Failure Analysis of Big Cloud Service Providers Prior to and During Covid-19 Period (2210.08006v3)

Abstract: Cloud services are important for societal function such as healthcare, commerce, entertainment and education. Cloud can provide a variety of features such as increased collaboration and inexpensive computing. Failures are unavoidable in cloud services due to the large size and complexity, resulting in decreased reliability and efficiency. For example, due to bugs, many high-severity failures have been occurring in cloud infrastructure of popular providers, causing outages of several hours and the unrecoverable loss of user data. There are prior studies about cloud failure analyses are limited and use sources such as news articles. However, a detailed cloud failure focused study is required that provides analyses for cloud failure data gathered directly from the vendors. Furthermore, the Covid-19 cloud failures should be studied as cloud services played a major role throughout the Covid-19 period, as individuals relied on cloud services for activities such as working from home. A program can be made for this task. As a result, we will be able to better understand and mitigate cloud failures to reduce the effect of cloud failures.