-
Notifications
You must be signed in to change notification settings - Fork 14.4k
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
fix(docker): error around missing requirements/base.txt #27608
Conversation
Closes #27606 I'm unclear on why exactly CI succeeded while this was an issue running locally. I'm guessing it has to do with `--mount` and how it works around caching, so I'm moving to a COPY-based approach, which is more standard and easy to reason about. It also leaves a useful artefact on the machine as to how it was built. It does add a tiny layer and payload on the image, but it's super cheap for the utility/introspection it provides.
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'll admit I had GPT walk me through some of this, but it seems like it all ought to be fine from what I/we can tell.
I think the issue here was that |
Oh right, totally, that was the issue. Originally in my PR I was confused as to why CI didn't fail / catch this issue in the first place which I found along the way and commented on the PR body ->
|
Hi,
https://github.com/apache/superset/actions/runs/8235126931/job/22862070545?pr=26180 |
@xyy94813 rebase and you should be fine |
SUMMARY
I'm unclear on why exactly CI succeeded while this was an issue running
locally. I'm guessing it has to do with
--mount
and how it worksaround caching, so I'm moving to a COPY-based approach, which is more
standard and easy to reason about. It also leaves a useful artefact on
the machine as to how it was built. It does add a tiny layer and payload
on the image, but it's super cheap for the utility/introspection it provides.
Closes #27606
I found the issue as to why this defect in the dockerfile wasn't caught in CI, it's because the
--preset dev
parameter was not getting passed down properly, and resulting in targeting thelean
layer, avoid that issue altogether. I fixedsupersetbot
, added unit tests and published a new version on npm.TESTING/VALIDATION
tested
docker-compose up
and validated that CI is now building the right target as opposed to always be building thelean
target.