Step 2: Prepare Packages and Repositories

To install services correctly, each node must have access to the package files.

The Data Fabricsoftware distribution is separated into two repositories that contain the package files:

  • Data Fabric packages. These provide core functionality for Data Fabric clusters, such as the file system.
  • Ecosystem packages. These packages are not specific to HPE Ezmeral Data Fabric. Examples include the packages for Hive and Spark.

You can make packages available to each node, as described in subsequent sections, using the Data Fabric Internet repository, a local repository, or a local path with rpm or deb package files. For information about packages and repositories for Data Fabric software and Hadoop Ecosystem tools, see Data Fabric Repositories and Packages.