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

Re: [Condor-users] preempt and then hold?

On Tue, 02 Aug 2005 13:40:24 -0500  Derek Wright wrote:

> sorry, that's not really true (granted, the startd's code is
> confusing, and i'm sure the behavior isn't obvious just from looking
> at it).  the startd is recomputing the imagesize of the starter and
> all it's children much more frequently than the UPDATE_INTERVAL.  you
> only see the changed version with condor_status every UPDATE_INTERVAL
> and/or state change (which might be what mike's talking about).
> however, anything that's potentially changing rapidly about a job or
> machine which a machine owner might want to use for policy expressions
> is recomputed every "POLLING_INTERVAL" (defaults to 5 seconds).
> examples are the load average and condor load average (don't ask),
> memory usage, etc.  so, the startd itself has a new version of the
> imagesize every 5 seconds, and you could watch that with
> "condor_status -direct" if you wanted to.

ugh, i'm sorry.  i wasn't entirely right, either. ;) everything i said
above is true.  however, due to a small bug in the startd (which i
just noticed), even though the startd has accurate values for the
image size every POLLING_INTERVAL seconds, it only *uses* the new
value when evaluating the policy expressions (like PREEMPT) every
UPDATE_INTERVAL seconds!  whoops. ;) unfortunately, i noticed this bug
after i already released 6.7.10.  but, it'll be fixed in 6.7.11 and
probably i'll put the fix into 6.6.11, as well.

apologies to everyone, especially mike yoder, who probably saw the
code i consider a bug, he thought it was a feature, and gave the
accurate answer to the original question. ;)

mea cupla,