In network continuous increase and speeds of CPU is fostering the new class development of applications which are new depending upon streaming, for example BBC iPlayer and YouTube. Unlike traditional applications of the client server which use transport mechanisms which have best-effort, QoS support is required by these new applications to deliver performance which is acceptable.

Demand misses are also supported by storage systems that happen when the data block is required by program to sustain execution of it.

 Misses of the demand should thus be satisfied promptly or else delays of the server execution will be experienced by applications. As a result it is required to balance the prefetching demands by making use of the clustering as well as the requirement to satisfy promptly misses of demand. To explore the storage servers design this has high performance, at Middlesex University Network Memory Server (NMS) is developed. 

The Network Memory Serve characteristics

With many key goals, NMS has designed. Firstly the NMS actions are considered as atomic by clients of it. These properties indicate that reason is presented to assume distribution which is Poisson for the rate of arrival and the service can assumed also to be exponential because the memory less system is presented with observe to transaction which is preformed previously.