On Tue, Mar 23, 2010 at 12:59 PM, Jed Brown <[email protected]> wrote:
> On Tue, 23 Mar 2010 12:48:59 -0500 (CDT), Roy Stogner 
> <[email protected]> wrote:
>
>> it looks like bumping up the output file sizes by 60% would take up an
>> extra 6MB in the examples.  Trivial normally, possibly the last straw
>> for someone on a machine with tiny quotas.
>
> I have no interest one way or the other, but it's worth pointing out
> that 1 GB costs well under a dime on consumer disks, somewhat more on a
> higher performance file system.  I think the point where 6 MB of disk is
> lost in the noise has past several years ago, and administrators that
> fail to acknowledge this ought to be notified.  Of course output from
> production runs is an entirely different matter, but that's not what the
> present discussion is about.

Definitely correct in general.  Ben and Roy are worried about the very
specific case of lonestar.tacc, where $HOME's have only 200Mb quotas.
It would certainly be nice (and many users have asked us) to increase
that, but I guess there is no funding allocated specifically to that
purpose.  Also, I believe it's NFS over GigE so there may be some
questions of the scalability of that setup if only the disk and not
the network itself were upgraded.

-- 
John

------------------------------------------------------------------------------
Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Libmesh-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libmesh-devel

Reply via email to