Today, we commonly see IT budget reduction, cost cutting, and barriers for potential network circuit upgrades. In this tough economic situation, the motive of a network administrator should be optimizing the current infrastructure for future accommodations.

Optimization plays a major role when it comes to distributed network architecture and when users are around the globe.

How can we optimize the network with the current infrastructure without adding a hardware or software to the network? We’ll discuss this below.

The network administrator’s role is to ensure that the network is always up and running, and that the performance of the entire network is always running smoothly, even when accessed from a remote location on the other side of the globe.

When accessing the database server or the in-house custom application from a remote site, the main concern is bandwidth over WAN. The bandwidth over WAN or internet on a corporate network is utilized by various business critical applications and non-business critical applications (bandwidth eaters).

Shaping the bandwidth over WAN circuit plays a major role in optimizing network performance. This is where QoS plays a pivotal role. Class-based Quality of Service (QoS) in the Cisco devices helps the network administrator define various classes based on criticality and assigns bandwidth percentage, and priority marking among other things.

If available bandwidth on the network is like an ocean, we could ignore optimizing it. What if it was like preserved water on reservoir? The administrator should be aware of which gate to open for effective utilization. QoS acts as gate, which helps classify the important traffic and how to treat it.

In a distributed network, for cost saving purposes and flexibility, most of the important meetings are conducted using VoIP (Voice over IP) or Video conferencing. Both these applications are heavy in nature and have to be treated effectively over WAN so that the quality of a voice call or video is not compromised.

Setting up a QoS policy for Voice and Video on Cisco routers:

QoS Treatment for Video Streaming or Conferencing:

For classifying the video traffic to appropriate QoS values, the device should be capable of identifying this type of traffic. We can classify the video traffic using DSCP. The appropriate DSCP values are:

Expedited Forwarding (EF):

Packets with EF marking will be given high priority on the network.

Cisco implements EF service under low latency queueing (LLQ). EF keeps the high-priority queue very small to control delay and to prevent starvation of lower-priority traffic. As a result, packets can drop if the queue is full. Usually, EF is most appropriate for VoIP.

Assured Forwarding (AF):

Cisco design guides recommend AF41 (DSCP value 100010) for video. We are not going to get a better advantage if we treat the audio portion of video conference better than the video packets. Therefore, use AF41 as the DSCP value for both voice and video media in a video conference.

Below is the sample QoS configuration for Video Traffic:

Creating an Access list for Video specific traffic:

 Cisco#(Conf)access-list 108 permit ip any any dscp cs4

Cisco#(Conf)access-list 108 permit ip any any dscp af41

Creating Class:

class-map Video_Conference

match access-group 108

Policy Creation:

policy-map QoS-Policy

class Video_Conference

priority 450 30000

Attaching to WAN interface:

Interface Multilink1

Ip access-group 108 out

service-policy output QoS-Policy

 QoS Treatment for VoIP:

Packets with EF marking will be given high priority on the network. This marking is suited for voice only traffic and provides a high quality call.

The sample configuration mentioned below helps treat voice traffic effectively on network over WAN.

class-map match-any VOIP-SIGNAL

match ip dscp cs5

match ip precedence 4

match ip precedence 3

class-map match-any VOIP-RTP

 match ip dscp ef

 match ip precedence 5

 !

 !

 policy-map QOS-Policy

 class VOIP-SIGNAL

 priority percent 5

 class VOIP-RTP

 priority percent 70

 class class-default

 fair-queue

 random-detect

!

!

Interface Multilink

service-policy output QoS-Policy

After implementing QoS policies on the network, administrators can verify the performance of policies using tools that are capable of generating reports on CBQoS. ManageEngine’s NetFlow Analyzer is the one tool that supports CBQoS reporting and generates reports for policies and classes as shown below:

ManageEngine’s NetFlow Analyzer also supports Cisco IP SLA, Medinet and Mediatrace, which helps analyze the performance of VoIP and video traffic on a real-time basis.

We consider QoS the most important component of traffic engineering, which is why we joined a webinar with Cisco to create awareness. Refer to the video below, which talks about QoS and its importance.

Praveen Kumar

NetFlow Analyzer Technical Team

You Can Learn More About the ManageEngine Product Line By Going to manageengine.optrics.com

The original article/video can be found at Quality of Service (QoS): A Good Traffic Engineering Component