Quality of service of a network

Quality of service of a network CONGESTION MANAGMENT: Networks which are designed to support most different traffic types which share a single data path between routers. Congestion management techniques should be considered in such cases to ensure the quality and treatment for the various traffic types. Traffic prioritization especially important for delay-sensitive, interactive transaction based application for instance, for example takes video conferencing that requires higher priority than the file transfer applications. However use of WFQ (weighted fair queuing) ensures that all traffic is treated fairly. Prioritization is most effective in WAN links where the combination of bursty traffic and relatively lower data rates can cause temporary congestion if there is no congestion on the WAN link, there is no reason to implement traffic prioritization. APPLICATION QUALITY: By various parameters used for network quality of service which allows the specification of quality metrics. Such quality of service metrics are monitored and analysed using network quality parameters and important aspect of quality which is always ignored in many cases and overlooked is the behaviour of each and every individual applications in the internet recognised .most applications try to possess a majority part of network resources possible, immediately it may not or influence the quality of an application. NETWORK QUALITY OF SERVICE: The Ability to provide better service to a selected traffic.Quality of service refers to the ability of a network to provide highest quality service to any selected network traffic using various technologies in the latest networks used with a combination of traffic Qos ensures the exact application to be accessed to the resources of the network first, it is a combination of various technologies which allows application to recive request most acceptable and predicted service levels in terms of data and its Bandwidth (throughput capacity) latency variation in jitter, packet loss and delays.Qos provide the best features and more predictable network service by following methods(cisco 2009) Ø Dedicated bandwidth and support Ø Loss characteristics Ø Congestion avoidance and mangment of network Ø Network traffic shaping OPERATION OF QOS: Ø Qos distinguishes the traffic and splits it with very exact timing requirements Ø It improves resources in the network so that all the traffic reaches the specified destinations reliably and faster Ø It doesnt create any bandwidth it simply manages it effectively to meet the application requirements (OPNET 2008) Here the working of qos follow a series flow of simple rules in order to execute the file or a packet and at first the packet is pass through the classification which means that the following packet must be in which group does it fit then the packet is moved on to the next level which is the pre-queuing which means that the following packet is arranged in a order according to the source and the destination numbers and then it is again moved on to the next level which is the queuing and scheduling its main work is to set the stream to the destination point so that it automatically enforces the bandwidth allocations then moves on to the last level which is the post queuing which increases throughput on the lower speed links so that it reaches the destination point within the time it has decided and thus the working of the qos is completed. ADVANTAGES OF QOS: Ø Dedicated bandwidth Ø Controlled network latency and jitter Ø Improved loss characterstics Ø Control and predictability beyond best effort concept (OPNET NETWORK 2008) NEED OF QOS IN A NETWORK: All packets in a network are given equal access to all the resources , the priorty acn only be given when we can distinguish data packet from a voice packet. For a company network to effectively use the network and its resources it must identify which network traffic is the appropriate and which is not to allocate the right recourses to effectively support the traffic streams . The priority should be given to all the data streams to failing to do so can create low quality of voice and data, because most of the audio and visual application is delay and jitter sensitive. Only a good Qos can give the audio and video packets the best priority access. Qos evolution: The networks which are used with Qos enabled are called the Best -effort network. Where each and every packet is treated in the same way which is significantly important in the network design for preferable results, when even space in the cpu is available these type of networks work prominently. There are two major architectures of qos, both the architecture have different approach and different ways of quality of service in their own preferable ways. The main architecture of Qos is integrated services and differentiated services. Traffic characteristics that Qos tools can effect 1) Bandwidth 2) latency 3) jitter 4) compression 5) queuing BANDWIDTH: Bandwidth refers to number of bits per second that can reasonably be expected to be successfully delivered across same medium. firstly bandwidth techniques are mainly used to describe or define network traffic and to do this we can set some special type of techniques which are mainly used by the network devices such as blue coat for instance and by using this we can setup a type of service (TOS) in the packet of the internet protocol header and by using the required information of the type of service we can set the device in action and allow the traffic to flow in different direction and for another instance we can choose some factors to involve this issue such as audio clarity, if the sound is audible in two different direction then it is said to be working in perfect condition and this can be measured by the MOS (which means the mean opinion score) , so by this we can rate from 1-5 which means that 1 scores the worst and 5 scores the best and the average typical analogue call whi ch rates from 4.1 to 4.7 and a avg cell phone ranges