M.Sc. Lirim Osmani defends his doctoral thesis "Multicloud management plane: Experiments with Scientific and Telco Clouds" on Friday the 11th of August 2023 at 12 o'clock in the University of Helsinki Chemicum building, Auditorium A129 (A.I. Virtasen aukio 1, 1st floor). His opponent is Professor Pasi Tyrväinen (University of Jyväskylä, Finland) and custos Professor Sasu Tarkoma (University of Helsinki). The defence will be held in English. It is possible to follow the defence as a live stream at https://video.helsinki.fi/unitube/live-stream.html?room=l42.
The thesis of Lirim Osmani is a part of research done in the Department of Computer Science and in the Content-Centric Structures and Networking group at the University of Helsinki. His supervisors have been Professor Sasu Tarkoma and Laboratory Engineer Tomas Lindén (University of Helsinki).
Multicloud management plane: Experiments with Scientific and Telco Clouds
Virtualization is the foundation of what we experience today as Cloud Computing. Delivering software and services initially through virtual machines and later through containers, has the industry and research efforts centred around OpenStack and Kubernetes. Both projects are community driven and industry supported, thus providing flexible frameworks for virtualizing large scale production infrastructures, yet primarily within a single administrative domain.
In this thesis, we provide architecture solutions that extend further OpenStack and Kubernetes for managing heterogeneous resources across multiple cloud environments. More specifically, with OpenStack we use GlusterFS for a storage solution that brings stability and scalability for a private deployed cloud, and Host Identity Protocol as a secure and scalable network technique for scaling workloads across many environments. With Kubernetes we use KubeFed and Network Service Mesh in providing federation and cross-cluster connectivity for container workloads.
Nonetheless, adding flexibility also brings new challenging problems in terms of performance and stability. We perform experimental evaluations using workloads and scenarios particular to our representative use-cases from science and industry, namely CERN and Telecom providers. The results indicate the techniques employed have multicloud capabilities for managing and deploying workloads at large scale respective to our chosen use-cases.
Availability of the dissertation
An electronic version of the doctoral dissertation will be available on the e-thesis site of the University of Helsinki at http://urn.fi/URN:ISBN:978-951-51-9350-6.
Printed copies will be available on request from Lirim Osmani: lirim.osmani@helsinki.fi.