Abstract (EN):
The quality of the digital experiences delivered by engineers and their business success depends on empowering developers and operators with an effective method for continuously assessing a system's health, diagnosing possible issues, and recovering from service outages. In other words, monitoring is essential to ensure the quality of an application. However, monitoring best practices may not be apparent to everyone and, most of the time, are not sufficiently explained or documented to be learned quickly and communicated effectively. Therefore, practices usually lack formalisation and a standard structure that would make all of them easy to communicate and share among practitioners. To tackle this issue, this paper describes three proactive monitoring practices as design patterns: Liveness Endpoint, Readiness Endpoint and Synthetic Testing. Design patterns provide enough structure and detail to be easily reused by practitioners and have space to accommodate different needs and quirks depending on the usage context. The proposed patterns are based on existing literature and tools, stemming from industry best practices that are further detailed and adapted to design patterns. Relations to existing monitoring patterns are also analysed to point the reader to more patterns that complement the ones proposed in this work. © 2022 Owner/Author.
Language:
English
Type (Professor's evaluation):
Scientific