Computing and Informatics

Availability and performance are key issues in SOA cloud applications. Those applications can be represented as a graph spanning multiple Cloud and on-premises environments, forming a very complex computing system that supports increasing numbers and types of users, business transactions, and usage scenarios. In order to rapidly find, predict, and proactively prevent root causes of issues, such as performance degradations and runtime errors, we developed a monitoring solution which is able to elicit the end-to-end behavior of those applications. We insert lightweight components into SOA frameworks and clients thereby keeping the monitoring impact minimal. Monitoring data collected from call chains is used to assist in issues related to performance, errors and alerts, as well as business and IT transactions.