slurm_submit_workers(1)

✨ slurm_submit_workers(1) — the Cooperative Computing Tools one ✨

okay bestie, gather round, because this is the little command that yeets your work_queue_worker onto a SLURM cluster so you don't have to babysit it 💅

---

🌸 NAME 🌸

slurm_submit_workers — submit work_queue_worker to a SLURM cluster. that's the whole job, no cap.

---

🦋 SYNOPSIS 🦋

slurm_submit_workers [options] <servername> <port> <num-workers>

---

💖 DESCRIPTION 💖

so here's the vibe: slurm_submit_workers schedules work_queue_worker(1) to run on the SLURM batch system, going through its job submission interface, qsub. and how many workers are we scheduling and running? that's literally what the num-workers argument is for. it's the headcount 🫶

the servername and port arguments are the hostname and the port number of the manager that the work_queue_worker is gonna connect to. iconic, straightforward.

and here's the plot twist: those two arguments become optional when the auto mode option is specified for work_queue_worker. so if you're in auto mode, you can just... not. love that for you.

---

🎀 OPTIONS 🎀

one bullet per option, no skips, we're doing all of them:

• -M — name of the preferred manager for the worker.
• -N <name> — preferred project name for work_queue_worker to connect.
• -c <cores> — set the number of cores each worker should use (0=auto). default is 1.
• -C <catalog> — set the catalog server for work_queue_worker to <catalog>. the <catalog> format is HOSTNAME:PORT.
• -t <seconds> — abort work_queue_worker after this amount of idle time. default is 900s.
• -d <subsystem> — enable debugging on the worker for this subsystem (try -d all to start).
• -w <size> — set the TCP window size.
• -i <time> — set the initial value for the backoff interval when the worker fails to connect to a manager. default is 1s.
• -b <time> — set the maximum value for the backoff interval when the worker fails to connect to a manager. default is 60s.
• -z <size> — set the available disk space threshold (in MB). when it gets exceeded, the worker will clean up and reconnect. default is 100MB.
• -A <arch> — set the architecture string for the worker to report to the manager, instead of the value from uname.
• -O <os> — set the operating system string for the worker to report to the manager, instead of the value from uname.
• -s <path> — set the location where the worker's working directory gets created.
• -p <parameters> — SLURM qsub parameters.
• -h — show the help message. always here for you 🫶

---

🔥 EXIT STATUS 🔥

• on success, returns zero. slay.
• on failure, returns non-zero. not slay, but informative.

---

🌈 EXAMPLES 🌈

exhibit one — submit 10 worker instances to run on SLURM and connect to a specific manager:

slurm_submit_workers manager.somewhere.edu 9123 10

exhibit two — submit 10 work_queue_worker instances to run on SLURM in auto mode, with their preferred project name set to Project_A and the abort timeout set to 3600 seconds:

slurm_submit_workers -a -t 3600 -M Project_A 10

---

👑 COPYRIGHT 👑

The Cooperative Computing Tools are Copyright (C) 2005-2019 The University of Notre Dame. This software is distributed under the GNU General Public License. See the file COPYING for details.

---

🧚 SEE ALSO 🧚

• Cooperative Computing Tools Documentation
• Work Queue User Manual
• work_queue_worker(1) work_queue_status(1) work_queue_factory(1) condor_submit_workers(1) sge_submit_workers(1) torque_submit_workers(1)

---

okay that's the whole page, you ate it 💖 this little number is from CCTools 8.0.0 DEVELOPMENT, and yeah, it's slurm_submit_workers(1) in case you forgot which manpage you were living in. go submit some workers bestie, ily, bye 🦋✨