Hardware Sizing Model For Optimum Application Deployments A Focus On Small And Medium Enterprises In Kenya

Abstract

Sizing is the measuring of system resources required to serve a deployed system and other running systems and processes in their working environments. The sizing model, like any theoretical model, is just an approximation. Research can be addressed differently depending on the system needs, task specifications, and sizing turnaround time. The most common method is to enter all load-related parameters into a modelling tool that is constructed on different hardware using load simulation results. The requirements for hardware and software are determined by the tool's mathematical model. No sizing can be a hundred percent accurate without performing a test on the actual hardware environment to be used. Nonetheless, there is a need to forecast capacity in real life while budgeting equipment, evaluating technological risk, validating technical design, predicting power requirements for production systems, and estimating project costs. These scenarios require a quick way to estimate the requirements of the system. There is a need to develop credible and accurate estimates of sizing when dealing with prospects without spending a lot of time. One of Kenya’s SMEs challenges is the amount of effort and time spent on sizing and also failure to meet the standards mostly due to sizing related issues as they can’t predict the performance and scalability needs of a system before deployment hence, they are rated poorly. This project attempts to explore alternative methods for rapid sizing. It is possible to establish a relationship between sizing factors and CPU/Memory usage by combining the empirical data collected from production system environment and statistical technique. This can be used when deploying new applications.