The case for QoS over the enterprise medianet WAN/VPN is largely self-evident, as these links are often orders of magnitude slower than the (Gigabit or Ten Gigabit Ethernet) campus or branch LAN links to which they connect. As such, these WAN/VPN edges usually represent the greatest bottlenecks in the network and therefore require the most attention to QoS design. QoS policies on the WAN/VPN serve to control jitter and packet loss.
Classification and marking are typically performed within the campus or branch. As such packets may be assigned QoS treatments on the WAN/VPN edge based on their DSCP markings. Queuing and congestion avoidance are critical QoS functions that are performed over the medianet WAN/VPN.
Three strategic QoS design principles that apply to WAN/VPN QoS deployments include:
(Webtorials membership required. Click here to register or if you forgot your username/password.)
Classification and marking are typically performed within the campus or branch. As such packets may be assigned QoS treatments on the WAN/VPN edge based on their DSCP markings. Queuing and congestion avoidance are critical QoS functions that are performed over the medianet WAN/VPN.
Three strategic QoS design principles that apply to WAN/VPN QoS deployments include:
- Always perform QoS in hardware rather than software when a choice exists
- Enable queuing policies at every node where the potential for congestion exists
- Enable congestion avoidance mechanisms to improve the efficiency of elastic flows
(Webtorials membership required. Click here to register or if you forgot your username/password.)
This "at a glance" paper summarizes key considerations in appropriately a QoS design that ensures good performance over the WAN - where the available bandwidth is going to be orders of magnitude less than on the corporate campus or even in the branch office.
The more technically oriented members of our community will especially appreciate the level of detail contained in this paper.