Skip to content
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

Add image substitution feature to either DSL compiler or KFP client #487

Closed
Ark-kun opened this issue Dec 6, 2018 · 7 comments
Closed

Comments

@Ark-kun
Copy link
Contributor

Ark-kun commented Dec 6, 2018

We need a way to properly perform container image substitution.
The substitution can be performed either during the pipeline compilation or before the run submission. (It can also be performed in the backend.)

run = client.run_pipeline(exp.id, 'my-tfx', 'my-tfx.tar.gz', params={}, image_substitution_map={'gcr.io/old-image': 'gcr.io/new-image'})

@gaoning777
Copy link
Contributor

The name is currently substituted during the compilation stage.

@gaoning777
Copy link
Contributor

#879

@Ark-kun
Copy link
Contributor Author

Ark-kun commented Apr 8, 2019

#879

That is a different issue. This issue is about replacing gcr.io/ml-pipeline/ with gcr.io/ml-pipeline-test/ in the whole pipeline during compilation or testing.

@Ark-kun Ark-kun reopened this Apr 8, 2019
@gaoning777
Copy link
Contributor

I understand the usefulness of the image replacement but I'm leaning towards supporting the image replacement by the user. It is simply one sed bash command. WDYT?

@elikatsis
Copy link
Member

PipelineParams may also be used in the image's place. So they can be set that way on run creation.

@gaoning777
Copy link
Contributor

JFYI, The current DSL compiler already supports substituting images with runtime PipelineParam.
I can only see the use case where the pipeline is uploaded to the KFP while users want to submit a run using that pipeline but need image substitution.
In such case, I would recommend uploading another pipeline since this is another pipeline version, which is simply another API call. WDYT?

@gaoning777
Copy link
Contributor

@Ark-kun I'm gonna close this issue now. Feel free to reopen this if you think otherwise.

Linchin pushed a commit to Linchin/pipelines that referenced this issue Apr 11, 2023
…kubeflow#487)

* See kubeflow/kfctl#58 junit artifacts aren't being uploaded for some of our
  tests because the GCS bucket is being set incorrectly.

* It looks like the problem is that our Argo workflows are setting BUILD_ID
  and not BUILD_NUMBER

* To avoid breaking workflows that might still be relying on BUILD_NUMBER
  we try BUILD_ID and fall back to BUILD_NUMBER
magdalenakuhn17 pushed a commit to magdalenakuhn17/pipelines that referenced this issue Oct 22, 2023
HumairAK pushed a commit to red-hat-data-services/data-science-pipelines that referenced this issue Mar 11, 2024
* add configuration for tekton pipeline spec

* sort annotation order
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

5 participants