[HTCondor-users] dynamic allocation of RAM

[Thomas Hartmann <thomas.hartmann@xxxxxxxx>]

hi,
we have been using htcondor quite successfully so far. yet, there is one 
issue that still needs resolving:

at the moment, users are required to request the amount of RAM their 
jobs need. if the requested amount of RAM is exceeded, the job gets 
killed. this, of course, gets quite annoying for the users, especially 
when they run jobs that take several days. this leads to the non-optimal 
behavior of requesting excessive amounts of RAM, thereby blocking other 
jobs from running.

my question is, if there would be a better solution available. 
brainstorming myself, i came up with the following ideas:

1. punish users by giving the a worse priority if they ask for excessive 
amount of RAM. i did not find a way yet to do that....

2. handle RAM allocations more dynamically. for instance:
2.1. if a job wants to use more RAM than previously requested, see 
whether the machine on which it runs still has this amount of RAM available.
2.2. if it does, update the request_memory to a safe value and continue 
running the job.
2.3. if the extra RAM is not available, stop the job, update the 
request_memory to a safe value and put it back into the queue.

any idea, how the second option could be done?

thanks in advance and best regards,
thomas

--
Dr. Thomas Hartmann

Centre for Cognitive Neuroscience
FB Psychologie
UniversitÃt Salzburg
HellbrunnerstraÃe 34/II
5020 Salzburg

Tel: +43 662 8044 5109
Email: thomas.hartmann@xxxxxxxx

"I am a brain, Watson. The rest of me is a mere appendix. " (Arthur Conan Doyle)