What is Apache Drill?
Apache Drill is a multi functional SQL service that allows users to run SQL queries over datasources that traditionally don't support SQL. The most common use case is running Hadoop workloads and queries over CSV/JSON and Parquet output, but other supported datasources include generic JDBC(Data Federation use cases), MongoDB, S3 and more.
Apache Drill both supports direct querying via its REST, Web or CLI clients or use as a JDBC server so that more traditional SQL based tooling can use it as a proxy server to translate and process the queries.
At Anssr we believe that SQL is here to stay, but sometimes the data you process dictates a less traditional data processing environment that still requires some SQL support. Drill facilitates this by allowing you to process large volumes of data in Hadoop or similar and then running SQL queries over the output without having to stream the data to another engine.
Deploying Apache Drill
Deploying Apache Drill using Anssr is very straightforward, to do so you put the Apache Drill and the Zookeeper charms on the canvas create a relation between the two and hit deploy, or run:
juju deploy ~spiculecharms/apache-drill juju deploy zookeeper juju add-relation zookeeper apache-drill
This will spin up Apache Drill in your cloud provider of choice.
Once deployed, then expose the Drill machine either via the GUI or by running
juju expose apache-drill
Login to the Drill server by visiting the exposed IP on port 8047. You should see the Drill user interface.
From there you can then click the query button in the toolbar.
Lets run the test query to get started. Paste
SELECT * FROM cp.`employee.json` LIMIT 20
into the query window and press Submit.
In a few seconds the server will render a table. As you can see from the query, this table of data is generated by running an SQL query over a JSON file. Pretty cool huh?
If you use Apache Drill in production and would like professional support. AnssrIO provides a range of support options. To find out more please visit our support page