How do you build and operate systems that can ingest millions of events per second, store petabytes of historical data, and run thousands of queries per second, all at subsecond response times?

It’s not easy: you need to deal with workloads that mix datasets of different sizes and change rates the need for quality of service managing fluctuating performance across fleets of servers (there is always at least one slow server in every cluster) making every query highly efficient with minimal CPU time keeping costs under control

But it can be done! Gian Merlino, Apache Druid® committer and co-founder of Imply will share tales of scale, both of database design and real-world deployment patterns removing the mystery of how to build and deploy high-performance systems.

Audience: developers, architects, and leaders with at least a basic understanding of analytics  technologies, with an interest in analytics databases at scale

Takeaways:

How to build and operate systems that can ingest millions of events, store petabytes of historical data, and run thousands of queries

What’s the right mix of compute-storage design, scatter/gather query engines, and cluster management

How Apache Druid delivers high-performance for interactive data conversations with high concurrency and low latency, for both streaming and batch data