CONDOR_WAIT(1)

📦
version 25.14.1 · 12 versions
, in fedora 43

✨ condor_wait — the HTCondor Manual, bestie edition ✨

okay so real talk, this is condor_wait from the HTCondor Manual, and it is literally the "are we there yet" of the HTCondor universe. she waits for your jobs to finish. that's the whole gig. and she's so good at it 💖

🌸 NAME 🌸

condor_wait — HTCondor Manual

wait for HTCondor jobs to finish. no cap, that's it, that's the function.

🎀 SYNOPSIS 🎀

condor_wait [-help | -version]

condor_wait [-debug] [-status] [-echo] [-wait seconds] [-num
number-of-jobs] log-file [job-ID]

🦋 DESCRIPTION 🦋

okay so condor_wait is that friend who just sits there and watches. specifically, she watches a job event log file — the one you made with the log command inside a submit description file — and she comes back to you the moment one or more jobs from that log have completed or aborted. peak loyalty behavior fr fr 🫶

here's the important plot twist though: condor_wait expects to find at least one job submitted event in that log file. so at least one job has to have been successfully submitted with condor_submit before you run condor_wait. if nothing was ever submitted, she's got nothing to watch, and that's simply not the vibe.

and with no other instructions? condor_wait will wait forever for jobs to finish. forever forever. unless you specify a shorter wait time. so maybe set that timeout, bestie 😭

✨ OPTIONS ✨

  • -help — show the usage info. just the friendly little "here's how I work" moment.
  • -version — show version info. the girlie's ID card 💅
  • -debug — show extra debugging info. she'll spill all the details about what's going on.
  • -status — show job start and terminate info. so you can see when things kicked off and when they dipped.
  • -echo — print the events out to stdout. narrating the whole saga live, iconic.
  • -wait seconds — wait no more than that integer number of seconds. the default is unlimited time (see above: forever, bestie, literally forever).
  • -num number-of-jobs — wait for that integer number-of-jobs number of jobs to end. the default is all jobs in the log file. so if you only care about a few, this is the move.
  • -allevents — keep reading the log even after all jobs have ended. most useful when you pair it with -echo, so you can transcribe an entire job event log to standard output. oh and heads up: this one may not be combined with -num. they don't share the aux, it's giving enemies 😤
  • log-file — the name of the log file to watch for information about the job. this is your main source of tea.
  • job-ID — a specific job, or a whole set of jobs, to watch. here's the breakdown:
    • if the job-ID is only the job ClassAd attribute ClusterId, then condor_wait waits for all jobs with that given ClusterId.
    • if the job-ID is a pair of the job ClassAd attributes, written as ClusterId.ProcId, then condor_wait waits for the specific job with that job-ID.
    • if you don't specify this at all, then every job that exists in the log file at the moment condor_wait is invoked will be watched. they're all getting her attention, periodt.

🌈 GENERAL REMARKS 🌈

condor_wait is an inexpensive way to test or wait for the completion of a job or a whole cluster — perfect if you're trying to get a process outside of HTCondor to sync up with a job or a set of jobs. very "I'll hold your spot" energy 💜

she can also be used to wait for the completion of a limited subset of jobs, via the -num option. so you don't have to commit to the whole cluster if you don't want to. boundaries, bestie, we love them.

🍓 EXAMPLES 🍓

wait for all jobs in a log file to complete:

$ condor_wait logfile

wait for all jobs in cluster 40 to complete:

$ condor_wait logfile 40

wait for any two jobs to complete:

$ condor_wait -num 2 logfile

wait for a specific job to complete:

$ condor_wait logfile 40.1

wait for job 40.1 with a one hour timeout:

$ condor_wait -wait 3600 logfile 40.1

💀 EXIT STATUS 💀

  • 0 — success! the specified job or jobs have completed or aborted. slay.
  • 1 — failure. this is the unrecoverable-errors zone, like a missing log file, a non-existent job, or a timeout that expired. she tried her best 😭

👑 SEE ALSO 👑

condor_submit, condor_q, condor_history

go read those too, iykyk.

🧚 AVAILABILITY 🧚

Linux, MacOS, Windows — she's everywhere, global icon behavior. 🌍

💖 AUTHOR 💖

HTCondor Team

1990-2026, Center for High Throughput Computing, Computer Sciences Department, University of Wisconsin-Madison, Madison, WI, US. Licensed under the Apache License, Version 2.0.

and that's the whole page, bestie 🎀 this was CONDOR_WAIT(1) from the HTCondor Manual, dated Oct 16, 2026 — go wait for your jobs in style, you've got this 💅✨

📦 where this page comes from