This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.
QoS Has Four Parts
First is the QoS in the routers and switches. The predominant QoS mechanisms used today are DiffServ for the routed infrastructure, and IEEE 802.1p for the switched infrastructure. These mechanisms are in most enterprises' network infrastructure, and are waiting to be turned on if they are not already on. Turning them on isn't hard. What takes some time is determining the right plan for what traffic will use which levels of service, how many levels of service to have, and what markings to use for each class. Also how to map the priority classes across the layer 2 to layer 3 boundary, and how well that plan fits with the services offered by the WAN vendors.
The second part of QoS is classification. Packet streams need to be marked so that the switches and routers can recognize them as higher or lower priority. Think of this as an airline agent who meets you at the door of the airport, and gives you a red jacket or a blue jacket indicating you are flying first class or coach. Now the agents inside the airport can quickly identify which line you should be in without having to engage every person in a conversation.
The third part of QoS is bandwidth management. When we look closely we find that DiffServ and IEEE 802.1p are really class-of-service mechanisms, which means they know how to provide different levels of service but they don't know how to control a rush of traffic or oversubscription. So additional mechanisms are needed to ensure that we don't overutilize the classes of service we have established in the network. This is especially important for voice and video streams, because their quality can degrade quickly when oversubscription occurs.
The fourth part of QoS is monitoring. We have to know that our network is continuing to provide high quality transport for voice and video streams, and we would prefer to know about issues before our users find them through poor application performance. Because voice and video have special needs, new types of monitoring tools are required for this task. This part of the process is both a technology investment as well as a new process to be learned and managed by the IT team.
I'll take a closer look at each of these areas over the next few weeks--watch this space!