Hi all,
Some days ago on a freshly installed 6.2u5 I've done some tests and
almost every serial job run on the same machine which is already
loaded from interactive use and all the other (empty) machines where
ignored from sge.
After some days of testing, now I can see the "usual" logic and the
use of the least loaded nodes.
Previously I was thinking that SGE always allocate resources on least
loaded nodes but now I'm not so sure...
Where can I read some info about how sge choose the slots for a job?
Is that logic somehow configurable?

Thanks
Stefano
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users

Reply via email to