End of week 2/20

And thus the second week is over. That is 10% of the master’s thesis done already. It feels like I’m just getting started though, I’ve only tinkered around in Kubernetes, read some research literature and produced a project plan. I haven’t even gotten started on doing any work on anomaly detection within Kubernetes yet.

However, I’ve done some stuff!

Summary of week [2/20]

• Monday
  • Did some tinkering around in Kubernetes and Minikube. Tried to install metrics-server 0.3.0, however this did not work out of the box from the Minikube installment (only version 0.2.1). This proved to be a good learning experience though.
• Tuesday
  • More work on installing metrics-server 0.3.0.
  • Had a headache so I worked from home.
  • Read CPI²: CPU performance isolation for shared compute clusters by Zhang et. al.[1]
• Wednesday
  • Even more work on installing metrics-server 0.3.0, eventually I gave up due to the situation not being a good learning experience anymore.
  • Had a headache so I worked from home.
  • Read Virtualization Techniques Compared: Performance, Resource, and Power Usage Overheads in Clouds by Selome Kostentinos Tesfatsion, Cristian Klein and Johan Tordsson. [2]
• Thursday
  • Elastisys are moving to new offices, so I followed on a tour of the new offices.
  • Wrote project plan
  • Wrote time plan
• Friday
  • Had a meeting with my supervisor regarding the project plan
  • Finalized project plan
  • Sent in project plan
  • Wrote this post

Experiences

I spent two days at home this week and I’m not nearly as productive at home as when I’m at the office. So maybe I’ll go to the offices even if I have a slight headache…

Next week

I have two main tasks for the next week:

• Reproduce earlier experiments performed here at Elastisys within the area of anomaly detection
• Determine a suitable, reproducible infrastructure/environment in which all my anomaly detection experiments will reside in. Preferably a completely scripted setup process to get a container cluster in AWS, run normally for a while and then introduce various kinds of anomalies.

And the week after that I will begin doing “real work”: implementing anomaly detection methods. Exiting!

References

1. [1]X. Zhang, E. Tune, R. Hagmann, R. Jnagal, V. Gokhale, and J. Wilkes, “CPI^2: CPU performance isolation for shared compute clusters,” in SIGOPS European Conference on Computer Systems (EuroSys), Prague, Czech Republic, 2013, pp. 379–391 [Online]. Available at: http://eurosys2013.tudos.org/wp-content/uploads/2013/paper/Zhang_2.pdf
2. [2]S. K. Tesfatsion, C. Klein, and J. Tordsson, “Virtualization Techniques Compared: Performance, Resource, and Power Usage Overheads in Clouds,” in Proceedings of the 2018 ACM/SPEC International Conference on Performance Engineering, New York, NY, USA, 2018, pp. 145–156 [Online]. Available at: http://doi.acm.org/10.1145/3184407.3184414