Data Modeling Metadata Package

Metadata
Metadata is information generated during the development of IT solutions that defines both business and technical understanding of the data. There are many different types of metadata, however most metadata originates during the data modeling process. The data modeling process includes identifying the business requirements and then deriving data requirements. The data requirements are then captured in an E-R diagram (also known as a logical model) for analysis and normalization. The E-R model is then converted into a physical model that represents the database characteristics.

The data modeling process generates a wealth of metadata. The business requirements, data requirements, E-R diagrams, and the physical models all consist of metadata. This metadata continues to evolve during the development process to include process metadata, application metadata, reporting metadata, and sourcing metadata. The evolution of the metadata requires that the metadata created during the modeling process be accurate and of the highest quality. Absent or deficient metadata from the modeling process hinders the evolution of metadata during later phases of the development process. This is likely to lead to deficient applications.

Obviously, metadata is important. In today's world, most data modelers utilize tools to perform the data modeling process. The industry leading modeling tool is the CA ERwin® Data Modeler (ERwin). This software allows the data modeler to create and capture modeling metadata. Development projects depend on good metadata to be produced and used to create the relational structures to store the data.

Jump-Start Package
The Data Innovations' metadata jump-start package establishes the environment to create the highest quality metadata during the modeling process. Metadata standards are defined for data modeling efforts. A validation process is created to verify the metadata standards are being met. The package also provides a metadata repository and processes to extract ERwin metadata and populate the metadata repository. This metadata repository makes the model metadata accessible for sourcing or reporting needs.

The following sections identify what is included in the package:

Requirements:
The metadata package will require the following:

  • An existing data modeling environment that utilizes CA ERwin® Data Modeler and CA ERwin® Model Manager (Model Manager).
  • Purchase of the metadata repository and metadata extraction utilities from Data Innovations.
  • A minimum of 160 hours of consulting services.

These requirements allow Data Innovations to provide discounts for both software and professional services.

Establish the Environment:
The metadata repository and extraction utilities will need to be installed. The following identifies the task to establish the environment:

  • Install metadata repository:
    • Acquire access to the Model Manager software and the RDBMS.
    • Save the metadata repository physical model to Model Manager.
    • Create the metadata repository in the target RDBMS using the physical model.
    • Integrate the metadata extraction utilities into the modeling environment.
    • Extract the physical model’s metadata into the metadata repository using the extraction utilities.
    • Validate that all the metadata was extracted and loaded into the metadata repository properly.
  • Establish metadata standards:
    • Review the current metadata standards; naming standards, database standards, etc.
      • Create a document identifying the new metadata requirements for data modeling.
      • Create a document identifying how the metadata requirements will be validated.
    • Review the current model metadata:
      • Create a document identifying the metadata currently captured in a logical data model.
      • Create a document identifying the metadata currently captured in a physical data model.
      • Create a document identifying metadata missing from logical and physical models.
      • Review current metadata with client and determine appropriate logical and physical metadata requirements going forward.
      • Create a document that identifies the new metadata requirements. Include example metadata for the new requirements.
    • Create a process for validating model metadata:
      • Review the model development process and identify the points where logical and physical models will be review for metadata content.
      • Create SQL to validate model metadata within the metadata repository.
      • Create reports against the metadata repository to identify the existing and missing metadata.
      • Identify roles and responsibilities for reviewing the metadata content of the models.
      • Create a document identifying the procedure for validating metadata. Include a step-by-step procedure for migrating metadata into the repository for evaluation.

    Deliverables:
    The following bullets identify the primary deliverables of this package:

    • Metadata standards. A document is created that identifies the metadata requirements for logical and physical data models. This document is used by the data modelers and the database administrators to create logical and physical models that meet the metadata requirements.
    • Metadata reports. A metadata report is created against the metadata repository to identify existing and missing metadata. These reports are used to validate the metadata content for logical and physical models.
    • Metadata repository. The Data Innovations’ metadata repository and the extract utilities are installed and tested.
    • Metadata validation. A document is created that identifies the procedure for validating metadata content for logical and physical data models. The document identifies a step-by-step process for loading the metadata into the metadata repository and creating the metadata validation reports.