|
I've noticed that my Cullen Prime Search (LLR) 6.09 task llrCUL_91107183_2 is regularly checkpointing once more some 5 minutes after being preempted (while it has previously checkpointed exactly at the preemption moment):
16.06.2011 10:17:18 | PrimeGrid | [task] task_state=EXECUTING for llrCUL_91107183_2 from start
16.06.2011 10:17:18 | PrimeGrid | Restarting task llrCUL_91107183_2 using llrCUL version 609
16.06.2011 10:22:29 | PrimeGrid | [task] result llrCUL_91107183_2 checkpointed
....
16.06.2011 13:07:47 | PrimeGrid | [task] result llrCUL_91107183_2 checkpointed
16.06.2011 13:07:49 | PrimeGrid | [cpu_sched] Preempting llrCUL_91107183_2 (left in memory)
16.06.2011 13:07:49 | PrimeGrid | [task] task_state=SUSPENDED for llrCUL_91107183_2 from suspend
16.06.2011 13:12:55 | PrimeGrid | [task] result llrCUL_91107183_2 checkpointed
------
16.06.2011 13:53:13 | PrimeGrid | [task] task_state=EXECUTING for llrCUL_91107183_2 from unsuspend
16.06.2011 13:53:13 | PrimeGrid | Resuming task llrCUL_91107183_2 using llrCUL version 609
16.06.2011 13:53:56 | PrimeGrid | [task] result llrCUL_91107183_2 checkpointed
....
16.06.2011 16:45:03 | PrimeGrid | [task] result llrCUL_91107183_2 checkpointed
16.06.2011 16:45:03 | PrimeGrid | [cpu_sched] Preempting llrCUL_91107183_2 (left in memory)
16.06.2011 16:45:03 | PrimeGrid | [task] task_state=SUSPENDED for llrCUL_91107183_2 from suspend
16.06.2011 16:51:00 | PrimeGrid | [task] result llrCUL_91107183_2 checkpointed
Is it a known and expected behavior (for some unknown reason), or some hidden error (sort of post-mortem spasms? :-)
Peter | 
