Folks, When I use sacct, it always shows all tasks of an array job individually. This is supposed to be switched on by --array, but it seems to be the default. How do I switch it off? Thanks in advance, Herbert -- Herbert Fruchtl (he/him) Senior Scientific Computing Officer / HPC Administrator School of Chemistry, IT Services University of St Andrews -- The University of St Andrews is a charity registered in Scotland: No SC013532
Perhaps "squeue" will show you what you need(?). This: ``` squeue --all --state=all --long --array-unique ``` seems to have overlapping information. There're a number of fields that can be specified with `-O, --Format=<output_format>` that might have the info you need.
Hi Ron, Thanks for the response. I need information about historical jobs (squeue only shows what's in the queue now). I'll have to write a script to sum up the times for the individual jobs. Herbert On 20/01/2026 20:44, Ron Gould via slurm-users wrote:
Perhaps "squeue" will show you what you need(?). This:
``` squeue --all --state=all --long --array-unique ```
seems to have overlapping information. There're a number of fields that can be specified with `-O, --Format=<output_format>` that might have the info you need.
-- Herbert Fruchtl (he/him) Senior Scientific Computing Officer / HPC Administrator School of Chemistry, IT Services University of St Andrews -- The University of St Andrews is a charity registered in Scotland: No SC013532
Hello Herbert, I modified some SQL from my MariaDB history. This gets jobs that were started or ended after the first of the month. Does this help at all? ``` set @TimeStamp := UNIX_TIMESTAMP(DATE_FORMAT(CURDATE(), '%Y-%m-01 00:00:00')); select `account`, `partition`, `id_user`, `id_group`, `cpus_req`, `mem_req`, `nodelist`, `nodes_alloc`, `node_inx`, `id_job`, `job_name`, `job_db_inx`, `timelimit`, from_unixtime(`time_submit`, '%Y/%m/%d %H:%i:%s') as 'time_submit_HUMAN', from_unixtime(`time_eligible`, '%Y/%m/%d %H:%i:%s') as 'time_eligible_HUMAN', from_unixtime(`time_start`, '%Y/%m/%d %H:%i:%s') as 'time_start_HUMAN', from_unixtime(`time_end`, '%Y/%m/%d %H:%i:%s') as 'time_end_HUMAN', from_unixtime(`mod_time`, '%Y/%m/%d %H:%i:%s') as 'mod_time_HUMAN', `time_submit`, `time_eligible`, `time_start`, `time_end`, `mod_time`, `state`, `derived_ec`, `exit_code`, `work_dir`, `submit_line`, `flags`, `id_array_job`, `id_array_task`, `tres_alloc`, `tres_req` from cluster_job_table where ( time_start >= @TimeStamp OR time_end >= @TimeStamp ) order by `account` asc, `partition` asc, `id_user` asc, `id_job` asc ; ```
participants (2)
-
Herbert Fruchtl -
Ron Gould