Automating the Process of Identifying the Causes of the Events in the IT Infrastructure

This paper addresses cause-effect relations among events revealing problem. Nowadays IT-infrastructures tends to become more and more complex due to increasing complexity of informational systems. On the other hand, IT support engineer who knows all systems’ features and grounding principles tends to be unattainable ideal. When incident occurs, IT-support engineers are responsible for resolving the issue as fast as possible. However, in many cases resolution time is unpredictable because incident’s root cause should be identified first, and that is a real problem. Main objective of the concerned Event Information Analysis System (EIAS) is to reveal hidden cause-effect relations and provide some insights for search directions leading to increased efficiency of the IT support staff. Concerned EIAS is based on open and free software, specifically Java, Rsyslog, and Weka, and uses machine-learning techniques. Keywords - system logs; system log analysis; cause-effect relations; big data; machine learning.