direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Inhalt des Dokuments

MA: Monitoring Integrations for AIOps platforms

 

Monitoring servers, kubernetes clusters, VMs, switches, etc. is highly important to determine problematic behaviors of system infrastructure. As the industry shifts increasingly more applications to the cloud and experts to administrate such systems are rare, AIOps tries to provide solutions to help the DevOps to detect anomalies automatically. The main problem remains, that the best anomaly detection algorithms just can detect anomalies when they are feed with suitable input data. Consequently, the selection of valuable monitoring data is important.

The main focus of the thesis are:

  • Selection of open source monitoring tools, which are used widely in industry
  • Connector to an existing AIOps platform (bitflow format https://github.com/bitflow-stream/)
  • Selection of a service to be monitored and creation of service specific anomalies
  • Creation of actions to solve the service anomaly
  • Evaluating the impact of the different metrics for existing anomaly detection algorithms

Prerequisites for working on this topic are provable and advanced knowledge in algorithmic design, docker, kubernetes and very good programming skills in either Java, Python or Go.

 

 

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe

Auxiliary Functions

Contact

Florian Schmidt
+49 30 314 28306
Room TEL 1205