After the Crash: Designing Forgetful Systems

Any message passing through your application can be forgotten. The process crashes. The integration broker is down. The database is corrupted. The hacker achieves their goal. An entire data center is flooded, leaving the servers underwater. How do we prepare for this inevitability? Micro-batching, queuing theory, the actor model, and a continuously available, self-healing cluster can help. Every system can forget. Use the BEAM to help it remember.

Flora Petterson

Flora is a developer on the Waterpark team at HCA. She worked as a puppeteer, bartender, and writer before teaching herself Elixir to build hobby projects with Nerves. She lives in her hometown of New Orleans with her husband, three kids, and one loud cat.