10 cpus in a cluster. User A has submitted 100 jobs out of which 10 are running and 90 are idle. Then user B submits 100 jobs. All jobs of A and B are vanilla and of same size. Once the first 10 jobs of A are finished, B wants to get his first 10 jobs running. Once B's first 10 jobs are done, A wants to get his second 10 jobs running, ...., and so on until all jobs of A and B are done. What policy setting do I have to define to make that possible? I am currently using the TESTMODE policy and the dynamic priority calculation in condor does not seem to give equal shares to both with this policy.
No virus found in this outgoing message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.0 - Release Date: 29/4/2005