[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Condor-users] Re: split processors among multiple jobs

hmm.. adding this to my .sub file doesn`t seem to be changing anything.. am i doing this correctly ?

MyClusterPriority = 1 priority = (MyClusterPriority * 1000) -$(Process)

The Priority for each submitted job still shows up at 0, and when 2 jobs are submitted, they don`t get "interleaved"

any ideas ?


Matt Hope wrote:

On Mon, 21 Mar 2005 17:16:41 -0600, Derek Wright <wright@xxxxxxxxxxx> wrote:

On Mon, 21 Mar 2005 11:34:33 +0000  Matt Hope wrote:

Note that pushing a new job to the top of the queue may trigger a
preemption of an existing job which might not be optimal, I would
suggest any accelerator script should find out the lowest priority of
the currently running jobs and never exceed that...

not true for user-defined job priorities (what we're talking about in
this thread). the schedd will never preempt a job because of these
priority values, so you don't have to worry about that. user
priorities (the fair-share stuff enforced by the negotiator, which you
can view/manage via condor_userprio) can cause preemption (if the
PREEMPTION_REQUIREMENTS expression evaluates to true), but job
priorities never do.

Does this apply if the jobs are, in essence, running only via the ranking by the execute machines... So if the schedd puts another job to the top of its list which is ranked lower by the execute machines this will not cause the currently running job previously at the top (which presumably had at least as good a machine rank) to be preempted.

I have behaviour like the above I assumed what was happening was down
to this mismatch (which did exist on their schedd) and educated the
user on making sure their TIER levels matched their user supplied
priority but if I was wrong I probably need to find out why the
preemption occurred.

Condor-users mailing list