>>>
>>>> no wrr-queue random-detect 2
>>>>
>>>> no wrr-queue random-detect 3
>>>>
>>>> wrr-queue cos-map 1 3 0
>>>>
>>>> wrr-queue cos-map 2 1 1
>>>>
>>>> wrr-queue cos-map 2 2 2
>>>>
>>>> wrr-queue cos-map 2 3 4
>>>>
>>>> wrr-queue cos-map 3 2 3
>>>>
>>>> wrr-queue cos-map 3 3 6 7
>>>>
>>>> mls qos trust dscp
>>>>
>>>>
>>>> Here is the remote end:
>>>>
>>>> interface GigabitEthernet1/36
>>>>
>>>> description GIGABIT E/NET 1/36
>>>>
>>>> ip address X.X.X.X  X.X.X.X
>>>>
>>>> ip flow ingress
>>>>
>>>> ip pim sparse-dense-mode
>>>>
>>>> load-interval 30
>>>>
>>>> no wrr-queue random-detect 2
>>>>
>>>> no wrr-queue random-detect 3
>>>>
>>>> wrr-queue cos-map 1 3 0
>>>>
>>>> wrr-queue cos-map 2 1 1
>>>>
>>>> wrr-queue cos-map 2 2 2
>>>>
>>>> wrr-queue cos-map 2 3 4
>>>>
>>>> wrr-queue cos-map 3 2 3
>>>>
>>>> wrr-queue cos-map 3 3 6 7
>>>>
>>>> mls qos trust dscp
>>>>
>>>
>>>
>> _______________________________________________
>> cisco-voip mailing list
>> cisco-voip at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-voip
>>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/cisco-voip/attachments/20120319/0cca620c/attachment.html>


More information about the cisco-voip mailing list