A metadata catalog is a set of relational tables. They contain information about data and how to convert data from nonrelational to relational formats. The Boomi server accesses the information that is stored in these catalogs to perform later operations.
Boomi maintains data sets that are the metadata representation of the data source. As data sources are added to Boomi using a connector to an external database, file or application discovery is performed. During discovery, Boomi categorizes and examines the data in place, and then copies a sample of the data to itself. Data remains on the source. Only sample data are loaded temporarily to perform analysis on the data, and metadata are stored for Boomi functionality.
The metadata catalog has two components:
- Technical metadata, which is gleaned from the source during discovery
- Business metadata, which users add collaboratively based on what the data means to the business
The technical data consists of:
- Table names
- Column names
- Data types
- Key constraints
- DBA comments to the columns
If the source is a file system (XML, JSON, Parquet files, and so on), Boomi creates a schema for the file. It then
infers the data type information for the metadata.
Boomi can also identify sensitive or PII information from the source, which is classified and tagged automatically. The internal security settings hide content unless an individual has the required privileges to see it.
All this information is searchable using the Boomi web interface.