The IB-groep is a governmental organisation for student administration. The IB-Groep E-service offers direct web access to student information via the internet. The end users on the website complained about problems getting to their data. Mr Ritter was asked to find the root cause of these problems.
The website directly connects via an MQSeries based middleware infrastructure to their back office systems running on IBM mainframe, AS400 and Windows 2000. IB-Groep experienced a wide range of errors, from lost messages to timeout situations and lost connections. Mr Ritter was asked to determine the cause of all these problems, find solutions and to give advice for improving the infrastructure.
Mr Ritter first setup a measurement system to collect information on current system behaviour. This included network- , computer system traces and response time overviews. This information was presented in spreadsheet format to management on a daily basis showing availability and periods of system outage. This proved to be a very fruitfull approach. Mr Ritter concluded from all collected data that there was a serious network problem, some internal systems had performance problems and the Service Level Agreement with its external mainframe supplier was not met.
Meetings were arranged with the external mainframe supplier to improve performance to aggreed SLA levels. Also network and system specialists where involved to find the bottlenecks and improve system performance. Mr Ritter coordinated work between these specialists and conducted further tests. All problems were identified and solved. Mr Ritter compiled a final report on the current state of the infrastructure and improvements to be made.