[Beowulf] Definition of HPC

[Joe Landman]

On 04/16/2013 03:50 PM, Max R. Dechantsreiter wrote:
> ...Perhaps HPC should be defined problem-down, rather than hardware-up.
>
> Regarding your "Issue #1:"
>
> There are some _nearly_ EP problems that require very minimal, but crucial,
> communication (perhaps at long intervals).
>
> As to your "Issue #2:"
>
> "Owned compute" has some advantages over "rented compute."  In general, the
> control one has over one's owned resources enables applications to run with
> greater performance.  Some optimizations just demand root access!

[stunt pilot: do not try this at home or anywhere else]

     pdsh rm -rf /

[/stunt pilot]

Must be the cold medicine talking ...

I've seen well intentioned researchers destroy carefully built systems 
with root access.  I'd be very very careful about having them ever get 
used to running as root.  If there's a process they need to run with 
elevated privs for whatever reason, there's no dearth of mechanisms to 
enable that specific task.

Having spent many man-days/man-weeks on phones/in cages with 
researchers/it staff after the dejecta hit the air movement device ... 
yeah ... I'm not so thrilled about giving sleepy grad students access.  
Talk about giving them enough rope to blow their toes off with (mixed 
metaphor, on purpose).

-- 
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics, Inc.
email: landman at scalableinformatics.com
web  : http://scalableinformatics.com
        http://scalableinformatics.com/siflash
phone: +1 734 786 8423 x121
fax  : +1 866 888 3112
cell : +1 734 612 4615