I am currently planning a somewhat ambitious series of articles on scalability.
In the next couple of weeks and months, I am planning to write about 12–14 articles, as well as a few side notes that will delve deeper into specific topics. I’ll cover everything from a single service instance to distributed systems and the available resources and methods, as well as how to scale the organisation building the systems. After all, the same mechanisms that apply to technical systems can also be applied to development teams.
To kick off the series, I’ll share a true story: An enterprise portal started displaying error pages with just 50 concurrent users.
https://steinacker.name/articles/fifty-users/
The first substantive section reminds us of a fundamental concept from computer science studies: Little's Law. Here, I’d like to demonstrate a practical application of the law and, incidentally, explain how to avoid the unnecessary provision of service instances when the issue is actually just a configuration problem.
https://steinacker.name/articles/littles-law-in-practice/
The first diversion in this series is an article about caching. It explains why caching is mostly a solution for performance issues rather than scalability issues.
https://steinacker.name/digressions/zipfs-law-en/
