Hi! I have been using and playing with a lot of
vm scenaries with Condor for this last month, and I really need a little help
in my learning. I want to serve a "Biological" online
videogame that requires the processing of certains actions, for instance,
scripts like: START GEN=1 So in this simple example, the user should
see how every 10 seconds a new gen is generated. (The example is not real, the
simulations aren't that simple) The beauty of this system is that almost
any user isn't processing his "game scenario" (gens, etc). Questions are: 1) I managed to get a condor install for
windows in less than 11 MB (I think can be less than that), but I don't really
understand why condor does not start jobs behind a NAT configuration. Can I
redirect everything that condor needs by the port 80? 2) Task are a bit lenght (the should run
receiving modifications to the simulation for more than 6-12 hours), but every
task has a lot of idle times. Can I run multiple jobs in just one cpu without
dealing with cpu cores/dynamic slots? Thanks! |