Plume Design Achieves 27 Billion Ops Per Day with Distributed SQL
summary
Plume Design builds an Intelligent Wi‑Fi platform that supports millions of smart homes and achieves 27 billion ops per day using a 60-node YugabyteDB cluster holding 35 TB of data. The stack uses range sharding, geo‑replication, batched writes, connection pooling and careful compaction and tombstone tuning to control tail latency. Engineering retros describe how schema changes, hotspots and backups are automated, how load balancing and instrumentation reduced failure modes, and how those operational practices make the scale sustained and repeatable.
AI-written summary · about
related talks
30mFrom monolith to petabyte scale: Built a cloud-native, AI-ready data layer
29mEnterprise Scale: How a Distributed Architecture Can Deliver Scale and Resilience
26mA Winning Multi Cloud Data Strategy
32mDay 2 Operations Best Practices
46mBig databases, many servers, data center or cloud - it's all here!
57mNo Downtime, No Limits
25mHow Sharekhan Scaled Trading Platform with Yugabyte's Distributed SQL
51mEmbrace, Extend, Evolve - Making PostgreSQL Work for a Cloud Native World