You can install and run Drill on any number of nodes in your Hadoop cluster. Install the
mapr-drill package on each node that you want to run Drill. The
mapr-drill package installs the Drillbit daemon and the Drill shell.
The Drillbit daemon is the core Drillbit service that runs on a node. Drill’s processing capacity increases with the number of Drillbit services running in a cluster. Each node running the Drillbit service can receive, plan, and execute queries sent from a client. Warden manages the Drillbit service, which simplifies the installation process and management of Drill. For more information about Warden, refer to Warden in the MapR Architecture Guide.
The Drill shell is the command line interface, a pure-Java console-based utility, for connecting to relational databases and executing SQL commands.
When you install Drill from the MapR RPM, Ubuntu, or SUSE packages (or any builds that include the MapR profile), the installation includes the 4.1.0 MapR client. The client is designed in such a way that Drill can take advantage of Hadoop 2 in the file system. The MapR client is forward- and backward-compatible and can connect to 4.x and 5.0.0 MapR clusters.
To install the latest version of Drill on MapR, see Installing Drill on MapR.
Component and System Compatibility Matrix
See the Interoperability Matrix pages for information about the compatibility of Drill with operating systems and ecosystem projects.