Not sure the best channel for this Just tried uv ...
# best-practices
v
Not sure the best channel for this Just tried uv for one of our projects. This one the normal clean install time is ~45 seconds. With uv and a non-warm cache it goes down to ~15 seconds. With uv and a warm cache
meltano install --clean
is 3 seconds. Holy crap we are switching
👀 3
dancingpenguin 1
Thanks @Edgar Ramírez (Arch.dev) 😄
3
💯 2
m
yeah, there’s no going back 😂
😄 1
r
@aaron_phethean it might be time 🫡
1
💪 1
v
@Reuben (Matatika) steps for us locally pipx uninstall meltano pipx install meltano[uv] meltano config meltano set venv.backend uv meltano install --clean Ci was similar
🔥 1
Windows docker container build time went down 30 percent. Running directly on Windows the build time improved pretty well but we didn't see the crazy gain from a warm cache. Haven't dove into why
I tested Windows, and it definitely gets the same increase with a warm cache. My problem is related to Docker builds and caching, which don't have anything to do with this. I think I'll see a massive improvement, though, if I make this work!
https://github.com/astral-sh/uv/issues/2810 Windows parallel installs fail fairly consistently due to this one. Swapping to
meltano install -p 1
works
👍 1
e
I've put up a tracker for common uv issues: https://github.com/meltano/meltano/issues/8539