As we know cloud services are affected by multi-factors. One of them is the nature of workload. My question is actually about the stochastic nature of workload.
This is a very good question. In addition to the excellent answer given by @Mohamed-Mourad Lafifi, there is a bit more to add.
In cloud computing, an element of uncertainty is inherent. In other words, uncertainty is a persistent characteristic of cloud computing. Hence, the stochastic nature (seemingly random probability distributions) is a biproduct of cloud computing.
Exploration of the uncertainty assocated with cloud computing is explored in
Dear Tanvir, I'm working with systems for the past several years. Since my first year on university from then to now on as a research fellow. I got different experiences through working and would love to share with you. Most of the systems follow random probability distributions because of their different work aspects in different time distribution processes. Time is the most significant dependent metric for every systems. By varying different workloads systems have to follow different queuing theories or models.
To know more you have to read or gone through the papers that are given above by Sir James F Peters andSir Mohamed-Mourad Lafifi. Some of mine are given below as an attachment.