The 7 x 24 hour cloud database operation of a relational database management system such as Db2 for z/OS at DATEV generates a large amount of important data. These include, for example, the number of calls to individual applications, individual SQL statements and many other detailed information collected at system level – the so-called “detector data”. These mass data are currently only cumulated / aggregated to a small extent and can therefore only be evaluated over a limited period of time (10 or 30 days). To make them easier to evaluate regarding their extremely valuable information:
- like trends in applications
- progressively evolving vulnerabilities
- Typical patterns in application operation etc.
meaningful aggregations are required. These would then enable the following:
- make the mass data quickly and efficiently evaluable
- to significantly extend the period of current data storage and the possibilities of retrospective analysis – objective “annual / multi-year evaluations
The aim of the work is an appropriate aggregation of the available data via SQL on a database to be implemented. Based on this, first evaluation examples are to be created, with which resulting weak points and the possibilities of trend analyses can be pointed out.