I'm trying to troubleshoot some performance issues on a 6500 where we have QoS enabled (to support a non-Cisco IP telephony environment).
The following is the interface configuration that uplinks from the core to segment where the IPT equipment is located (pretty much straight from Cisco's QoS SRND).
interface GigabitEthernet4/13
no ip address
speed nonegotiate
wrr-queue bandwidth 5 25 70
wrr-queue queue-limit 10 45 15
wrr-queue random-detect min-threshold 3 50 60 70 80 90 100 100 100
wrr-queue random-detect max-threshold 1 100 100 100 100 100 100 100 100
wrr-queue random-detect max-threshold 2 100 100 100 100 100 100 100 100
wrr-queue random-detect max-threshold 3 60 70 80 90 100 100 100 100
mls qos trust dscp
switchport
switchport trunk encapsulation dot1q
switchport trunk allowed vlan 223
switchport mode trunk
On the port I'm seeing the following output from "show queueing interface..."
Packets dropped on Transmit:
queue dropped [cos-map]
---------------------------------------------
1 586883 [0 1 ]
2 0 [2 3 4 ]
3 0 [6 7 ]
4 0 [5 ]
The packets dropped steadily increases in queue 1 over time.
I don't think this is effecting the priority traffic, but it does seem to effect un-prioritiezed traffic (e.g. pings from the monitoring systems at the NOC, and potentially TFTP transfers).
We have tried tweaking the queue-limit parameters to see if we can minimize the drops, but with little difference.
Any suggestions on how to go about tuning the QoS better? Or to get more detail on how to troubleshoot?
Note: port 4/13 is on a 6724 module with a DFC3BXL. QoS is only enabled for Voice (no Video, scavenger-class, etc.. is currently enabled).
Thanks for the help.
- Klaus
