[Aces-support] memory

[Florian Bleibinhaus]

Hi,

I have a problem with jobs getting killed - probably because they want more than 1 GB of RAM. Most of my jobs are not parallel, and I submit them to the "one" queue, requesting one node.
How can I make sure they end up on a node that has 2 GB - other than requesting a specific node? Specifying a range of nodes=aE34-500-011/aE34-500-081 doesn't seem to work.

Thanks,
Florian

Florian Bleibinhaus, PhD - bleibi at mit.edu
MIT - Earth Resources Lab - E34-454
42 Carleton St, Cambridge, MA 02142
(++1) 617.253.5857 - Fax .6385 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://forge.csail.mit.edu/pipermail/aces-support/attachments/20070819/18d109d4/attachment.html