from 3.5 to 3.9 and if we considered the VOIP CALLS ranges from the 4.0 to 4.4 so which means that the VoIP also has the best average rating in audio clarity and if we comes to reliability comes to 99.999 so which is more reliability to the customers and the usage of them is also more and the techniques used in this kind of symphony which is sophisticated and the consistency is also more which is mixed out with the qos so this is in short all about bandwidth in qos LATENCY: It is mainly specified as some serious problem to qos which is also known as the delay and if we speak in technical words its the same amount of time taken by a packet from source to destination and if the latency and the bandwidth are defined they are for speeding the network so for instance we can say that a normal average person can hear a call upto 250 ms approx and 200 ms in sensitive person ear so if the call doesnt return in that range the caller is going to be disappointed. JITTER: It also refers to the same problem but for connectionless or wireless so which is also called as a delay which must be a serious and most inappropriate but business customers such as for company which is dealing with important calls so in this here if we see the database server is connected to some system and the employees are storing some information to the database systems and in between there are calls to handled and the calls doesnt seems to go through the cloud of the similar database server so we can find the disturbance which are caused by the latency and jitter so in order to watch and control the traffic we must maintain the jitter in control and the specification for this is it must be less than 100 ms for the communication less than 100 ms for normal database because if the jitter took place a bit slow in voice band it doesnt have a problem but if it takes under the packet series then it might be some kind of serious issue so in order to reduce this we must all keep the j itter in control. bandwidth refers to number of bits per second that can reasonably be expected to be successfully delivered across same medium Compression: Either the payloads or the headers compressed by reducing the total number of bits required to transmit the data. Call Admission control: reduces all the overall load of the network by denying any new incoming voice and video calls. Queuing: Qos refers to a broad collection networking technologies and techniques. The goal of Qos is to provide guarantee on the ability of a network to deliver predictable results. Qos involves prioritization of a network, Qos is a method to guarantee the bandwidth relationship between individual or application or protocols. Qos refers to the capability of a network to provide a better service to selected network traffic over various underlying technologies including frame relay,(Cisco systems Inc. 1999). There are seven Qos mechanisms and tools that are used to implement Qos in a computer network. Ø Congestion management Ø Congestion avoidance Ø Control admission control Ø Shaping and policing Ø Bandwidth reservation Ø Link efficiency Ø Classification and marking 1.1.3 Aim of Qos is to provide a dedicated bandwidth sufficient to deliver for the service of the applications by controlling latency and jitter, and by reducing data loss. Network characteristics managed by quality of service. Category of quality of service mechanisms: 1) Admission control 2) Traffic control 1.2.1 Admission control: Gives the information of number of users and the applications used to network resources it allows only specific users and resources which can be used in a network segment (subnet). 1.2.2 Traffic control: It controls and regulates the data flow by classifying and marking the packets based on priority and by stopping traffic, service class assigned to a traffic flow which evaluates the quality of service treatment, the traffic receives. 1.2.3 Call admission control: It provides the overall quality for all the networks, it controls the voice disturbance from the voice traffic, and video from the other video traffic 1.3.2 Classification and marking: For defining a quality of service identifying the traffic is the first case involved in the procedure which is treated either differently or preferentially which is done by classification and marking. 1.3.3 Bandwidth reservation: Bandwidth reservation provides guarantee to the bandwidth, i.e., bandwidth is provided whenever needed without reserving it for a specific application or flow in a network. 1.3.4 Shaping and policing: The significant issues related to quality of service in a network are solved by traffic shaping , the delay and loss in a network are solved by traffic shaping which is called as(egress blocking)The data which is sent or received are measured by traffic shaping and traffic policing . traffic policing in a network is used to remove all the excess packets which helps to overcome the policed rate. The excess packets are again en-queued by shaping. Both shaping and policing prevent the traffic from exceeding the bit rate defined. Link efficiency: Link efficiency is used increase the quality of service of a network,particularly if the given note continuously increase the bandwidth rate on a network which causes the sudden change in behaviour of the network. Which slows down data applications significantly.if the load exceeds the given bandwidth for a period of time the application slows down completely or even stops down ata particular point because of the continuous queues which can be avoided by using the link efficiency. 1.3.6 Congestion management: Whenever queuing occurs in a network congestion management gives the ability to rearrange the packets. 