GiAPA is an acronym of the words describing the aim of the software product:
Global iSeries Application Performance Analyzer
The main goal of GiAPA is to enable any programmer and operator to automatically pinpoint performance issues. This is obtained by applying AI (Artificial Intelligence) - all the complex data collection and analysis is hidden in algorithms within the program logic.
In addition, with detailed data for all jobs and tasks available every 15 seconds, the GiAPA data base has become an unprecedented tool for operations being responsible for controlling use of resources and producing usage information and trend reports to management.
GiAPA basically works in 3 steps:
- Collection of performance data, stored in a compressed binary format.
- Expansion and analysis of the collected data.
- Reporting – Interactive use of displayed or printed reports and graphics
Runs in batch typically using less than 0.1 % of one CPU allowing data collection 24/7. The data is stored in a compressed binary format using minimum disk space.
Complete attributes and use of 22 types of resources are received for all active jobs from operating system APIs every 15 seconds. This data is processed real-time.
If a job within a 15 seconds collection interval is using excessive resources indicating that it might have or cause a performance problem, GiAPA also retrieves data for all open files and the program call stack.
CPU efficient !
Uses < 0,1 % CPU
Collects data for all jobs every 15 sec's
Fully automatic - no "experts" needed.
Output detail level determined by installation parameters
Runs in a batch job creating a number of data base file members containing the results to be used as input for the numerous exception reports and charts. Data for a day, an hour, or a whole week, etc. may be selected. Can be scheduled to run e.g. every night, or requested interactively if e.g. a recent incident should be analyzed immediately.
Data from jobs suffering from or potentially causing performance problems are stored individually by GiAPA for further analysis. Data for other jobs or tasks are kept summarized by type and interval, unless all details explicitly are requested.
All data analysis is performed in a manner transparent to the user. Installation parameters control the limits for when detailed data should be kept.
After expansion and analysis a large number of displayed or printed reports and graphics can be requested interactively. The output can be summarized in many levels, or shown very detailed down to job-id per 15 seconds, combined with extensive selection and sort options.
Details for jobs that used any significant amount of resources are automatically stored and detailed performance analysis requires only few keystrokes rather than expensive external assistance. The results are available as traditional reports including all needed details, but can within seconds be converted to charts.
Apart from the above basic functions of GiAPA there are quite a number of additional features including additional performance investigations, analysis of data base files, data import/export and housekeeping.