Weekly Notes: A couple of good talks from reInvent about graviton and how riot gaming thinks about managing their backend servers
- AWS re:Invent 2024 - Optimizing performance and cost with AWS Graviton (DEV302): Liz Fong-Jones talks about Honeycomb’s use of arm64 and how they got there
- Mercari’s production readiness checklist for new services: List of stuff to think about while building a new service. Also described are service levels with minimum preparedness criteria. I guess the sre / platform team won’t help keep a service up if these designed in good production practices aren’t they. (The minimum bar is pretty low and a team can choose what their practice will look like.)
- AWS re:Invent 2024 - Effortless game launches: How League of Legends runs at scale on AWS (GAM307): Riot games talks about game provisioning platform and their migration from datacenters running backend game servers to aws
- Tame Your Telemetry: Introducing the Honeycomb Telemetry Pipeline: Configurable sampling behaviour in Refinery as well as rehydration of logs not added to the analysis engine that weren’t interesting on the way in but you’d like to see as part of an investigation now. (Logs are always stored) Neat!