1.3.7 Congestion avoidance: Congestion avoidance tools are used to avoid congestion. It enables queue to avoid congestion. Whenever the rate of transmission load and offer load exceeds the line rate send by various senders.Queues are formed which may cause congestion. The queue are managed by congestion avoidance tool by dropping the packets randomly which are selective which reduces the congestion level. WHY DO WE NEED QOS Each and every packet will be given equal access to resources when we not consider the QoS policies. We cannot give voice priority if we cannot tell a voice packet from a data packet. In order to utilize its network resources efficiently for a company, so to support those traffic streams it must identify which network traffic is critical traffic and allocate appropriate resources to support those traffic streams. Otherwise, the result could be intermittent with voice quality complaints. Applications like voice and video are delay and jitter sensitive. Voice packet will be given first priority access to the interface queue, when we use a good QoS policy. For example, both FTP and an voice packet arrive at the same time at an outbound router interface. When we not consider the QoS policies the voice packet may need to wait in the queue until the FTP packet has been processed out the interface. This may results delay of unacceptable amount of delay into the voice path which will depends upon the interface speed. Traffic flow with out qos (global knowledge,whitepapers,[2],author:gardiner2008) The voice packet could be given priority first over the FTP packet when we use QoS configuration. The FTP packet may be fragmented to make sure that the voice packet does not show any excessive delay, if the interface speed is less than T. QoS Evolution The main reason for the wide transformation of best-effort models to more complex differentiated services models is by privately owned enterprise and service providers networks, the meaning that the network gives different applications differing levels of service. 2.6.1 Congestion Management (Queuing): QoS queuing tools provide you with a variety of queuing methods. Queuing tools define a number of queues. The queuing tools are as follows: a. Priority Queuing(PQ) b. Custom Queuing(CQ) c. Weighted Fair Queuing(WFQ) d. Class-Based Weighted Fair Queing(CBWFQ) e. Low Latency Queuing(LLQ) f. Modified Deficit Round-Robin(MDRR) a. Priority Queuing: Priority Queuings most distinctive feature is its scheduler. PQ schedules traffic such that the higher-priority queues always get serviced, with the side affect of starving the lower-priority queues. With a maximum of four queues, called High, Normal, and Low, the complete logic of the scheduler can be easily represented, as shown in figure. (Cisco systems, 1999) b. Custom Queuing: As with most queuing tools, the most interesting part of the tool is the scheduler. The CQ scheduler reserves an approximate percentage of overall link bandwidth to each queue. CQ approximates the bandwidth percentages, as opposed to meeting an exact percentage, due to the simple operation of the CQ scheduler. The CQ scheduler performs round-robin service on each queue, beginning with queue 1. CQ takes packets from the queue, until the total byte count specified for the queue has been met or exceeded. After the queue has been serviced for that many bytes, or the queue does not have any more packets, CQ moves on to the next queue, and repeats the process. (Morgan, 1991) c. Weighted Fair Queuing: Weighted Fair Queuing differs from PQ and CQ in several significant ways. The first and most obvious difference is that WFQ does not allow classification options to be configured. WFQ classifies packets based on flows. A flow consists of all packets that have the same source and destination IP address, and the same source and destination port numbers. So, no explicit matching is configured. The other large difference between WFQ versus PQ and CQ is the scheduler, which simply favors low-volume, higher-precedence flows over large-volume, lower-precedence flows. Also because WFQ is flow based, and each flow uses a different queue, the number of queues become rather large up to a maximum of 4096 queues per interface. And although WFQ uses tail drop, it really uses a slightly modified tail-drop scheme- yet another difference. (Cisco systems Inc, 1999) d. Class-Based WFQ: CBWFQ is most like CQ, in that it can be used to reserve minimum bandwidth for each queue. It does differ from CQ in that you can configure the actual percentage of traffic, rather than a byte count. CBWFQ is like WFQ in that CBWFQ can actually use WFQ inside one particular queue, but it differs from WFQ in that it does not keep up with flows for all the traffic. e. Low Latency Queuing (LLQ): LLQ combines the bandwidth reservation feature of CBWFQ with a PQ-like high priority queue, called as Low Latency Queue, which allows delay-sensitive traffic to spend little time in the queue. But first, this section begins with WFQ, which uses a completely different scheduler. Table 2: Comparison of Queuing Tools. (Odom W, et al, 2005) Tool Maximum Number of Queues Classification Capabilities Queue Service Algorithm/ End Result of Algorithm PriorityQueuing (PQ) 4 IP ACL Input interface Fragments Strict service; always serves higher-priority queue over lower queue. Custom Queuing (CQ) 16 IP ACL Input interface Fragments Serves a configured number of bytes per queue, per round-robin pass through the queues. Result: Rough percentage of the bandwidth given to each queue under load. Weighted Fair Queuing (WFQ) 4096 Automatic, based on flows, (Flow identified by source/destination address and port numbers, plus protocol type.) Each flow uses a different queue, Queues with lower volume and higher IP precedence get more service; high volume, low precedence flows get less service. Class-Based Weighted Fair Queuing (CBWFQ) 64 IP ACL NBAR Same as CB marking Service algorithm not published; results in set percentage bandwidth for each queue under load. Low Latency Queuing (LLQ) N/A Same as CBWFQ LLQ is a variant of CBWFQ, which makes some queues priority queues, always getting served next if a packet is waiting in that queue. It also polices traffic. Modified Deficit Round-Robin (MDRR) 8 IP precedence Similar to CQ, but each queue gets an exact percentage of bandwidth. Supports LLQ mechanism as well.