

Atlas Proxy

In order to make the Atlas-Amundsen integration smooth, we’ve released a python package, amundsenatlastypes that has all the required entity definitions along with helper functions needed to make Atlas compatible with Amundsen.

Usage and Installation of amundsenatlastypes can be found here

Metadata Entities Explained

In order to support different kind of statistics for Table, Column (or even DataSet) we need to add new entities Metadata (table_metadata, column_metadata etc.) and make a relationship between Metadata and DataSet.

Also, in order to store User information and User’s statistics and actions, we’ll need a new User type entity, which will be linked with Metadata as Readers, Owners, Managers etc.

We are not going to add custom attributes directly to a Table/DataSet entity, to make it backward compatible and clean of course.

Configurations

Once you are done with setting up required entity definitions using amundsenatlastypes, you are all set to use Atlas with Amundsen.

Other things to configure:

• Popular Tables