sh int counters errors

[Nisterio]

Hi All, 

I am seeing this "Outdiscards" on all the interfaces of 2960-X Switch.. Could someone point me in the right direction so I know how to fix them? I run a simple router-on-a-stick topo with dot1x, QoS and rspvt+  enable on all the ports/vlans.  

OFF-SW01#sh int counters errors

Port Align-Err FCS-Err Xmit-Err Rcv-Err UnderSize OutDiscards
Gi1/0/1 0 0 1 0 0 347443
Gi1/0/2 0 0 1 0 0 246623
Gi1/0/3 0 0 0 0 0 21592
Gi1/0/4 0 0 1 0 0 523678
Gi1/0/5 0 0 1 0 0 200460
Gi1/0/6 0 0 0 0 0 179933
Gi1/0/7 0 0 1 0 0 215674
Gi1/0/8 0 0 1 0 0 513777
Gi1/0/9 0 0 1 0 0 511430
Gi1/0/10 0 0 0 0 0 0

interface GigabitEthernet1/0/1
description End User Port
switchport access vlan 56

witchport mode access
switchport voice vlan 32
ip access-group ACL-DEFAULT in
logging event trunk-status
srr-queue bandwidth share 1 30 35 5
priority-queue out
authentication event fail action next-method
authentication event server alive action reinitialize
authentication host-mode multi-auth
authentication open
authentication order dot1x mab
authentication priority dot1x mab
authentication port-control auto
mab
snmp trap mac-notification change added
mls qos trust device cisco-phone
mls qos trust cos
dot1x pae authenticator
dot1x timeout tx-period 6
flowcontrol receive desired
spanning-tree portfast edge
end

[Remi Astruc]

Hello,

It is likely your QoS scheduler is dropping egress packets.

We need to determine which queue is dropping, what is the related setting and which cos/dscp packets are guilty. Please output the following commands:

sh mls qos int gi1/0/1 stat | beg drop

sh mls qos int gi1/0/1 queueing

sh mls qos queue-set

sh mls qos map | beg outputq-threshold

Remi Astruc

[Nisterio]

Remi. 

Thanks for this input, I have attached the file with the requested output 

[Remi Astruc]

Hello,

So you probably have bursts from default traffic with COS and/or DSCP 0.

Assuming the default traffic is well assigned to the expected queue/threshold as designed, you can try to raise the SRR bandwidth for the affected queue 3 (queue 4 is also slightly concerned):

srr-queue bandwidth share 1 30 60 9

And increase the related buffers to absorb the peaks:

mls qos queue-set output 1 buffers 5 20 50 25

Then clear all the counters and see if drops decrease.

Remi Astruc

[Leo Laohoo]

Turn off QoS.

[Nisterio]

I added the srr-queue shared and srr-queue shaped weights on the interfaces but still seeing some discard. from my little understanding of QoS, I think the traffic with COS 0 / dscp 0  (which is the most part of the traffic) goes into the queue 3 threshold 3. Now I don't know the best way to configure the cos-map. Any input on that?

mls qos map cos-dscp 0 8 16 24 32 46 48 56
mls qos srr-queue output cos-map queue 1 threshold 3 4 5
mls qos srr-queue output cos-map queue 2 threshold 1 2
mls qos srr-queue output cos-map queue 2 threshold 2 3
mls qos srr-queue output cos-map queue 2 threshold 3 6 7
mls qos srr-queue output cos-map queue 3 threshold 3 0
mls qos srr-queue output cos-map queue 4 threshold 3 1
mls qos srr-queue output dscp-map queue 1 threshold 3 32 33 40 41 42 43 44 45
mls qos srr-queue output dscp-map queue 1 threshold 3 46 47
mls qos srr-queue output dscp-map queue 2 threshold 1 16 17 18 19 20 21 22 23
mls qos srr-queue output dscp-map queue 2 threshold 1 26 27 28 29 30 31 34 35
mls qos srr-queue output dscp-map queue 2 threshold 1 36 37 38 39
mls qos srr-queue output dscp-map queue 2 threshold 2 24
mls qos srr-queue output dscp-map queue 2 threshold 3 48 49 50 51 52 53 54 55
mls qos srr-queue output dscp-map queue 2 threshold 3 56 57 58 59 60 61 62 63
mls qos srr-queue output dscp-map queue 3 threshold 3 0 1 2 3 4 5 6 7
mls qos srr-queue output dscp-map queue 4 threshold 1 8 9 11 13 15
mls qos srr-queue output dscp-map queue 4 threshold 2 10 12 14
mls qos queue-set output 1 threshold 1 100 100 100 100
mls qos queue-set output 1 threshold 2 100 100 100 400
mls qos queue-set output 1 threshold 3 100 100 100 400
mls qos queue-set output 1 threshold 4 100 100 100 400
mls qos queue-set output 2 threshold 3 3100 3100 100 3100
mls qos queue-set output 1 buffers 5 20 40 35
mls qos queue-set output 2 buffers 5 15 50 30
mls qos

[Remi Astruc]

Hi,

No, setting the shape weights override the share configuration and behaves the opposite way. Please remove shape statement if just added.

Also you have not set the queue-set buffers like suggested above.

Ensure to have the following on int g1/0/1 for example (without any "shape" command):

srr-queue bandwidth share 1 30 60 9

And in global, increase the related buffers to absorb the peaks:

mls qos queue-set output 1 buffers 5 20 50 25

According to your previous output, you should also raise threshold 1 of queue 4:

mls qos queue-set output 1 threshold 4 100 100 100 400

Then clear counters and check the values.

clear counters

clear mls qos interface statistics

You will surely see improvements.

Remi Astruc