-
Notifications
You must be signed in to change notification settings - Fork 99
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
Create a dedicated docker image for ArcticDB development #2086
base: master
Are you sure you want to change the base?
Create a dedicated docker image for ArcticDB development #2086
Conversation
push: | ||
paths: | ||
- 'docker/**' | ||
- '.github/workflows/dev_docker_image.yml' |
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 don't see the other jobs depend on Docker Image for Development
job. So changes in the dev docker creation won't be reflected until the image got pushed in Docker Image for Development
job.
I feel like the checksum checking method in cibw_docker_image.yml
for triggering the creation of new docker image a better alternative
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.
We make changes to this docker very infrequently so I prefer to keep the flows a bit simpler and to not add dependencies between them unless it is needed.
Also this image will be used in other repos as well so it has to be tested manually in any case.
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.
Yes I am torn. I understand this docker image won't be changed frequently so it's hard to justify complicating already complicated wheel building process.
At the same time, if any changes made to this docker image, the CI can't tell us whether it works until the change is merged and tag latest
is updated. Only manual test evidence can be relied on. This kinda defy the purpose of CI.
I'll approve it and leave it to @poodlewars for thoughts
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.
It would be nice to be able to run the CI with a candidate new image. How about a run parameter to choose the image tag rather than hardcoding latest
? Would then be easy to try out a CI run with a candidate image.
Or having downstream jobs depend on this one sounds reasonable too. Either way works as long as it's all documented and easy for people to figure out.
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.
good point , I've added a parameter to override the default tag
docker push $image_name:$image_ver | ||
|
||
- name: Publish Docker image to GHCR as latest | ||
if: startsWith(github.ref, 'refs/tags') || github.ref == 'refs/heads/master' |
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.
Why do we push this for tags?
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.
copy-paste snafu, I've removed it
Please squash the commits when you merge |
Reference Issues/PRs
#2081
What does this implement or fix?
This PR aims to:
Successful manual run of the benchmark flow can be seen here.
It needed to be run manually because the changes to the workflow file are not picked up correctly in the automatic run here.
Any other comments?
Checklist
Checklist for code changes...