Building an Environment for Processing Logs and Performance Data of Containerized Applications

OData support
Dr. Maliosz Markosz
Department of Telecommunications and Media Informatics

Living our life in a modern technological environment, IT systems and devices have become the part of our life. Because of the development of science, people are having more and more expectation on technological services. Nowadays continuous operation is indispensable. Time boxes for debugging and for maintaining are less and less acceptable. In highly available systems the faults have to be identified and fixed as soon as possible. Moreover, systems have to be always under supervision to minimize downtime.

In my thesis, I am building a system which is responsible for the monitoring of applications running in a modern environment. Because of the spread of cloud computing and virtualization, it is very important to build all the components of our system compatible with virtual or containerized environment.

In my thesis, the provided monitoring system is capable to monitoring logs and metrics of containers running for testing purposes. The monitoring system is based on the well known Elasticsearch and the less known Prometheus. The data collected by Elasticsearch is visualized by Kibana. Prometheus has it`s own graphical interface, therefore no other application is needed. The cluster contains four physical servers and having Ubuntu distros on them. The applications under monitoring are running in Docker containers which were managed either by "hand" or by the popular Kubernetes container manager application.Applications take place in containers which are running in pods. One pod could contain one or more containers. My provided system is monitoring a complex virtual environment and the physical hosts of the system. The collected data is processed and visualized in different ways. As a conclusion my thesis provides a short recommendation on the tested monitoring tools.


Please sign in to download the files of this thesis.