A critique of the CAP Theorem
Ever since I started working on distributed systems, I’ve found the “CAP Theorem” very confusing mostly because it’s over simplified and not well defined. It’s more of a rough guideline than a theorem.
Sacrificing P is not really an option. CP and AP are hard to map to practical systems; supposedly AP systems have interesting consistency properties. I find that the whole thing is actually too restrictive and doesn’t cover the design space.
So I searched online to see what others think and was happy to find “A Critique of the CAP Theorem”. Glad someone (Martin Kleppmann) documented this. Recommended read if you want to understand the confusion around the “CAP Theorem”.
Update (07/05/2020): I stumbled upon some more interested reads on this topic: