Monitoring Of Distributed Systems Using Historical Debuggers

Historical debuggers are tools that give insight into the history of programs execution. This functionality is usually achieved by recording logs (execution traces). There are many tools of this kind but only limited number of them are suitable for working with distributed systems. It is a serious problem because the majority of contemporary information systems has distributed nature. This paper proposes how to monitor and trace distributed systems using historical debugger that were designed to work locally. The proposed conception includes an approach to the correlation of execution traces that were recorded on nodes of a system being monitored.