Auto Repair in Apache Cassandra by Andy Tolbert, Jaydeepkumar Chovatia скачать в хорошем качестве

Auto Repair in Apache Cassandra by Andy Tolbert, Jaydeepkumar Chovatia

Are you tired of the operational headaches that come with manually repairing your Apache Cassandra cluster? Dive into the future of Cassandra maintenance with this deep-dive session from Community Over Code 2025! 🚀 Join Jaydeepkumar Chovatia from Uber and Andy Tolbert from Apple as they unveil the new automated repair scheduler, a groundbreaking feature merged into Apache Cassandra. Say goodbye to the complex ecosystem of external tools and control planes! This talk reveals how Cassandra can now manage its own data consistency, intelligently and efficiently, right out of the box. The speakers, who were instrumental in developing and battle-testing this feature at a massive scale, unpack everything you need to know. They cover the historic pain points of repair, the elegant architecture of the new scheduler, and its powerful capabilities like smart token splitting, which minimizes network load. A significant portion of the talk is dedicated to a masterclass on incremental repair. Andy Tolbert demystifies its complexities, including the dreaded "anti-compaction," and presents clear, actionable strategies for migrating your clusters safely and effectively. Whether you're an SRE, a DBA, or a developer running Cassandra in production, this session is packed with invaluable insights to make your life easier and your clusters more reliable. In this talk, you'll learn: The "Why": A quick refresher on why repair is absolutely critical for data consistency and handling tombstones in Cassandra. The Solution (CEP-37): How the new, built-in scheduler works just like compaction—no external dependencies needed! Smart & Safe: Discover key features like dynamic configuration, built-in safety nets (like disabling repair on low disk space), and extensibility for new repair types. Incremental Repair Demystified: Understand the pros and cons, how to avoid common pitfalls like anti-compaction, and the best migration paths for your workload. Observability is Key: See the new metrics and tools that give you a crystal-clear view of repair status across your entire cluster. The Future Roadmap: Get a sneak peek into what's next for automated data management in Cassandra. Recorded at Community Over Code 2025 in Minneapolis. #ApacheCassandra #Cassandra #Database #DistributedSystems #SRE #Uber #Apple #BigData #DatabaseRepair #CommunityOverCode