No description, website, or topics provided.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
README.md
_config.yml

README.md

Software Developer Targeted Performance and Cost Management for Autonomic Cloud Applications (ACOSTT)

This is a project announcement for a fully sponsored PhD under the advisory of Dr. Philipp Leitner. The position is part of Chalmer's ICT Area of Advance.

Please apply using this link. The hard deadline is November 30th, 2017. However, applications will be evaluated as they come in.

Project Summary

Scale-out cloud services, such as Netflix or Microsoft Bing, often consist of hundreds of independent microservices running on thousands of virtual machines or containers. Deployment automation and self-management (e.g., through circuit breakers or autoscaling) are fundamental to the management of such systems. However, these autonomic principles have made testing and operations cost estimation at development time challenging. In this PhD project, the goal is to combine empirical and experimental research methods to devise approaches to test the autonomic behavior of changes to a cloud application prior to deployment. Further, models shall be investigated to estimate the impact of such changes on the financial costs of deployment, allowing to quantify the monetary costs of development decisions.

Expected Outcomes

As with any PhD project, the detailed goals are not specified upfront. Instead, it will be one of your tasks as a PhD student to find the most interesting and relevant open challenges (in the broader scope of performance testing and cost estimation for autonomic cloud apps) that you can (and want to!) address during your PhD. However, generally speaking, as a PhD student in the larger field of software technologies, you will be asked to produce three kinds of scientific outcomes:

  • New empirical knowledge, for instance through interview or survey studies, software repository mining, or mapping studies.
  • New concepts, approaches, techniques, or methodologies. These need to be developed following the principles of design science, and require extensive scientific evaluation (e.g., through controlled studies or other experiments).
  • Proof-of-concept implementations of the above in concrete prototypes and tools, which we usually make available as open source software.

All of these are to be communicated to other scientists through high-quality publications. We typically target a mixture of top-ranked and smaller, more specialized venues. As a young scientist, you will be judged based on your published work, so writing regularly will quickly become a non-optional part of your work routine.

The best PhD theses in our field have elements of all three of the above. This also means that you should have fairly broad interests and be willing to study and apply a vast array of different scientific methodologies, spanning from interviewing and survey research over statistical analysis to programming and system development.

Contributions are often developed in cooperation with industrial partners, such as IBM, SAP, Microsoft, Volvo, or Ericsson. However, this is not a purely industrial PhD project. That is, a focus will be put on novel concepts and fundamental research in software technologies, although evaluation of these concepts will often be on an industrial context.

Related Publications

The best way to find out what to expect when doing your PhD in this area is by looking at publications of my previous or current students. Below you find a few recent, representative examples:

Empirical Works:

Conceptual / Technological Works:

Things We Ask for

We require a MSc degree or equivalent, and a strong background in software engineering and Web-based systems (services- and cloud computing). Particularly, my research requires interest, knowledge, and experience in programming in a Web context, using languages such as Java, JavaScript, TypeScript, Ruby, or Go. Further, robust statistical and machine learning method knowledge (along with some hands-on experience in R and/or Python) is required for many paper projects we do, but a lot of it can be learned "on the job". Prior knowledge in autonomic computing is a plus.

Support You Will Get

As part of your PhD, you will be a salaried employee at the Chalmers University of Technology. You will be part of the Software Engineering division, one of the largest and most successful software engineering research groups in Europe. Your office will be located in beautiful Lindholmspiren, in a technology park shared between Chalmers, Ericsson, and others.

Full funding for your PhD for up to 5 years will be provided. This includes not only your salary, but also funds to cover equipment, travels to conferences, and other costs of doing research. There are no tuition fees for PhD studies at Chalmers. The tentative starting date is beginning of 2018, but somewhat flexible in both directions.

In addition to frequent meetings with your advisors, you will also be embedded in a closely-knit and successful research group operating in Sweden and Switzerland, with collaborators all over Europe and the world.

Chalmers continuously strives to be an attractive employer. Equality and diversity are substantial foundations in all activities at Chalmers.

Further Questions?

If you have any open questions, please contact Philipp via email.