Overview
Over the period, hardware architecture and capacity of storage have been improved a lot, however improvement of I/O performance was limited resulting lower enhancement of disk storage performance. In terms of presently available disk storage performance enhancement matrices, the prominent ones are Storage Performance Council (SPC), Microsoft Exchange Solution Reviewed Program (ESRP) and Standard Performance Corporation (SPEC) respectively.
Storage performance (as found in “https://www.techopedia.com/definition/28330/storage-performance”) or performance of hard disk (Negi, Singh, & Sahu, 2015) is typically measured in terms of Storage (Stallings, 2009) capacity, throughput and utilization of storage. Storage performance (Tanenbaum, 2008) metrics are detailed out as follows:
Input/output operations per second (IOPS) is measured by the reverse of sum of average seek time and average rotational latency. However, in case of constant average rotational latency, IOPS parameter is dominated by the average seek time.
Transaction processing workload (as found in “https://docs.microsoft.com/en-us/azure/virtual-network/virtual-network-bandwidth-testing”) manages transaction-oriented application over the internet. Meant time between failures (MTBF) is measured by reliability of hardware components and products. MTBF refers the amount of time between two consecutive failure. Meant time to recovery includes the time to find out about the failure of system, diagnose the problem and subsequently repair the same. Response time which is measured by milliseconds, is the time taken by storage system to perform an I/O operation from start to finish. A device has a read or write speed of 32 MBPS means that the device can read or write 32-megabyte record per second. Percent utilization means proper space utilization of storage. Storage performance (Bernstein, Bernstein, Sankar, Diamond, & Morrow, 2009) parameters are further detailed out as follows: Seek time refers to the time consumed by the read or write head on a hard disk to move from one track to another. Rotational Latency is the additional time consumed by the read or write head on hard disk between two desired sectors. Access time is the sum of seek time and rotational latency. Transfer time is depending on the rotation speed (Huerta-Canepa & Lee, 2010) of disk. T = b /rN; where T = Transfer Time; b = Number of bytes to be transferred; N = Number of bytes on a track; r = Rotation speed in revolutions per second; Total average access time can be expressed as follows: Ta = Ts + 1 /2r + T; where, Ts = Average seek time, r = Rotational speed in revolutions per second; T = Transfer time; Cloud computing (Chard, Caton, Rana, & Bubendorfer, 2010) is on demand computer resources which is delivered to user over Internet. It is a set of computing services which typically do executions locally at server side, and perform data transfer remotely across the Internet to exhibit a global approach. It is used around the world, in many forms, by several users.
Cloud computing (Tso, White, Jouet, Singer, & Pezaros, 2013) can run any application, anywhere, anytime, at any scale without having own resources. It makes life easier by saving time and money (Shahri, Hosseini, Ali, & Dalpiaz, 2014). The cloud is elastic (Silberschatz, Galvin, & Gagne, 2005) means it can scale up and down quickly in run-time (as found in “https://www.online-tech-tips.com/computer-tips/check-hard-drive-rpm”). In cloud computing, virtualization is one key technology which can improve the utilization of resources like CPU, network interface, etc. (Jain & Laksmi, 2014). Multicore CPUs, high capacity memory and virtualization technologizes have increased the density of overall infrastructure especially when it is dealing with high capacity disks (Malensek, Pallickara, & Pallickara, 2016).