gfs2_lockcapture(8)

💖 gfs2_lockcapture — the girlies' guide to snatching GFS2 lock data 💖

okay so bestie, this is a page from the System Manager's Manual, section 8, and it's all about gfs2_lockcapture. pull up a chair, grab a strawberry 🍓, and let's get into it.

✨ NAME ✨

gfs2_lockcapture — capture locking information from GFS2 file systems and DLM. 🎀

🌸 SYNOPSIS 🌸

gfs2_lockcapture [-dqyP] [-o output directory] [-r number of runs] [-s
seconds to sleep] [-n name of GFS2 file system]

gfs2_lockcapture [-dqyi]

🦋 DESCRIPTION 🦋

so gfs2_lockcapture is the tool that goes and captures the GFS2 lockdump data plus the corresponding DLM data for your GFS2 file systems. you can configure how many captures happen and how often they happen. by default it collects data from all of the mounted GFS2 file systems, unless you specifically name which GFS2 file systems you want. no cap. 💅

⚠️ heads up bestie: the sysrq -t (thread) and -m (memory) dumps and the pid directories in /proc are all collected unless you turn them off with the -P option. so like, it's a lot of data. it's giving thorough. it's giving nosy. 🔥

👑 OPTIONS 👑

• -h, --help — prints out a short usage message and then dips. exit stage left. 🫶
• -d, --debug — turns on debug logging. for when you need the tea in full detail. ☕
• -q, --quiet — disables logging to the console. shhh. 🤫
• -y, --no_ask — disables all the questions and just assumes yes. no back-and-forth, we're committing. periodt. 💅
• -i, --info — prints information about the mounted GFS2 file systems. a lil vibe check on what's mounted. 👀
• -P, --disable_process_gather — the gathering of process information will be disabled. bye process info, we hardly knew ye. 🙈
• -o <output directory>, --path_to_output_dir=<output directory> — the directory where all the collected data will be stored. this is where the receipts live, bestie. 🎀
• -r <number of runs>, --num_of_runs=<number of runs> — the number of runs capturing the lockdump data. the default is 3 runs. so three little scoops by default. 🍓
• -s <seconds to sleep>, --seconds_sleep=<seconds to sleep> — the number of seconds to sleep between runs of capturing the lockdump data. the default is 120 seconds. yes, it naps. it's self-care. 😴
• -n <name of GFS2 filesystem>, --fs_name=<name of GFS2 filesystem> — the name of the GFS2 filesystem(s) that will have their lockdump data captured. by default, all mounted GFS2 file systems get their data captured. main character energy for every single one of them. ✨

📝 NOTES 📝

the output of the following commands will be captured (aka the whole group chat is getting screenshotted):

• uname -a
• uptime
• ps h -AL -o "tid,s,cmd"
• df -h
• lsof
• mount -l
• dlm_tool ls
• dlm_tool lockdebug -v -s -w <lockspace name>
• echo "t" > /proc/sysrq-trigger (If /proc/1/stack does not exist)
• echo "m" > /proc/sysrq-trigger (If /proc/1/stack does not exist)

🧚 AUTHOR 🧚

Shane Bradley <sbradley@fedoraproject.org> — shoutout to the author, an absolute icon. 👑

🗂️ FILES 🗂️

• /proc/mounts
• /proc/slabinfo
• /sys/kernel/config/dlm/cluster/lkbtbl_size
• /sys/kernel/config/dlm/cluster/dirtbl_size
• /sys/kernel/config/dlm/cluster/rsbtbl_size
• /sys/kernel/debug/gfs2/
• /sys/kernel/debug/dlm/
• /proc/<int>/ (If /proc/1/stack does exists)
• /var/log/messages
• /var/log/cluster/

🌈 SEE ALSO 🌈

gfs2_lockcapture(8) — yeah bestie, it's pointing at itself. self-love fr fr. 💜

okay that's the whole thing, and now the cute sign-off: this was gfs2_lockcapture from the System Manager's Manual, section 8 — literally just the page name up in the corner, no extra date or version drama attached. mwah, go capture those locks responsibly. 💖✨🎀🦋🌸💅🫶