CONDOR_GATHER_INFO(1)

📦
version 25.14.1 · 12 versions
, in fedora 43

💖 condor_gather_info — the HTCondor Manual, but make it a group chat 💖

okay bestie, pull up a chair because we're talking about condor_gather_info, a section 1 (user commands) manual page from the HTCondor Manual. she's giving diagnostic detective, she's giving collect everything, and honestly? she ate. 🕵️♀️✨

✨ NAME ✨

condor_gather_info — from the HTCondor Manual. It gathers information about an HTCondor installation and about a queued job. 🌸

🎀 SYNOPSIS 🎀

condor_gather_info [--jobid ClusterId.ProcId] [--scratch /path/to/directory]

that's it. that's the whole vibe. two optional flags and a dream. 💅

🦋 DESCRIPTION 🦋

so condor_gather_info is a Linux-only tool (yes, Linux only, don't come for her) and what she does is collect and output info about three things:

  1. the machine she's being run on,
  2. the HTCondor installation that's local to that machine,
  3. and optionally, if you asked nicely, a specific HTCondor job.

and the tea? the info she gathers is most often used as a debugging aid for the developers of HTCondor. so like, you're basically handing the devs a whole scrapbook of your machine's lore. iconic behavior. 📸

no --jobid? we do a little identity moment

if you run her without the --jobid option, she gathers info about the local machine and its HTCondor installation and drops it into a file called condor-profile.txt, in the current working directory. and that info lives under the category called Identity. 🪪

with --jobid? we go full main character

if you do pass --jobid, she gathers additional info about the job you gave her on the command line — the one identified by its ClusterId and ProcId ClassAd attributes. now you get both categories: Identity and Job information. 💖

and because the quantity of information can be extensive (read: a LOT, like a concerning amount), she puts it all into a compressed tar file. it lands in the current working directory, and it's named like this:

cgi-<username>-jid<ClusterId>.<ProcId>-<year>-<month>-<day>-<hour>_<minute>_<second>-<TZ>.tar.gz

every single value inside those <> gets substituted with the current values. no cap.

⚠️ heads up bestie: building this potentially large tar file can require a fair amount of temporary space. like, she needs room to cook. 🍳

so if you specify the --scratch option, that identifies a directory where she'll build the tar file. and if you don't specify --scratch, then the directory will be /tmp/cgi-<PID>, where that process ID is the one belonging to the condor_gather_info executable. 🫶

the info she gathers, in full, no skips 🧚

1. Identity 🌈

  • User name who generated the report
  • Script location and machine name
  • Date of report creation
  • uname -a
  • Contents of /etc/issue
  • Contents of /etc/redhat-release
  • Contents of /etc/debian_version
  • Contents of $(LOG)/MasterLog
  • Contents of $(LOG)/ShadowLog
  • Contents of $(LOG)/SchedLog
  • Output of ps -auxww -forest
  • Output of df -h
  • Output of iptables -L
  • Output of ls 'condor_config_val LOG'
  • Output of ldd 'condor_config_val SBIN'/condor_schedd
  • Contents of /etc/hosts
  • Contents of /etc/nsswitch.conf
  • Output of ulimit -a
  • Output of uptime
  • Output of free
  • Network interface configuration (ifconfig)
  • HTCondor version
  • Location of HTCondor configuration files
  • HTCondor configuration variables:
    • All variables and values
    • Definition locations for each configuration variable

2. Job Information 👑

  • Output of condor_q jobid
  • Output of condor_q -l jobid
  • Output of condor_q -analyze jobid
  • Job event log, if it exists:
    • Only events pertaining to the job ID
  • If condor_gather_info has the proper permissions, it runs condor_fetchlog on the machine where the job most recently ran, and includes the contents of the logs from the condor_master, condor_startd, and condor_starter.

💅 OPTIONS 💅

  • -jobid <ClusterId.ProcId> — data mine information about this HTCondor job from the local HTCondor installation and condor_schedd. give her the job receipt basically. 🧾
  • -scratch /path/to/directory — a path to temporary space needed when building the output tar file. defaults to /tmp/cgi-<PID>, where <PID> is replaced by the process ID of condor_gather_info. so if you don't pick a spot, she picks one for you. 🌸

🍓 FILES 🍓

  • condor-profile.txt — the Identity portion of the information gathered when condor_gather_info is run without arguments. the no-flags special. ✨
  • cgi-<username>-jid<cluster>.<proc>-<year>-<month>-<day>-<hour>_<minute>_<second>-<TZ>.tar.gz — the output file which contains all of the information produced by this tool. the whole scrapbook. 📦

🔥 EXIT STATUS 🔥

condor_gather_info will exit with a status value of 0 (zero) upon success, and it will exit with the value 1 (one) upon failure.

zero = slay. one = something went 💀. simple as that.

🫶 AUTHOR 🫶

HTCondor Team — the collective. the group chat admins. the ones who made her. 💜

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, we're done, no more notes 🎀 this was the CONDOR_GATHER_INFO(1) page from the HTCondor Manual — dated Oct 16, 2026, periodt. go gather some info, drink water, and remember: when in doubt, --scratch somewhere with space. iykyk 💖✨

📦 where this page comes from