Performance slowing down: is a monitoring tool the right solution? Are you capable to resolve the problems with APM solutions or any other monitoring tool.
We have 2 different monitoring solutions. One is based on infrastructure. The other, latest ones, are based on Application Monitoring.
Both are capable of pinpointing the area responsible for a delay. The first problem however, is how to interpreted the results of the monitors.
Let me give you an example: you have a database running on one server and the monitoring shows you 99% of memory use. The system has 32GB of Memory. Most of the people but also most of the monitors will tell you add more memory to the server.
Now you add another 32GB to the server and surprise; memory stays on 99%. Do I need to expand now to 128GB or do I have another problem?
It is all about parameter management and relationships between different applications, hardware and parameter settings. If you leave your database in this example on automatic memory it will take all the memory it can get and your system start paging like crazy.
The problem is the combination of relationships are almost endless. So you need an intelligent analytic system to determine if a parameter is the reason for the delay.
ITPA has built an analytic system on top of IBM ITM, IBM APM and Dynatrace this system has already solved several performance problems in many different applications, like JDEdwards, Peoplesoft, Maximo and a few home written applications.
28 januari 2017