Want to help out?

If you would like to help out with the Clumon project and have the skills, ambition or even just an idea or sugestion, e-mail Joshi Fullop and we'll see what we can work out.

Here are a few ideas where Clumon could use a hand that are not overly technical. You will certainly get credit for your contributions if they get used.

For more current info on these and other contribution topics, see the Clumon Message Board.


Host Page Layout (machine Dashboard)

Currently : Clumon just displays all the metrics it collects in an unorganized table.
Task : Design a layout that intuitively gives vital machine statistics concerning memory, cpu, various disk data, process tree (can be expandable), etc. to take the place of host.php.
Restrictions : Use of the current clumon.css and color scheme is preferred. But if you have a really great layout, send it to us anyway.


Process Tree Layout

Currently : Clumon displays the processes with their PID in a rasterized tree.
Task : Design a layout that shows the process hierarchy and is able to work in more process data possibly by means of mouse-overs or expanded tabular view.
Restrictions : Must work in the majority of browsers. This will generally rule out some of the cooler, more sophisticated tools and techs. However things like Flash or Shockwave can be pretty cool, and the cool-factor can sometimes override the need for some users of Clumon to have a vanilla browser compliant compatibility and they may require or just preferr to use IE or Netscape. Use of the current clumon.css and color scheme is preferred. But if you have a really great layout, send it to us anyway.


More Charts and Graphs

Currently : Clumon has a few graphs it shows for load history for a machine and aggregate load history across a job.
Task : Create mock graphs (or code real ones based on the graph.php or graph-job.php) that show useful information.
Restrictions : None really. Be creative. Just avoid any refference to purple (or green) dinosaurs.



Currently : Clumon shows a spatial correlation of running jobs to the actual machines.
Task : Create mock graphs (or real ones if you are up to it) that show data in a useful format.
Restrictions : None. Again no purple or green dinosaurs.
Ideas : Network weather patterns (for entire cluster or a subset, e.g a job's nodes); job que/reservation data - what jobs are up next, how many nodes are they slated to use and what is the time they are set to consume