Apache Druid : A Hands-on Course

Install Apache Druid on a local machine via the single-server quickstart, meet Linux or macOS requirements and Java 8 update 92 or later, then start micro-quickstart.

Explore the Druid console through its tiles, data sources, segments, and services, learn how to load data, manage ingestion tasks, and run queries with real-time insights.

Load data from a local disk file into a Druid data source via the console, create and submit an ingestion spec, then query with sql syntax and apply countryName filter.

Explore multiple methods to load local disk files into druid, including the console load data, the ingestion tab json tasks, the bin post-index-task script, and curl calls.

Transform data during load in Apache Druid by adding transformed columns, such as uppercasing country names and computing comment lengths, using transform expressions like upper and strlen.

Apply a pre-row filter at load time with a like condition on countryName to pass only Australia; the sample confirms Germany yields no records.

Load a nested json file, flatten the address fields into city, state, and pincode, then set parse time, disable roll up, set day granularity, and query to confirm five rows.

Learn how to load data from kafka into druid, including installing kafka, creating a wikipedia topic, producing json records, and configuring druid load spec for kafka stream ingestion.

Load csv data from https into Apache Druid by connecting data, parsing headers, and publishing a real-estate data source.

Learn how Druid SQL translates into native queries, explore JSON native query formats, and use intervals, data sources, and result formats to query data efficiently.

Explore Druid architecture and how coordinators, overlords, brokers, middle managers, and historical processes coordinate ingestion, segment management, and query serving with ZooKeeper, metadata storage, and deep storage.

Learn to integrate Apache Druid with Superset, configure security and host settings, start Superset via Docker, connect to Druid, and create datasets and time series visualizations.

See how Netflix uses Apache Druid for real-time insights at scale, ingesting two million events per second and delivering subsecond to a few-second queries with roll-up and compaction.