Live Visualization of Database Behavior for Large Software Landscapes: The RACCOON Approach
Databases are essential components within large software landscapes, since they are employed in almost every information system.
Based on the growing complexity of software systems and a steadily increasing amount of data which is collected, processed, and stored in databases, it is difficult to obtain a live overview of these software landscapes. This often leads to an insufficient knowledge of the actual internal structure and behavior of employed databases. Furthermore, databases are often involved in performance issues within information systems. A solution to these problems is employing live visualizations of databases and related communication from applications within the software landscape.
These visualizations allow operators to understand their databases in detail and to analyze database queries performed by applications. Based on established visualization concepts like the entity relationship diagrams and the 3D city metaphor, operators can be supported in the task of database comprehension. Established monitoring techniques, like dynamic and static analysis, can be used to capture necessary information from applications and databases.
In this paper, we present our live visualization approach of databases and associated communication for large software landscapes. Our visualization offers two different views – a landscape-level and a database-level perspective.
The landscape-level perspective provides an overview of monitored applications and related databases. The database-level perspective reveals database schemas within a database, shows contained tables and relationships, and allows for the inspection of executed queries based on the monitoring information collected at runtime.