Why QoS is Critical to VoIP/UC in Enterprise Networks

IT organizations implement QoS policies to ensure that mission-critical applications receive the highest priority when they need to access network resources. For delay-sensitive applications such as VoIP and unified communications, QoS policies are especially critical and must be constantly fine-tuned to provide the right priority without impeding data applications.

Implementing effective QoS policies requires an understanding of the network infrastructure, how well existing applications are performing across the network, and the idiosyncrasies of QoS in VoIP and UC environments.

To ensure consistent performance of converged voice and data networks, network teams need insight into the composition, volume, and performance of all network traffic before and after implementing QoS policies. The IT organization needs to be able to answer questions such as:

• What is the composition of traffic on each network link?
• Which hosts, conversations, and protocols are consuming the most bandwidth?
• Which users are experiencing poor application performance?
• Where is most of the latency occurring, and why?
• And, are our QoS policies helping us provide consistent, acceptable end-user response times for our critical applications?

Traffic profiling is a critical step in establishing QoS policies on network links. The IT organization needs to identify all applications consuming bandwidth (including who, what, when and where) to ensure adequate bandwidth provisioning. This raw data is available from devices supporting Internet Protocol Flow Information eXport or Cisco (News - Alert) IOS NetFlow, which can be easily exported to network traffic analysis tools.

In addition, IT teams need to baseline the performance of business-critical applications to understand how any changes may impact them. IT organizations armed with the right network and application performance information beforehand can apply effective QoS policies that give higher priority to VoIP and other business-critical applications. Then, IT organizations need to monitor continuously with real-time visibility into how QoS is prioritizing network traffic to determine if applications are being classified appropriately.

Ongoing monitoring of QoS policies is especially crucial for VoIP. The PSTN did a great job of guaranteeing the resources needed for a call and maintaining them throughout that call. By contrast, in an IP network, the resources are shared by many different users. Congestion at a router caused by another user’s large file download can impair the quality of a VoIP call.

Even with QoS prioritization, as the packet traverses the network, routers along the way may alter the marking of the packets. This is particularly true of MPLS networks, where a packet entering a carrier network with one type of QoS marking may leave the network with a different QoS marking. If the QoS marking is different at these ingress and egress points, the quality of a VoIP call is likely to suffer. This is typically a configuration problem that would not be solved by additional bandwidth. NetFlow information can be used to show the breakdown of QoS packet markings for packets passing through a router interface.

As with VoIP, QoS mechanisms are required to enable unified communications applications to run with acceptable quality. The real-time nature of UC applications and their strict latency and jitter requirements dictate the usage of QoS. Due to differences in the protocols and bandwidth consumption, IT teams will need multiple network traffic classes to represent SIP, audio, and video flows. Cisco’s Class-Based QoS is an ideal mechanism to provide the traffic class mappings and queues that UC applications require.

CBQoS allows a network administrator to define application traffic classes that should receive different handling in the network. In a UC environment, IT teams could define separate voice, video, and SIP traffic classes. Each of these classes would be prioritized appropriately by intermediate network devices that carry the UC traffic. Different queues and bandwidth allocations can be used for each class so that the voice, video, or SIP traffic does not become stuck behind other data traffic.

Implementing QoS policies in today’s complex network environments requires lots of care and feeding. To meet the challenge of voice, video and data convergence with effective QoS policies, IT personnel need to track performance before, during and after deployment with a unified view into critical network and application statistics, including traffic composition and application performance. IT

Jeff Hicks is senior software architect for NetQoS (News - Alert) (www.netqos.com), a CA company.

» Internet Telephony Magazine Table of Contents