We all know that having a single-point-of-failure is not something you want, but dealing with multiple machines doing the same thing at the same time takes a lot of coordination. How do they keep your data in sync while dealing with things like network outages, incorrect data or systems lagging behind in time? It’s not as easy as it sounds but we use it all on a daily basis.
This talk is about RAFT, a popular distributed consensus protocol that allows clusters to keep their data available and consistent. I will talk about how the cluster will elect leaders, how it makes sure everybody is kept up-to-date and how it deals with things like server outages.