Scientific Experiment management with Commodity Tools
From Java CoG Kit
Contents |
Who?
? is a Ph.D student at ?. His supervisor is ?. His research interrests are? and ? .... As part of his school requiremnet he will be working on a Masters thesis "Scientific experiment Management with Commodity Tools" under supervision of Gregor von Laszewski from Argonne National Laboratory.
Deepak: link to your resume.
Deepak: I need a name of a school oficial that I can corespnd to in case of an emergancy or other wise ... E.g. in case I spend a lot of time with you and your would decide for example to drop out, I like to at least have a name to discuss this with somone at the school.
What ?
Proposal by Gregor von Laszewski: Scientific Experiment Management with Commodity Tools
This project will identify and evaluate practically how commodity tools can be utilized for scientific experiment management.
The first task is to identify what scientific experiment management is. Keywords that hel in identifying this are:
- Commodity Toolkits
- Parameter Studies
- Problem Solving Environments
- Scientific Gateways
- ...
You will also be able to run through an experiment use case yourself such as
- How is a physics or chemistry experiment conducted
- study, idea, plan, setup, conduct exp., collect data, evaluate results
A question to be answered is which commodity technologies are useful for the gaol of scientific experiemnt management. As you can see from the usecase a lot of technologies could be helpful at different stages of the experiemnt. It woudl be good to more carefully analyse one or more of these use cases and identify promissing technologies that support the various sub tasks.
One of the most prominent examples today is the conduct of parameter studies on suprecomputing centers. Let us assume a scientist is running multiple scientific parameter studies or experiements refered to from now on. Each experminent contanisn a number of tasks to be done that may or may not depend on each other. Each experiment could have a particular priority (compare to Nimrod and to our experiment management tool on sourceforge, keyword cogkit).
Now one could go ahed and implement a user interface to help assiting in setting up and configuring such experiments. However there are for example in the MS tools already a number of commodity tools that may with modifications support this quest.
a) Microsoft Visual studio 2006 could be used to create sophisticated models that than can be executed and support the design of the workflow model of the experiment or multiple experiments. Interesting here is to evaluate how foar this can be done, where the shortcomings are, and how it can be practically enahnced to be made more useful.
b) Microsoft Project is a good tool to organize once tasks based on dependencies. Here we can ask the same question. Can we extend project to assist the user in coordinating a large number of tasks?
In your project you will be required to at least develop a components and software that can be integrated within microsoft project to assist in the creation of experiment tasks submitted to the Grid.
We assume a new window gets poped up where a task or group of tasks could be defined that upon closing would be integrated into the gant chart of microsoft project. manipulation of the task and the status monitoring should be enabled.
We recommend that you use the Java CoG Kit (a bridge for a small number of tasks to C# should be easy,). We prefer that you program in C# rather than VB.
How this should look like should be proposed by you through a design document taht we itterate between each other.
References:
Deepak identify some references
- [1] ???
Related Research and Efforts
Progress
The progress of the project is monitored weekly. It is mandatory for the student to fill out this information. In case the student has not worked in that particular week it is supposed to be entered acordingly. If it takes you more than ten minutes to fill out the information, you do something wrong. In case you read references and they are avialable online, please add hyperlinks:
In reverse order (latest week first):
Week Dec 4, 2005 - Dec 6, 2005
- What have I achieved this week?
- What have I not achieved this week?
- What do I plan to achive next week?
