partitioning unstructured with SF for parallel use

Sasha

Hi. Can somebody suggest me on the following questions for unstructered mesh built non-uniformly (with SF): * will auto-partitioning take account on SF and devide the domain in more or less equally distributed cell number partitions * is it worth it to partition it by myself trying to equally distribute the number of cells between the partitions * are there some specific guidelines for dealing with DO radiation model in parallel solver * how beneficial is using some non-fluent soft for partitioning