Packet Loss

This topic describes how packet loss can adversely impact QoS in a network and describes ways to manage packet loss so that QoS is not affected.

• Tail drops occur when the output queue is full. These are common drops, which happen when a link is congested.

• Many other types of drops exist, usually the result of router congestion, that are uncommon and may require a hardware upgrade (input drop, ignore, overrun, frame errors).

The usual packet loss occurs when routers run out of buffer space for a particular interface output queue. The figure illustrates a full interface output queue, which causes newly arriving packets to be dropped. The term used for such drops is simply "output drop" or "tail drop" (packets are dropped at the tail of the queue).

Routers might also drop packets for these other less common reasons:

■ Input queue drop: The main CPU is congested and cannot process packets (the input queue is full).

■ Ignore: The router ran out of buffer space.

■ Overrun: The CPU is congested and cannot assign a free buffer to the new packet.

■ Frame errors: There is a hardware-detected error in a frame—cyclic redundancy check (CRC), runt, giant.

Ways to Prevent Packet Loss

Upgrade the link; the best solution but also the most expensive.

Guarantee enough bandwidth to sensitive packets.

Upgrade the link; the best solution but also the most expensive.

Guarantee enough bandwidth to sensitive packets.

Prevent congestion by randomly dropping less important packets before congestion occurs.

Packet loss is usually the result of congestion on an interface. Most applications that use TCP experience slowdown because TCP adjusts to the network resources. Dropped TCP segments cause TCP sessions to reduce their window sizes. There are some other applications that do not use TCP and cannot handle drops.

You can follow these approaches to prevent drops of sensitive applications:

■ Increase link capacity to ease or prevent congestion.

■ Guarantee enough bandwidth and increase buffer space to accommodate bursts of fragile applications. There are several mechanisms available in Cisco IOS QoS software that can guarantee bandwidth and provide prioritized forwarding to drop-sensitive applications, such as these mechanisms:

— MDDR (on Cisco 12000 series routers)

- IP RTP prioritization

■ Prevent congestion by dropping other packets before congestion occurs. You can use weighted random early detection (WRED) to start dropping other packets before congestion occurs.

These are some other mechanisms that you can use to prevent congestion:

■ Traffic shaping: Traffic shaping delays packets instead of dropping them (Generic Traffic Shaping [GTS], Frame Relay traffic shaping [FRTS], and class-based shaping).

■ Traffic policing: Traffic policing can limit the rate of less important packets to provide better service to drop-sensitive packets (committed access rate [CAR] and class-based policing).

Was this article helpful?

0 0

Post a comment