University of Helsinki homepage Suomeksi På svenska In English
University of Helsinki Department of Computer Science

Department of Computer Science

Mikko Pervilä

Photograph of Mikko

  • Postdoctoral Researcher
  • Mobile: +358 44 7696086

I pursue a field of research that can be described as empirical system reliability. More specifically, I'm trying to figure out the effects of diversity in peer-to-peer systems, specially in relation to the fault tolerance of both P2P and warehouse-scale distributed systems.

In layman's terms, my research questions whether a system built of diverse components provides superior reliability when compared with a system built from (almost) identical components.

I've also been interested in saving the planet by reducing the power usage of I(C)T equipment on a large scale. This used to be a very trendy subject, and may be so once again. My research is definitely hands-on, not theoretical. Much of the data we collect is publicly available.

You can follow a local mad scientist by reading his blog about Helsinki chambers and experimental cooling setups in general. The blog provides some details on the computer-powered greenhouse on top of the Exactum building. The project's official web pages contain much more detail. I also maintain a wiki page which lists other projects.

Awards & Grants

Selected works

Data Center Energy Retrofits, Ph.D. thesis, December 2013.

Underfloor Air Containment, IEEE Online Conference on Green Communications, October 29-31th, 2013.

Harvesting Heat in an Urban Greenhouse (extended abstract), ACM Greenmetrics 2013 workshop in conjunction with SIGMETRICS 2013, Carnegie Mellon University, Pittsburgh, PA, USA, June 17th, 2013. (slides)

Harvesting Heat in an Urban Greenhouse, ACM CoNEXT 2012 workshop on Urban Networking, Nice, France, December 10th, 2012. (slides)

Implementation and Evaluation of a Wired Data Center Sensor Network, (read the web version) 1st International Workshop on Energy-Efficient Data Centres (E2DC), Madrid, Spain, May 8th, 2012, in connection with e-Energy 2012 (slides)

Cold Air Containment, Proceedings of the 2nd ACM SIGCOMM workshop on Green networking, August 19th, Toronto. (slides)

Running Servers around Zero Degrees, First ACM SIGCOMM Workshop on Green Networking, August 30th New Delhi, India. Republished in ACM's CCR Volume 41, Number 1, January 2011 (slides)

Performance of Ajax applications on mobile devices, Master's Thesis. It was awarded with the SMFL pro gradu price.

Using Nagios to monitor faults in a self-healing environment, graciously hosted by the Nagios Community. Written for the Seminar on Self-Healing Systems, spring 2007. The most recent version is also available from this server as paper8.pdf.


I've lectured or assisted in the following courses at the Department of Computer Science:


The Nmap3Nagios tool may be helpful to some.

The how-to for the Fall 2008 short course on MPI is online at MPI-s2008.

My conversion of the department's "Guide on structure and layout on thesis" for is available from the following folder. The first version uses Times New Roman and the second Century Schoolbook L. Use the former.

  • Help! prints blank pages after the first / summary / index page.
  • Tools / Options / Print: uncheck "Print automatically inserted blank pages"
    Alternatively: write your favorite poem on the blank page and send it to me at the address specified above. Internal mail works great.
  • Tomi J. Mikkonen remarked that using NeoOffice on MacOS, the table containing the abstract overflows to the following page. You can either manually drag the "Tiivistelmä - Referat - Abstract" cell a bit smaller, or change the table row size to 6,56 cm.

This Java-based password generator was a project work done in 2004. It can be used to generate pseudo-random passwords from text input, e.g., favorite poems, ascii pictures or any other text. Using the same text yields the exact same password listings.


"All that is necessary for the triumph of evil is that good men do nothing" (or words to that effect) -- A study of web quotation by Martin Porter, dated January 2002. I found the study immensely enjoyable, with the follow-up most interesting as well. Not to mention the weight of Edmund Burke's actual words.