SJSTAT(1)

📦
version 26.05.5 · 5 versions
, in epel 10

🌸 sjstat(1) — the SLURM job-stats glow-up 💅

okay bestie, so sjstat lives in the User Contributed Perl Documentation section of the manual (section 1, for the record ✨), and honestly? it's giving "let me peek at what's actually happening on this cluster" energy and I am obsessed.

✨ NAME ✨

sjstat — literally just: list attributes of jobs under the SLURM control. that's the whole vibe. 🎀

🦋 SYNOPSIS 🦋

sjstat [-h ] [-c] [-r ] [-v]

yes that's the exact spacing from the original, don't come for me 💀

💖 DESCRIPTION 💖

so the sjstat command is your new main character for displaying statistics of jobs under control of SLURM. the output is designed to give you the tea on resource usage and availablilty (yes, that's how it's spelled in the original, keeping it fr fr 🫶), plus all the info about jobs that are currently active on the machine.

this output is built using the SLURM utilities sinfo, squeue and scontrol — and no cap, the man pages for those utilities will give you more information and greater depth of understanding. like, go read them. iykyk. 📚

🌈 OPTIONS 🌈

  • -h — display a brief help message. tiny lil rescue when you're lost 💫
  • -c — display the computing resource information only. just the resources, no job gossip 🙈
  • -man — show the man page. iconic self-reference behavior, honestly 👑
  • -r — display only the running jobs. skip the pending girlies 💀
  • -v — display more verbose information. more fields, more lore, more everything 🔥

⭐ EXAMPLE ⭐

the following is a basic request for status. slay, let's look:

> sjstat

 Scheduling pool data:
 ------------------------------------------------------------
 Pool         Memory  Cpus  Total Usable   Free  Other Traits
 ------------------------------------------------------------
 pdebug      15000Mb     8     32     32     24  (null)
 pbatch*     15000Mb     8   1072   1070    174  (null)

 Running job data:
 -------------------------------------------------------------------
 JobID    User      Nodes Pool       Status        Used Master/Other
 -------------------------------------------------------------------
 395      mary       1000 pbatch     PD            0:00 (JobHeld)
 396      mary       1000 pbatch     PD            0:00 (JobHeld)
 375      sam        1000 pbatch     CG            0:00 (JobHeld)
 388      fred         32 pbatch     R            25:27 atlas89
 361      harry       512 pbatch     R          1:01:12 atlas618
 1077742  sally         8 pdebug     R            20:16 atlas18

🧚 the scheduling data, explained

the scheduling data contains information pertaining to the:

  • Pool — a set of nodes
  • Memory — the amount of memory on each node
  • Cpus — the number of cpus on each node
  • Total — the total number of nodes in the pool
  • Usable — total usaable nodes in the pool (their spelling, not mine, i promise 😭)
  • Free — total nodes that are currently free

🍓 the running job data, explained

the running job data contains information pertaining to the:

  • JobID — the SLURM job id
  • User — owner of the job
  • Nodes — nodes required, or in use by the job (Note: on cpu scheduled machines, this field will be labeled "Procs" show the number of processors the job is using.)
  • Pool — the Pool required or in use by the job
  • Status — current status of the job
  • Used — Wallclick time used by the job (again — exactly as written in the original 💀)
  • Master/Other — either the Master (head) node used by the job, or may indicate further status of a pending, or completing job.

💜 the common status values

  • R — the job is running 🏃
  • PD — the job is Pending ⏳
  • CG — the job is Completing 🎬

these are states reported by SLURM and more elaborate documentation can be found in the squeue/sinfo man pages. go get that deep lore, bestie 🫶

✨ an example of the -v option ✨

Scheduling pool data:
-----------------------------------------------------------------------------
                           Total  Usable   Free   Node   Time  Other
Pool         Memory  Cpus  Nodes   Nodes  Nodes  Limit  Limit  Traits
-----------------------------------------------------------------------------
pdebug      15000Mb     8     32      32     24     16     30  (null)
pbatch*     15000Mb     8   1072    1070    174  UNLIM  UNLIM  (null)

Running job data:
---------------------------------------------------------------------------------------------------
                                                     Time        Time            Time
JobID    User      Nodes Pool      Status        Used       Limit         Started  Master/Other
---------------------------------------------------------------------------------------------------
38562    tom           4 pbatch    PD            0:00     1:00:00  01-14T18:11:22  (JobHeld)

🌸 the added fields to the "Scheduling pool data"

  • Node Limit — SLURM imposed node limit.
  • Time Limit — SLURM imposed time limit, value in minutes.

🌸 the added fields to the "Running job data"

  • Limit — time limit of job.
  • Start — start time of job.

💀 REPORTING BUGS 💀

found a bug? report bugs to <eckert2@llnl.gov> — be nice, be specific, include the deets 🙏

bestie reminder: the man pages for sinfo, squeue and scontrol are your besties too. sjstat is literally built from them, so that's where the full picture lives 💅

that's the whole page, no cap. this edition of the manual was brought to you by perl v5.40.2, dated 2026-10-14 — mwah, go check your queue 🎀✨🦋

📦 where this page comes from