QoS Models

Two models exist for deploying end-to-end QoS in a network for traffic that is not suitable for best-effort service: IntServ and DiffServ. End-to-end QoS means that the network provides the level of service required by traffic throughout the entire network, from one end to the other.

Key Point

With IntServ, an application requests services from the network, and the network devices confirm that they can meet the request, before any data is sent. The data from the application is considered to be a flow of packets.

In contrast, with DiffServ, each packet is marked as it enters the network based on the type of traffic that it contains. The network devices then use this marking to determine how to handle the packet as it travels through the network.

IntServ

IntServ uses an explicit signaling mechanism from applications to network devices. The application requests a specific service level, including, for example, its bandwidth and delay requirements. After the network devices have confirmed that they can meet these requirements, the application is assumed to only send data that requires this level of service.

Applications in an IntServ environment use the Resource Reservation Protocol (RSVP) to indicate their requirements to the network devices. The network devices keep information about the flow of packets, and ensure that the flow gets the resources it needs by using appropriate queuing (prioritizing traffic) and policing (selectively dropping other packets) methods. Two types of services provided in an IntServ environment are as follows:

• Guaranteed Rate Service This service allows applications to reserve bandwidth to meet their requirements. The network uses weighted fair queuing (WFQ) with RSVP to provide this service. (WFQ is described in the "Congestion Management" section, later in this chapter.)

• Controlled Load Service This service allows applications to request low delay and high throughput, even during times of congestion. The network uses RSVP with weighted random early detection (WRED) to provide this kind of service. (WRED is described in the "Congestion Avoidance" section, later in this chapter.)

Because IntServ requires RSVP on all network devices, it is currently not used as much as DiffServ.

DiffServ

An application in a DiffServ environment does not explicitly signal the network before sending data. Instead, the network tries to deliver a specific level of service based on the QoS specified in the header of each packet. Network devices, typically on the edge of the network, are configured to classify and mark packets according to their source, the destination, or the type of traffic in them. Devices within the network then provide appropriate resources based on this marking. For example, packets that contain voice traffic are usually given higher priority than file transfer data because of the unique requirements of voice.

The Cisco Internet Operating System (IOS) incorporates QoS features that support DiffServ, as described in the following section.