On Fri, Mar 11, 2011 at 10:12 AM, Reuti <[email protected]> wrote:
> Am 11.03.2011 um 15:07 schrieb Lane Schwartz: > > > On Fri, Mar 11, 2011 at 8:58 AM, Reuti <[email protected]> > wrote: > > Am 11.03.2011 um 14:49 schrieb Lane Schwartz: > > > > > Rayson, > > > > > > Thanks for the pointer. In the qmon scheduler configuration, I have > > > "Job Scheduling Information" set to true. I assume that's the same > > > setting you're refering to? > > > > > > With this setting enabled, I still don't get very much info. When I > > > run qstat -j on my jobs, the only thing it tells me is that a queue > > > instance for a particular node is dropped because that node is > > > disabled. > > > > Because "disabled"? Did someone use `qmon` to disable the node or set up > any calendar? > > I believe our sysadmin intentionally disabled that particular node. The > message is: > > Maybe the sysadmin can also help you to investigate why your jobs are not > scheduled. Maybe it's by intention, e.g. by a resource quota or other > imposed limits. > We've got a pretty small group, and our sysadmin is as puzzled by this as we are. It's definitely not by intention. We don't have any resource quotas or other imposed limits (at least not that were intentionally set). The scheduler has a decay time of 07:30 with load adjustment of np_load_avg 0.5. The general parameters are: Algorithm: default Schedule Interval: 0:0:15 Reprioritize Interval: 0:0:0 Default Duration: INFINITY Max Jobs/User: 0 Sort by load Job Scheduling Information: True Load Formula: np_load_avg Flush Submit Seconds: 0 Flush Finish Seconds: 0 Maximum Reservation: 0 Params: none In our queues, all of the hard and soft limits values are set to INFINITY. Are there other settings that I should investigate that could be causing delays between when a job is queued and when it actually starts running? Thanks, Lane
_______________________________________________ users mailing list [email protected] https://gridengine.org/mailman/listinfo/users
