git-annex-benchmark(1)

✨ git-annex-benchmark — the benchmark girlie of the git-annex manual 💅

okay bestie, so this is the page for git-annex-benchmark(1) from the General Commands Manual, and it's literally about benchmarking git-annex commands. no cap.

---

🌸 NAME 🌸

git-annex-benchmark — benchmark git-annex commands 🔥

---

🎀 SYNOPSIS 🎀

git annex benchmark [criterionopts] ( -- commmand [; command] |
--databases=N )

(yes the commmand triple-m is exactly what the page says, i don't make the rules, i just vibe with them 🙈)

---

🦋 DESCRIPTION 🦋

so here's the tea: when git-annex is built with benchmarking support, this command lets you benchmark any other git-annex command. it's giving personal trainer for your repo. for example, git annex benchmark -- get . will benchmark git annex get. iconic.

the command you're benchmarking gets run in the current git-annex repository. and it does not run just once, bestie — the benchmarking process runs it several times so the result is actually statistically meaningful. it's giving science. ⭐

now, when you're benchmarking an action like git annex get, the first run will often do much more than subsequent runs. 😭 to make the benchmark repeat an action like getting a file each time, you can list additional commands separated by ;. heads up though: ; needs to be escaped from the shell. the combined script gets run repeatedly by the benchmark. example time:

git annex benchmark -- get . ';' drop .

also super important: git-annex benchmark does not fork new git-annex processes when benchmarking. it calls the command to benchmark internally, so it skips git-annex's startup overhead. so like… don't try to use it to optimise git-annex startup. that's not the vibe here. 💀

---

💖 OPTIONS 💖

• before the --, you can use any of the criterion library's command-line options. slay, borrowed greatness. ✨
• any options that git-annex usually accepts can be included after the command to benchmark. so you're not losing anything, bestie. 🫶
• --databases=N — this one benchmarks git-annex's use of sqlite databases instead of a command. N is the number of items to benchmark. database girlies rise. 🌈

---

👑 OUTPUT 👑

the output of the commands being benchmarked goes to standard output and standard error as usual. honestly it's often a good idea to use --quiet to avoid unnecessary output — unless, of course, the generation of that output is part of what you want to benchmark. iykyk. 🧚

the benchmark report itself is output to standard output by default, though criterion options can be used to redirect it to a file. for example:

git annex benchmark -o bench -- find >/dev/null

periodt. 💅

---

🌟 SEE ALSO 🌟

• git-annex(1)

---

💜 AUTHOR 💜

Joey Hess <id@joeyh.name> — thank you for your service king 🫡✨

---

and that's the whole page, bestie!! this was git-annex-benchmark(1) from the General Commands Manual — the footer's just the page name itself, no extra version or date tea to spill, so we're keeping it cute and simple. 🎀💖🌸