[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [HTCondor-users] Changed EXECUTE macro but TotalDisk = 0

Sorry to do this, but I just wanted to bump this thread.  It's pretty important to what I'm doing and I am very stumped on it.  Any suggestions would be appreciated!

Thanks again,

On Tue, Jan 15, 2013 at 1:39 PM, Michael Aschenbeck <m.g.aschenbeck@xxxxxxxxx> wrote:
Hello Everybody,

Our Machines' C drives (where condor is installed) are pretty full and we've had some 'lack of scratch space' problems lately.  Each machine has ample space on another drive (E: for this example) so on one Windows machine I modified the EXECUTE variable condor_config file to be

EXECUTE = E:\Scratch/execute

(I used the backslash and forward slash to exactly emulate what it was doing by default with LOCAL_DIR = C:\condor and EXECUTE = $(LOCAL_DIR)/execute, which worked fine.)  I created the directory on the machine and gave all users full permissions and restarted condor.

The machine shows up via condor_status, however, the Machine's classad shows TotalDisk = 0, Disk = 0.  I know Condor "sees" the directory because when I delete the directory the machine no longer shows up when running condor_status.

So the question is, how do I get Condor to refresh TotalDisk and Disk so that jobs can be matched on this machine?  I've seen this question a few times in the archives with no answer.  There is a work-around (although I haven't tried it) to add the requirement Disk >= 0 in the submit file, but I need a more graceful solution so I can better utilize the full classad system in the future.

Thanks in advance for your help,