ACCOUNTING(5) File Formats Manual ACCOUNTING(5) NAME accounting - Univa Grid Engine accounting file format DESCRIPTION An accounting record is written to the Univa Grid Engine accounting file for each job having finished. The accounting file is processed by qacct(1) to derive accounting statistics. FORMAT Each job is represented by a line in the accounting file. Empty lines and lines which contain one character or less are ignored. Accounting record entries are separated by colon (':') signs. The entries denote in their order of appearance: qname Name of the cluster queue in which the job has run. hostname Name of the execution host. group The effective group id of the job owner when executing the job. owner Owner of the Univa Grid Engine job. job_name Job name. job_number Job identifier - job number. account An account string as specified by the qsub(1) or qalter(1) -A option. priority Priority value assigned to the job corresponding to the priority param- eter in the queue configuration (see queue_conf(5)). submission_time Submission time (64bit GMT unix time stamp in milliseconds). start_time Start time (64bit GMT unix time stamp in milliseconds). end_time End time (64bit GMT unix time stamp in milliseconds). failed Indicates the problem which occurred in case a job could not be started on the execution host (e.g. because the owner of the job did not have a valid account on that machine). If Univa Grid Engine tries to start a job multiple times, this may lead to multiple entries in the accounting file corresponding to the same job ID. exit_status Exit status of the job script (or Univa Grid Engine specific status in case of certain error conditions). The exit status is determined by following the normal shell conventions. If the command terminates nor- mally the value of the command is its exit status. However, in the case that the command exits abnormally, a value of 0200 (octal), 128 (decimal) is added to the value of the command to make up the exit sta- tus. For example: If a job dies through signal 9 (SIGKILL) then the exit status becomes 128 + 9 = 137. ru_wallclock Difference between end_time and start_time (see above). The remainder of the accounting entries follows the contents of the standard UNIX rusage structure as described in getrusage(2). Depending on the operating system where the job was executed some of the fields may be 0. The following entries are provided: ru_utime ru_stime ru_maxrss ru_ixrss ru_ismrss ru_idrss ru_isrss ru_minflt ru_majflt ru_nswap ru_inblock ru_oublock ru_msgsnd ru_msgrcv ru_nsignals ru_nvcsw ru_nivcsw On Windows, only the values ru_wallclock, ru_utime and ru_stime are accounted. These values are the final usage values of the Windows Job object that is used to reflect the Univa Grid Engine Job, not the sum of the usage of all processes. project The project which was assigned to the job. department The department which was assigned to the job. granted_pe The parallel environment which was selected for that job. slots The number of slots which were dispatched to the job by the scheduler. task_number Array job task index number. cpu The cpu time usage in seconds. mem The integral memory usage in Gbytes cpu seconds. io The amount of data transferred in Gbytes. On Linux data transferred means all bytes read and written by the job through the read(), pread(), write() and pwrite() systems calls. On Windows this is the sum of all bytes transferred by the job by doing write, read and other operations. It's not documented what these other operations are. category A string specifying the job category. iow The io wait time in seconds. ioops The number of io operations. pe_taskid If this identifier is set the task was part of a parallel job and was passed to Univa Grid Engine via the qrsh -inherit interface. maxvmem The maximum vmem size in bytes. arid Advance reservation identifier. If the job used resources of an advance reservation then this field contains a positive integer identifier oth- erwise the value is "0" . ar_submission_time If the job used resources of an advance reservation then this field contains the submission time (64bit GMT unix time stamp in millisec- onds) of the advance reservation, otherwise the value is "0" . job_class If the job has been running in a job class, the name of the job class, otherwise "NONE" . qdel_info If the job (the array task) has been deleted via qdel, "@", else "NONE". If qdel was called multiple times, every invocation is recorded in a comma separated list. maxrss The maximum resident set size in bytes. maxpss The maximum proportional set size in bytes. submit_host The submit host name. cwd The working directory the job ran in as specified with qsub / qalter switches -cwd and -wd. As the delimiter used by the accounting file (colon ":") can be part of the working directory all colons in the working directory are replaced by ASCII code 255. submit_cmd The command line used for job submission. As the delimiter used by the accounting file (colon ":") can be part of the command line all colons in the command line are replaced by ASCII code 255. Line feeds being part of the command line will be replaced by a space character. The qacct client converts them back to colon. For jobs submitted via the DRMAA interface or via qmon graphical user interface the accounting file contains "NONE" as submit_cmd. wallclock The wallclock time the job spent in running state. Times during which the job was suspended are not counted as wallclock time. For loosely integrated jobs and for tightly integrated jobs with accounting summary enabled the wallclock time reported for the master task is the wall- clock time multiplied by the number of job slots. SEE ALSO sge_intro(1), qacct(1), qalter(1), qsub(1), getrusage(2), queue_conf(5). COPYRIGHT See sge_intro(1) for a full statement of rights and permissions. Univa Grid Engine File Formats UGE 8.5.4 ACCOUNTING(5)