I am working on a set of Stratix 5400 (IE4000 equivalent) switches (22) that to me appear to have an inconsistent qos configuration. I'm not sure if this is an actual "correct" configuration, a type of configuration that I'm not familiar with, or perhaps it's a cosmetic bug.
The switches are running IOS 15.2(7)E3. All of the switches have the global "auto qos srnd4" command. On many of the interfaces, the commands "auto qos trust dscp", as well as commands for associating user-defined QOS policies "service-policy input CIP-PTP-Traffic" and "service-policy output PTP-Event-Priority". But, it seems like the auto qos and user-defined policies applied to the same interface ought to be mutually exclusive, right?
If I run the "show auto qos" command & reference one of these interfaces, the output is "AutoQos is disabled", which again makes me think this is some kind of cosmetic. But any attempt to modify the QOS config on such interfaces results in a raft of QOS error messages about policies and/or classes already being applied and/or in use & not able to be deleted.
Even defaulting one of these interfaces & then trying to re-apply the QOS commands results in errors. I've also tried removing ALL of the auto qos and the service-policies from all the interfaces, and then removing the auto qos srnd4 gobal command to try and start with a "clean" QOS slate, but I still get errors when trying to re-renter any of the qos or service-policy commands.
Can anyone tell me what I have here? Here's an example interface config:
interface GigabitEthernet1/9
switchport trunk allowed vlan xxx,yyy
switchport trunk native vlan zzz
switchport mode trunk
switchport nonegotiate
logging event trunk-status
logging event bundle-status
logging event spanning-tree
logging event status
load-interval 30
udld port aggressive
auto qos trust dscp
alarm profile ab-alarm
spanning-tree link-type point-to-point
spanning-tree guard root
service-policy input CIP-PTP-Traffic
service-policy output PTP-Event-Priority
ip dhcp snooping trust
