Branch.io has been leveraging Apache Druid for real-time analytics for over eight years, handling more than 600,000 queries daily. In this presentation, we will explore the diverse use cases that Druid addresses for Branch.io, detailing our data ingestion methods through both real-time and batch processes. We will discuss various optimizations implemented to enhance query performance and reduce costs, including techniques such as cardinality capping, data rollup to higher granularity, and multidimensional partitioning. Additionally, we’ll cover our packaging and deployment strategies, and conclude with insights into our approach to query debugging.
Speakers:
– Ramesh Shanmugam, Principal Data Engineer, Branch.io
– Andrei Harbunou, Senior Software Engineer, Branch.io
[Timestamp] Table of Contents:
[0:00] Real-Time Analytics at Branch.io
[12:15] How we run Druid: Cluster Size, Packaging, Real-time Query Logging, Deployment
[17:50] Optimizations: Cardinality Protection, Daily Rollup, Data Partitioning
Keynote: Powering Event-Driven Data with Apache Druid
The distinction between OLTP and OLAP is becoming less relevant as data architectures shift toward entities and events. In this session, we’ll delve into how Apache Druid’s event-first approach synthesizes...
Watch nowClosing Keynote: Charting the Future of Druid
What lies ahead for Apache Druid? Join us as we explore the evolving landscape of Druid’s query and storage engines, and how they are positioned to address the biggest challenges in event data for the future. Speaker: Gian...
Watch nowSalesforce: Tracing Service Dependencies at Scale with Druid and Flink
At Salesforce, we manage approximately 300 million distributed spans to infer service dependencies. We have successfully utilized a combination of Druid and Flink to handle this scale with high availability....
Watch now