Cloudmesh Dirstributed Virtual Cluster Project

Motivation

Cloudmesh is a large set of python code for manageing cloud and HPC resources. You will have the chance to work on us on new commands and code improvements for cloudmesh. We will be developing a distributed cluster spawning multiple super computers for high throughput computing.

Task

The first part of this project project you will be providing a high quality scholarly survey and review of various HTC tools and technologies. This project will be conducted in co-authorship with Gregor von Laszewski. The document will be using Word. References must be maintained in endnote or Mendeley. If possible Microsoft Live will be used to share the document. Alternatively the paper can be written in LaTeX.

The second part of this project will be to develop a cloudmesh virtual cluster command that allows millions of jobs to be scheduled across multiple clusters.

Details will be discussed in our meetings in this project. Multiple students can participate in the project, but responsibilities will be clearly identified.

The authors of the paper will not plagiarize. We anticipate the maximum number of students will be three. If more students sign up we may identify a or additional topics to work on an additinal features. However the projects will not duplicate each other.

Artifacts

  • Working program submitted into github
    • integration with cloudmesh
    • application demonstartion
  • Scholarly paper
    • Report written in IEEE format in LaTeX or Word
    • Original images used in the report
    • To make citation of scholarly work easier we not only recommend but require that you use a bibliography management tool such as endnote, Mendeley for word or bibtex for LaTex. In case of MS Word, if you do not have access to endnote, please use the build in citation manager from Word. In case of bibtex, you can also use jabref.

Optional Artifacts

Samples of using the technologies identified.