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

[Condor-users] Interesting use of backfill

Hi guys,

Wanted to share with you our interesting experience with startd backfill
capabilities, which appear to be very powerful and nicely applicable out
of the scope of Condor.

We are using BOINC to gain access to our Technion resources. However we
found that there is a severe problem with BOINC due to the fact that it
does not monitor anything but keyboard and mouse. So if a machine has no
RAM available, or CPU load is high, it will start the job in any case.

Interestingly, we found that the new startd capability of starting
backfill BOINC client comes really handy here.
We specify our program be executed under BOINC as backfill, and startd
invokes it according to BACKFILL_START expression. Now, BOINC actually
starts startd, which in
turn monitors the machine and starts the actuall executable.

We also have a standalone version of the program which automatically
creates the config file with the correct settings, allowing to "backfill"
any program in a standalone mode, outside of BOINC.

The only problem was to kill startd when the actual program termintates.
So at the moment we make the program that is started by startd kill startd
right before the termination. But it's awkward. If we had a parameter in
startd, which would trigger it to suicide when backfill executable dies
itself - this would be fantastic. We have no problems with fixing that
ourselves, but we thought maybe this parameter can be added in 6.9.x

Any other ideas would be appreciated!