Overview
LinkedIn intends to run TiDB as a managed service offering at a very large scale. But before they can do that, they need to learn about the system’s scaling, resiliency, operability, and extensibility characteristics. To that end, they have been experimenting with TiDB in multiple ways, such as micro-benchmarking, simulated production-like workloads, and stress testing. This session recording from HTAP Summit 2024 unpacks some of LinkedIn’s learnings from their experiments and charts their path forward.