Metasafe in a Nutshell
MetasafeInANutshell.pdf (201.0 kB)
There are more and more applications which manage complex data - i.e. data with lots of entity-types and relationships. To squeeze such a real life problem into a set of tables of an RDMBS is costly and complex. You pay for that in many ways:
- The conceptual model is never really tested
- The original conceptual model gets lost
- The access to the database requires cryptic SQL-statements
Nobody will deny the benefits of the relational model - but it is not designed to store complex models and complex data in a graceful manner.
One type of database does not fit all!
- planning and reporting on management level is based on Excel
- applications dealing with complex data develop a thick layer on top of the RDBMS
- even complex development artefacts are stored in files and libraries
- lack of documentations (models, blueprints) leads to expensive and risky maintenance
Conclusion:
Complex data require a specialized DMBS. Metasafe is specialized to handle complex data.
The Metasafe Solution - an Executable Conceptual Model
Peter Chen explained once that ER-Modeling should not be restricted to modeling but should be extended to host instance data to bridge the well known "impedance mismatch".
Metasafe provides a seamless integration between conceptual models and instance-data in one integrated system. Changes and extensions of the data models are simple update transactions - without the necessity to unload and reload the database. The new structures on the instance level are ready immedately after committing the model update.
Conceptual Information Models and Submodels are managed in the database including the documentation of the models. They can be tested with instance data. They are protected by proper multi-user access control.
The Tool Box - to make Life easy
metaModeler - an Eclipse based multi-User multi-Model model Editor and Documentor: Conceptual Data Models and Submodels are defined including the user access rights.
metaEditor - an Eclipse based multi-user configurable data browser provides fast and easy acces to the instance data with automatic access rights control.
metaExcel-Interfaces - to load and export content of the repository to Excel
erSQL - query language for users and application developers based on the conceptual model
BIRT-Business Intelligence and Reporting - use erSQL-queries to retrieve data from the repository and create reports and BI-results using the cube-structure of BIRT.
Java-API - a powerful framework to write model driven applications and access the data models and the instance data stored in the metasafe repository
SIZE is important: Metasafe handles GB of data - just enough to support also large volumes of metadata or also sizable applications.
Instance Data – Entities, Relationships and Attributes – are stored in the same database as the models. Entities may exist in multiple manifestation of "Variants" (or Editions) and "Revisions" (or Versions) and can be stored in separate name spaces (Catalogs). This ensures that the data entered into the database are always consistent with the model.
Data driven Applications are flexible and easy to extend to new functionality. They are entirely controlled and documented by the model and use the metadata stored in the model. Applications work directly with the conceptual model – hence the impedance mismatch does not exist.
Benefits from Metasafe based Solutions
Metasafe is the obvious choice for applications with GBs of data, hundreds of types (entities, relationships, attributes), complex relationships, high demand for flexibility and ease of access.
Metadata is the ideal platform for metadata management, configuration management, CMDBs (configuration management databases), replacement for large Excel-populations, planning and analysis-tools, data infrastructure for modern development tools etc.