Improving the Job Success Rate through Analysis of User Logs in HPC

Abstract  !""#  ## !"   $ !" # "!""$""!!% #%! %!"  !  #   ! # !!  " "! " !&!!! !"  %#   ! ""! $!  !  "  #"!!" """! " !& !""  '  (! ""!   & #) &"! #   $ * &! !#  !   " +" ! """!  " !" #) &"!",!  '  !&) &"!!"# ! ) &"& !!   ! "$ (&! ) &" #  ! #    ! !"""  ""!  ) &,! #  !  "! !$"! " ! #) &"# ! !"*"! !""#  $ (! " ##+ !  $ ,!!"  #) &"! ! &"  ) &"!   !   '  ! ! "$ *" " ##) &"Keywords : HPC, Supercomputer, Scheduler, Batch job, Log Analysis※ Corresponding Author: ChanYeol ParkReceived : September 18, 2015Revised : October 19, 2015Accepted : October 22, 2015* Dept. of Supercomputing Center, KISTI Tel: +82-42-869-0581, Fax: +82-42-869-0569 email: jwyoon@kisti,re.kr** Dept. of Supercomputing Center, KISTI*** Dept. of Multimedia, Namseoul University**** Dept. of Supercomputing Center, KISTI

[1]  Henri Casanova,et al.  Dynamic Fractional Resource Scheduling versus Batch Scheduling , 2012, IEEE Transactions on Parallel and Distributed Systems.

[2]  Jemal H. Abawajy,et al.  An efficient adaptive scheduling policy for high-performance computing , 2009, Future Gener. Comput. Syst..

[3]  Bianca Schroeder,et al.  Reading between the lines of failure logs: Understanding how HPC systems fail , 2013, 2013 43rd Annual IEEE/IFIP International Conference on Dependable Systems and Networks (DSN).