Druid - Queries and Approximate Queries

Druid is an advanced time series database designed to enable interactive queries against arbitrary dimensions of very large scale datasets. Fast queries are possible because Druid indexes every column in a database with a highly optimized inverted index. In this session of Coding with AppsFlyer we will investigate the querying options available in druid. We will cover all the native druid query types -- TimeSeries, TopN, GroupBy, Search and Select -- by running queries ourselves against a live Druid database and learning the intended use cases for each of these query types. We will also explore the SQL translation options available for querying druid directly from SQL. Finally we will experiment with approximate queries available in Druid when the data has been indexed using either the hyperUnique or thetaSketch aggregators. Approximate queries can generate fast results for count distinct queries for example when counting daily or monthly active users.

As always, please bring your own device and come ready to try it yourself.