I just wanted to confirm my understanding on a sample QoS configuration I created here.
policy-map child
class video
priority percent 10
class voip
priority percent 15
class replication
bandwidth remaining percent 50
class gold
bandwidth remaining percent 15
class vpn
bandwidth remaining percent 30
class class-default
random-detect
policy-map parent
class class-default
shape average 10000000
service-policy child
This policy will be applied to a GigE interface that is connected to a metro ethernet connection with an SLA of 10Mbps. I want to make sure that when the priority traffic is not in using all of it's bandwidth (low amount of phone calls or no video conference going on) that other data classes are given the resources it needs. This is important because we are replication Data Domain traffic over this metro link as well. We are starting to have some backlogging issues. I want to allow Data Domain replication to comsume as much bandwidth as needed during normal business hours, however, when VPN traffic, and other important data traffic need resources, I want my QoS policy to allow it. Right now we shape replication to 7Mbps to allow room for other traffic, but I am considering removing shaping.
Will this policy do the trick? I am pretty sure this will work, but I wanted to double check before I promote something like this to my production network.
Thanks in advance.
