Dear Todd and Ian, I thought I needed to explicitly indicate
in the wrapper program the path to the executable file. But if I understood you right, Todd, Condor
tells to my job wrapper where the file is located. And as you said Ian, the wrapper doesn't
have to be in the same directory where the job will run from. That's great. Now, the question is: I would like that Condor runs the job
file using the following command (on Linux), wine cmd /c condor_exec.bat where condor_exec.bat
is the transferred file. How should the wrapper program (my_job_wrapper.sh)
look like? Could it simply be, ____________________________ #!/bin/sh wine cmd /c condor_exec.bat exec "$@" ______________________________ …doing like this the job gets
held. Or do I have to define _______________________________ #!/bin/sh wine cmd /c /var/lib/condor/execute/dirxxx/condor_exec.bat exec "$@" ______________________________ Thanks for your help! I still have lots to learn… but
this is quite interesting and very useful! : ) /Sónia -----Ursprungligt meddelande----- Sónia Liléo wrote: > Thanks Todd! > > > You wrote, > > ________________________________________________ >> Could you please also confirm if the following
is correct, >> >> If I define in the .sub file transfer_executable
= true, then the >> executable file will be transferred to the
var/lib/condor/execute >> directory of the execute machine. >> >> Is this correct? >> > > Yes. Actually, it will be transferred into a
dynamically created > subdirectory of the directory specified by EXECUTE
in the > condor_config. Note that transfer_executable is True
by default. > > __________________________________________________ > > According to the description of the USER_JOB_WRAPPER
setting, I need > to define in the wrapper program where the
executable file is > located. > > "The command-line arguments passed to this
program (wrapper program) > will include the full-path to the actual user job
which should be > executed, followed by all the command-line
parameters to pass to the > user job." > > I wonder how do I know the name of the dynamically
created > subdirectory into which the executable file is
transferred? > > On other words, how should I define the
"full-path to the actual user > job which should be executed"? As it says above, the full path to the user job is passed
to your job wrapper on the command line. Your job wrapper doesn't
need to worry about finding the dynamically created subdirectory, since
that subdirectory will be included as the basename in the
"full-path to the actual user job" which is passed to your job
wrapper. Did that help clarify? If not, then I am not understanding your
question... regards Todd _______________________________________________ Condor-users mailing list To unsubscribe, send a message to
condor-users-request@xxxxxxxxxxx with a subject: Unsubscribe You can also unsubscribe by visiting https://lists.cs.wisc.edu/mailman/listinfo/condor-users The archives can be found at: https://lists.cs.wisc.edu/archive/condor-users/ |