Ask the Expert: QoS on Catalyst Switches. - Page 4

ciscomoderator · ‎05-18-2012

With Shashank Singh and

Welcome to the Cisco Support Community Ask the Expert conversation. This is an opportunity to learn from Cisco experts Shashank Singh and Sweta Morga about implementation and working and troubleshooting QoS on Cisco Catalyst 2960, 3650, 3750, 4500 and 6500 switches.

Shashank Singh graduated in 2009 with a bachelor's degree in Computer Science and Engineering from VIT University, Vellore India. Prior to joining Cisco he worked at General Electric as a software engineer. Later on he joined the Cisco Technical Assistance Center as an engineer in October of 2009. He has been working on LAN Switching technologies in TAC since then. Shashank also holds a CCNP certificate. QoS on Catalyst switches is one of the areas of his interest.

Sweta Mogra is a Computer Science & Engineering graduate from VIT University, India. She has worked as a consultant with Tata Consultancy Services before joining Cisco's Technical Assistance Center (TAC) in 2011. She is currently working on LAN Switching technologies and QoS as one of her areas of expertise.

Remember to use the rating system to let Shashank and Sweta know if you have received an adequate response.

Shashank and Sweta might not be able to answer each question due to the volume expected during this event. Remember that you can continue the conversation on the Network Infastructure sub-community Lan Switching forum shortly after the event. This event lasts through June 1, 2012. Visit this forum often to view responses to your questions and the questions of other community members.

jkeeffe · ‎05-23-2012

--

You need to consider bandwidth ratios for the queues. If bandwidth ratios were the same, then they would be treated the same. If bandwidth ratios were different, scheduler will remove lesser number of packets from queue having lower bandwidth guarantee.

So if I want to have more packets in queue 1 dropped than queue 2 I need to have the bandwidth for queue 1 higher than queue 2 - am I understanding you correctly?

smogra · ‎05-23-2012

Packets in queue with lower bandwidth guarantee are more likely to be dropped. so, queue 1 needs to have lower bandwidth to drop more packets.

Jason Dance · ‎05-22-2012

Shashank,

In my limited understanding of dscp and cos, isn't dscp used at layer 3 and cost at layer 2?

Regards,

Jason

Shashank Singh · ‎05-22-2012

You are right Jason. Dscp is derived from the TOS field in IP header (layer 3) and CoS is a field in the 802.1Q tag (layer 2) in ethernet frame.

Regards,

Shashank

slugggerzzz · ‎05-23-2012

I am trying to convert the smartport profile on a CE 500 switch to a 2960 switch.

Can anyone help ?

smogra · ‎05-23-2012

Hi Lewis,

The smartport roles on CE500 are based on the type of devices to be connected to the switch ports. If you had Cisco IP phones role for a port on CE500, in that case you can use auto-qos feature on 2960 which suffice most of the times. Or if you are connecting to other switch and your traffic is coming premarked at the source, you can simply trust the marking (mls qos trust cos/dscp). If however you are planning to mark traffic on the switch, we will have to apply the service policy.

-Sweta

Jason Dance · ‎05-23-2012

On my 3750 switches, I can force the cos for all inbound traffic from the connected device with the following commands:

MLS qos cos 5

MLS qos cost override

What's the equivalent set of commands to override the cost values on a 6500 switch? I am running IOS on both sets of devices.

Regards,

Jason

smogra · ‎05-23-2012

Hi Jason,

The override suffix is not supported on 6500 but you can have an alternative configuration for that. Try configuring a policy map that matches the ingress traffic followed by using set command to configure the desired cos value.

Sample config below:

access-list 10 permit any

Class Map match-all TEST

Match access-group 10

policy-map COS

class TEST

set cos 5

Regards,

Sweta

glgersc · ‎05-23-2012

I have a question on 3560/3750 threshold counters.

