Slave STDOUT handler


#1

Hi All,

You can override stdout handlers in python per plugin. Is there also a way to add stdout handlers for slave logs?

We have an issue where slaves are unable to dequeue a job (for which I submitted a ticket). This doesn’t show as a job report as the job isn’t started. The only output that I get is on the slave itself, in the slave log.

Is there a way to add a handler to this, so I can change the priority to 1?

The output on the slave looks like the code below. I can’t seem to find it anywhere else.

Kind regards,
Marum

2019-01-11 15:10:47:  Scheduler - Performing Job scan on Primary Pools with scheduling order Pool, Priority, First-in First-out
2019-01-11 15:10:47:  Scheduler - Unexpected Error Occurred
2019-01-11 15:10:47:  >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
2019-01-11 15:10:47:  Exception Details
2019-01-11 15:10:47:  NullReferenceException -- Object reference not set to an instance of an object.
2019-01-11 15:10:47:  Exception.Data: ( )
2019-01-11 15:10:47:  Exception.TargetSite: Deadline.Jobs.Task DequeueNextTask(System.String, Int32, System.String, System.DateTime, Boolean ByRef, Boolean ByRef)
2019-01-11 15:10:47:  Exception.Source: deadline
2019-01-11 15:10:47:  Exception.HResult: -2147467261
2019-01-11 15:10:47:    Exception.StackTrace: 
2019-01-11 15:10:47:     at Deadline.StorageDB.MongoDB.MongoJobStorage.DequeueNextTask(String jobID, Int32 preferredTaskID, String slaveName, DateTime networkDateTime, Boolean& triggerStartEvent, Boolean& triggerFailedEvent)
2019-01-11 15:10:47:     at Deadline.Controllers.JobQueueController.DequeueNextTask(String jobID, Int32 preferredTaskID, String slaveName, DateTime networkDateTime, Boolean& triggerStartEvent, Boolean& triggerFailedEvent)
2019-01-11 15:10:47:     at Deadline.Controllers.DataController.t(Job bvn, Int32 bvo, String bvp, DateTime bvq, Boolean& bvr, Boolean& bvs)
2019-01-11 15:10:47:     at Deadline.Controllers.DataController.DequeueTasks(SlaveState slaveState, SlaveSettings slaveSettings, Boolean verboseLogging, Boolean secondaryScan)
2019-01-11 15:10:47:  <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
2019-01-11 15:10:47:  Scheduler - Performing Job scan on Secondary Pools with scheduling order Pool, Priority, First-in First-out
2019-01-11 15:10:47:  Scheduler - Job chooser found no jobs.
2019-01-11 15:10:47:  Scheduler - Returning all limit stubs.
2019-01-11 15:10:47:  Scheduler -   returning 5c388e476de04f3ee4ba91d9
2019-01-11 15:10:47:  Scheduler -   returning _fh
2019-01-11 15:10:47:  Scheduler -   returning redshift
2019-01-11 15:10:48:  Scheduler Thread - Seconds before next job scan: 11

#2

Unfortunatley, the standard output handling is only attached to processes the Sandbox launches as part of the run of a render plugin (say “[repo]\plugins\MayaBatch\MayaBatch.py”).

That error though is a pretty serious problem as it’s dying while trying to grab a task. Does it happen often?

Other question I suppose, are you using pre/post job scripts? We should make this error a bit more helpful if we can.