[gridengine users] execd load sensors timing
reuti at staff.uni-marburg.de
Mon Jul 9 14:56:41 UTC 2012
Am 09.07.2012 um 16:16 schrieb William Hay:
>> That the values are still reported from the last run in `qhost -F ...`. But when the reboot is taking only a few minutes the load sensor would report the same value as before. Or do you upgrade the OS in just a load_report interval, so that the old value would be wrong?
> Are you saying that a wait of load_report_time should be sufficient
> for grid engine to notice and trigger a load threshold/prevent
> scheduling? I could easily guarantee such a short delay
> before execd comes up.
In case you want to avoid that the old value is reported: yes.
But this won't prevent that a job is scheduled thereto between the startup of the execd and the next cycle of the report interval. Between these two events some values are just not reported. So, a job not requesting these load values, can be scheduled thereto. This I just confirmed in my cluster.
More information about the users