This is the default DSCP output threshold map, from the 'show mls qos map' command:

Dscp-outputq-threshold map:

d1 :d2 0 1 2 3 4 5 6 7 8 9

------------------------------------------------------------

0 : 02-01 02-01 02-01 02-01 02-01 02-01 02-01 02-01 02-01 02-01

1 : 02-01 02-01 02-01 02-01 02-01 02-01 03-01 03-01 03-01 03-01

2 : 03-01 03-01 03-01 03-01 03-01 03-01 03-01 03-01 03-01 03-01

3 : 03-01 03-01 04-01 04-01 04-01 04-01 04-01 04-01 04-01 04-01

4 : 01-01 01-01 01-01 01-01 01-01 01-01 01-01 01-01 04-01 04-01

5 : 04-01 04-01 04-01 04-01 04-01 04-01 04-01 04-01 04-01 04-01

6 : 04-01 04-01 04-01 04-01

All traffic is mapped to Threshold 1 in their respective queues.

However, when looking at the 'show mls qos int x/y stat' command, it shows extensive packets in the the other thresholds. Particularly queue 3, threshold 3:

output queues enqueued:

queue: threshold1 threshold2 threshold3

-----------------------------------------------

queue 0: 0 0 0

queue 1: 675 0 24

queue 2: 1 0 0

queue 3: 80 0 1848

Why?

Shashank Singh · ‎05-23-2012

Please let me know the following:

Is QoS enabled on the switch? (sh mls qos)

Is the counter against q3t3 incrementing as of now? Is it possible that this is stale data?

Is is possible that cos-output- map maps some markings to q3t3? Please check "sh mls qos maps cos-out". If not please share the following details:

sh ver

sh mls qos queue-set

Regards,

Shashank

glgersc · ‎05-23-2012

Hi,

Yes, QoS is enabled on the switch, and there is a fairly extensive port based marking/policing policy active on the edge ports, modeled on the Medianet 4.0 guidelines.

The cos-map is default as well:

Cos-outputq-threshold map:

cos: 0 1 2 3 4 5 6 7

------------------------------------

queue-threshold: 2-1 2-1 3-1 3-1 4-1 1-1 4-1 4-1

As is the queue-set:

Queueset: 1

Queue : 1 2 3 4

----------------------------------------------

buffers : 25 25 25 25

threshold1: 100 200 100 100

threshold2: 100 200 100 100

reserved : 50 50 50 50

maximum : 400 400 400 400

This is a 3750X stack running 12.2(53)SE2 code.

This switch may have had some auto-qos config on it at one time, which has since been removed, although not rebooted. Could the counters still be reporting from old config, even though all queueing has been defaulted?

Shashank Singh · ‎05-23-2012

I would believe that the counters date back to the time before reboot. Please see if you can reboot the switch and check again.

Regards,

Shashank

jmoss1 · ‎05-23-2012

I have an installation where the switch access port faces a phone (7942 or 7962) with a subtended PC.

The PC is also running a Cisco softphone (IP Communicator).

The switch config has the option of programming the port to trust either a phone OR a softphone

Switch(config-if)#auto qos voip ?

cisco-phone Trust the QoS marking of Cisco IP Phone

cisco-softphone Trust the QoS marking of Cisco IP SoftPhone

trust Trust the DSCP/CoS marking

I know that the switch uses CDP to recognise either in order to trust it.

I like the idea of CDP checking for a trusted source of QOS markings.

How can I make the port work with both a cisco-phone AND a cisco-softphone, without just trusting all DSCP marking (like from other apps on the PC)?

Or asking another way,

does the "cisco-softphone" setting exclude trusting a cisco phone?
does the "cisco-phone" setting exclude trusting a cisco softphone?

smogra · ‎05-23-2012

Hi,

Unfortunately we cannot implement auto-qos for IP phones and soft-phones together on an edge port. If you don't want to trust the ingress traffic using trust statement, we are only left with the option of using MQC (class-maps). You can create class maps to classify the two types of traffic received followed by mapping them to different queues.

Regards,

Sweta

jmoss1 · ‎05-23-2012

Thanks Sweta for clearing that up.