The metadata is data about data, so they are often considered as the element that defines the characteristics of a data. Its relevance in the Data Warehouse projects is given by a practical motivation. In a BI project, 80% of market tools are presented with a metadata layer that makes it possible to translate into a business concept content data from a field in a table.
In the world of DWH and integration processes, the developer or architect DWH has equivalent requirements that force you to find ways to manage this metadata, as usually happens in the integration processes.
The metadata come from many sources and this fact must be taken into account in terms of applicability of metadata in environments of integration processes, which typically include:
• Design phase of the integration project.
• ETL and workflow structure (stages).
• Design of data, data structures schemes provided by third parties or RDB CASE tools.
• ETL operational phase, performance summaries, records, return codes, etc ..
• Exploitation phase: the BI itself.
• Designs BI OLAP (online analytic process).
• Information Analyzer.
• Business Glosary.
• Re-engineering (Rational Rose, Erwin, …).
The lifecycle and metadata management
In the market, there are few tools to manage metadata throughout the cycle, from which are extracted from the beginning until it explodes in a report or dashboard. However, having the ability to manage this metadata from “born” is a crucial factor in the analytical environment.
To have access to traceability, have control over the lifecycle of the data it allows to reach a deeper understanding of it. Only in this way can you can learn things like expiration, timing, source, etc., that allow get more out of information when carrying out the strategic analysis.
Create and have a layer of metadata that results to cover end-to-end, facilitates the understanding of all processes, using business language and minimizing maintenance effort and cost while reducing the risk of error.