-
Notifications
You must be signed in to change notification settings - Fork 14
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
Using base image with python version 3.9.13 #318
Conversation
I don't think I am qualified to review this, but in any case I am quite confused by this sentence. |
Sure you are, this is an open source project, and you have exceeded qualification to review this from my perspective. :)
My bad. I mean [1] https://hub.docker.com/r/jupyter/minimal-notebook/tags?page=1&name=3.9.4 |
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.
Got it, thanks. Got confused since you're removing the linux/arm64
support while at the same time bumping to the image that supports it. :-)
Sure you are, this is an open source project, and you have exceeded qualification to review this from my perspective. :)
😊
Leaving approve for others to weigh in whether support should be removed here.
I'd say it has to be removed otherwise the build will fail in any case. The original support is a "fake" support, the docker buildx will fall to use the one with wrong architecture and set the image digest to wrong image. |
I am also a bit confused here. Why does switching to the base image that supports the |
If the image |
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.
When there is the image support arm64, it will use the correct one but our stack doesn't support arm64 yet such as the rabbitmq don't have conda forge for arm64.
So if I understand correctly, arm64 support will be added back in some future PR and besides upgrading the base image (done here) is blocked on other dependencies (RabbitMQ?).
Approving under the assumption that my understanding is correct. :-)
Thanks, @danielhollas. Yes, this is what my plan is. But for rabbitmq, I don't think it will have conda forge package supported for arm64 soon, even if it has, it may not support |
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.
LGTM, thanks @unkcpz!
This was an omission in #318. The Python version specified in build.json is used to automatically generate the image tags on Dockerhub. So currently, for example, the image aiidalab/full-stack:python-3.9.4 actually corresponds to python-3.9.13. The default BASE in `stack/base/Dockerfile` is set in docker-bake.hcl for building the image. Replace the `ARG BASE` with using the arg from the environment should be fine and good to have only one place to set the version.
As mentioned #316 (comment),
The new base jupyter/minimal-notebook start to support linux/arm64/v8 which is not for the old base image. This is also the wired reason that the arm64 build pass and what is being downloaded in my arm64 machine is actually amd64.
I propose to remove the support to arm64 and add it back after I also start and test by using the self-hosted CI runner.