Percona Live Online

There is a growing demand for real-time data warehouses by data-driven companies to implement real-time Online Analytical Processing analytics, real-time data panels, and real-time application monitoring. However, the architecture of real-time data warehouses has long been thought complex and difficult to operate and maintain.

As an open source and distributed Hybrid Transactional/Analytical Processing (HTAP) database, TiDB can be used as a backbone storage for real-time data warehouse in multiple use: business dDataSsource, dimension table DdataSsource and the analytical database for summarized data. The combination of stream processing systems (e.g. Apache Flink) and TiDB could become an efficient, easy-to-use, real-time data warehouse that features horizontal scalability and high availability.

In this talk, Qi Zhi will deep dive into what a real-time data warehouse is, how TiDB powers real world real-time data warehouses and the patterns on combining streaming processing systems and TiDB.