On Thursday 11 February 2010 2:19:26 pm Reddy, John wrote:
> One of my users, bless his heart, has requested I install a tool he called
> "modules".  That's all he's been able to describe it to me as in terms of
> name.  Apparently, it's a program that allows users to load or unload
> grouped sets of environment variables.
>
> Does anyone have this tool, know what it's called, who's developing it,
> etc?  Yes, I know, horribly, terribly vague software description.  Here's
> the context, which may help identify it.
>
> We've got a clustered processing environment with 120 dual-quad nodes
> running a variety of SL 4.x and 5.x with job control via Torque & Moab. 
> We've got three different compilers with multiple versions each, a variety
> of implementations of MPI, etc.  So a tool such as this would be useful for
> my users.
>
> Now the MPI selection is easily handled with mpi-selector.  I could
> probably (easily) enlist that tool for environment selection.  However, I'd
> like to see if I can find someone using the tool my user requested.
>
> TIA for any thoughts on the matter.
>
> -John
>
>
> ---
> John Reddy
> Technology Architect
> Information Technology Division - Unix Services
> Brookhaven National Laboratory
> [log in to unmask] - 631-344-3284

John:

Actually your user is correct.  "Modules" really is the name of the tool.

googling 'modules environment' provides a lot of information.  The first hits 
are the sourceforge page for the project, and a wikipedia article that should 
be very illuminating.

It is a very useful package for environments where you have different versions 
of the same tools.   "modules" are described by simple tcl scripts that can 
be defined by the administrator or by users themselves.  Then users can 
select between versions by simple commands.  

It was commonly used on Solaris, and I've used the linux version in a 
clustering environment much like you describe.  You may find it gives you 
much more flexibility than mpi-selector.  For example, if you have tools that 
are compiled with different versions of compilers, for different versions of 
MPI.


-- 
--------------------------------------------------------------
Jeffrey Anderson    			| [log in to unmask]
Lawrence Berkeley National Laboratory	| 
Office: 50A-5104E			| Mailstop 50A-5101
Phone: 510 486-4208			| Fax: 510 486-4204