Looks like it doesn't work for me. After the QoS policy is applied on the LAG, regular traffics in other sub-interfaces run through with no problem, but the sub-interface matching the policy stopped to pass any packets. As soon as the QoS policy was disabled, traffic started to flow again. The matching rule was based on source/destination zones, and a max bandwidth was applied. I guess it's probably not supported on LAG or something I didn't do right. Continue digging... But still, thanks for your replies!
... View more
Hi, this question may have been asked before, but I'm still curious what the best practice is in my situation. Here's what I need to do: a LAG (port-channel) with two 10gig interfaces is carved into mulitple subinterfaces. I'd like to cap the max bandwidth usage for one of the suninterfaces only. Apprently PAN5250 does not support QoS on subinterfaces. So, the question is: if I apply the QoS policy on LAG (physical interface) by defining the matching criteria (e.g. source/destination zones, or IPs), will the firewall find matching packets by stripping the dot1q tags, and put the tagged packets in the desired queue with bandwidth cap? If it works, would it affect other subinterfaces that I don't want rate-limit at all? Or any best practices in similar case when no subinterface QoS support is available? Any experience or real-world implemetation shared would be greatly appreciated. Joseph
... View more