New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Switch to Mamba #697
Switch to Mamba #697
Conversation
This comment has been minimized.
This comment has been minimized.
I'd love to see us use mambaforge instead of miniconda, big upvote on this PR! I'd be quick to review any additional changes to docs etc. I'd love to have someone deliberate on if this is a breaking change or not. I'm not sure myself, but I note that we have tests that first install the current version of things via the bootstrap script currently on GitHub in the repo's default branch, and that we then run it again etc.
|
`t2.micro` works but recommend `t3.small`
tljh/installer.py
Outdated
installer_sha256 = "a012c24e1cc3bcbe74a1e5693e510830e7c2956e85877b08d1e28707a0bd8d75" | ||
mamba_version = '0.15.2' | ||
|
||
if conda.check_miniconda_version(USER_ENV_PREFIX, mambaconda_new_version): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Do you understand the idea about this? It would be great to have a comment around here regarding these checks.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I did figure it out eventually but I can't remember the details now! It's something along the lines of check the conda version, from which you can infer the miniconda version that was used, and from that you can infer how old the TLJH installation is.
Edit: Or maybe it's the other way round, figure out which version of miniconda is installed, then choose which version of conda we want based on that (maybe because older miniconda versions installed an older version of Python which isn't suportted by newer conda?)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks! I consider this non-blocking for a merge. I've heard about feature creep, but this is more like a maintenance creep of me to ask about :D
Co-authored-by: Erik Sundell <erik.i.sundell@gmail.com>
I've rebased this PR, and I think I've responded to everything apart from #697 (comment) |
@GeorgianaElena @yuvipanda what do you think about this? |
Lets' do this! |
I think the upgrade logic we have means this isn't a 'breaking' change as such :) |
❤️ 🎉 WIEEEEEEEEEEEEEEE!!! |
[ ] Add testsCan't remember why I added this item since the tests were already in place and just needed updating for mambaforgeSwitches from conda to mamba. The main advantage is this should require less memory (I've reduced the integration test Docker limit to 900 MB), hopefully this is enough to get TLJH running on AWS t2.micro instances with 1GB RAM.
I haven't updated any docs yet, I'd like some feedback first.