I don't think you need to do extra anything, the Sheduler on JT will do it
for you. When there's node becoming free, Scheduler will assign new task to
On Mon, Jan 18, 2010 at 2:02 PM, Yanfeng Zhang <[EMAIL PROTECTED]> wrote:
> Hello, all
> I want to implement this.
> 1. In a 10-nodes cluster, there is a job J1 running on 5 nodes, at the same
> time there is another job J2 running on the other 5 nodes.
> 2. Then J1 finishes earlier and J1's 5 nodes join to work on job J2, so we
> have 10 nodes working for J2.
> Could anyone tell me how to implement it?
> Thank you very much!