Kubeflow is a machine learning (ML) toolkit that is dedicated to making deployments of ML workflows on Kubernetes simple, portable, and scalable.

What steps did you take:

In a multi-user enabled env, I created a notebook server on user's namespace, launch a notebook and try to call Python SDK from there. When I execute the code below:

pipeline = kfp.Client().create_run_from_pipeline_func(mnist_pipeline, arguments={}, namespace='mynamespace')

What happened:

The API call was rejected with the following errors:

~/.local/lib/python3.6/site-packages/kfp_server_api/rest.py in request(self, method, url, query_params, headers, body, post_params, _preload_content, _request_timeout)
    236 
    237         if not 200 <= r.status <= 299:
--> 238             raise ApiException(http_resp=r)
    239 
    240         return r

ApiException: (403)
Reason: Forbidden
HTTP response headers: HTTPHeaderDict({'content-length': '19', 'content-type': 'text/plain', 'date': 'Tue, 01 Sep 2020 00:58:39 GMT', 'server': 'envoy', 'x-envoy-upstream-service-time': '8'})
HTTP response body: RBAC: access denied

What did you expect to happen:

A pipeline run should be created and executed

Environment:

How did you deploy Kubeflow Pipelines (KFP)?

I installed the KFP on IKS with multi-user support KFP version: v1.1.0 KFP SDK version: v1.0.0

Anything else you would like to add:

[Miscellaneous information that will assist in solving the issue.]

/kind bug

Description of your changes:

Checklist:

• [ ] The title for your pull request (PR) should follow our title convention. Learn more about the pull request title convention used in this repository.

• [ ] Do you want this pull request (PR) cherry-picked into the current release branch?

  Learn more about cherry-picking updates into the release branch.

Description of your changes: This PR adds parallelism limits for sub-DAG:s. This is a continuation (https://github.com/kubeflow/pipelines/pull/4149) which relate to issue. Checklist:

• [ ] The title for your pull request (PR) should follow our title convention. Learn more about the pull request title convention used in this repository.

  PR titles examples:

  • fix(frontend): fixes empty page. Fixes #1234 Use fix to indicate that this PR fixes a bug.
  • feat(backend): configurable service account. Fixes #1234, fixes #1235 Use feat to indicate that this PR adds a new feature.
  • chore: set up changelog generation tools Use chore to indicate that this PR makes some changes that users don't need to know.
  • test: fix CI failure. Part of #1234 Use part of to indicate that a PR is working on an issue, but shouldn't close the issue when merged.

• [ ] Do you want this pull request (PR) cherry-picked into the current release branch?

  If yes, use one of the following options:

  • (Recommended.) Ask the PR approver to add the cherrypick-approved label to this PR. The release manager adds this PR to the release branch in a batch update.
  • After this PR is merged, create a cherry-pick PR to add these changes to the release branch. (For more information about creating a cherry-pick PR, see the Kubeflow Pipelines release guide.)

[April/6/2020] Latest design is in https://docs.google.com/document/d/1R9bj1uI0As6umCTZ2mv_6_tjgFshIKxkSt00QLYjNV4/edit?ts=5e4d8fbb#heading=h.5s8rbufek1ax

Areas we are working on:

• [x] [Frontend] Deploy ui artifact service for each namespace https://github.com/kubeflow/pipelines/issues/3554
• [x] [Frontend/Backend] Deploy visualization service for each namespace https://github.com/kubeflow/pipelines/issues/2899
• [x] [Backend] Use experiment for resource boundary for child resource CRUD https://github.com/kubeflow/pipelines/issues/2397
  • [x] Experiment https://github.com/kubeflow/pipelines/issues/3273
  • [x] Run https://github.com/kubeflow/pipelines/issues/3336
  • [x] Job https://github.com/kubeflow/pipelines/issues/3344
• [x] [Frontend/SDK/Backend] Skip specify namespace for CreateRun APIs https://github.com/kubeflow/pipelines/issues/3290
• [x] [Deployment] Enable MLMD functionality in multi-user mode https://github.com/kubeflow/pipelines/issues/3292
• [x] [Frontend] Block non public api from frontend (e.g. report api) in multi-user mode https://github.com/kubeflow/pipelines/issues/3293
• [x] [Frontend/Controller] Launch Tensorboard in user's namespace https://github.com/kubeflow/pipelines/issues/3294
• [x] [Frontend] Pass namespace as a parameter for experiment API https://github.com/kubeflow/pipelines/issues/3291
• [x] [Frontend] Pass namespace as a parameter for run API https://github.com/kubeflow/pipelines/pull/3351
• [x] [Frontend] UI should react when user changes namespace https://github.com/kubeflow/pipelines/issues/3296
• [x] [SDK] Pass namespace as a parameter for experiment APIs https://github.com/kubeflow/pipelines/pull/3272
• [x] [Deployment] KFP profile controller that configures KFP required resources in each user's namespaces https://github.com/kubeflow/pipelines/issues/3420
• [ ] [Test] Postsubmit test for multi user e2e scenario https://github.com/kubeflow/pipelines/issues/3288
• [ ] [Test] Backend integration tests for multi-user scenarios https://github.com/kubeflow/pipelines/issues/3289
• [ ] [Test] Network auth integration tests https://github.com/kubeflow/pipelines/issues/3646
• [x] [Deployment] Make user identity header configurable #3752
• [x] [Doc] documentation on kubeflow.org #4317

Release

• [x] How do we release KFP multi user mode? https://github.com/kubeflow/pipelines/issues/3645
• [x] Multi user mode early access release #3693
• [x] [Deployment] Merge changes to upstream kubeflow repo https://github.com/kubeflow/pipelines/issues/3241
• [x] Integrate with platforms other than GCP https://github.com/kubeflow/manifests/issues/1364

Areas related to integration with Kubeflow

• [ ] [Central Dashboard] Manage contributors for all namespaces I own https://github.com/kubeflow/kubeflow/issues/4569
• [x] [Central Dashboard] Support login to Kubeflow cluster without creating his/her namespace for a non-admin contributor https://github.com/kubeflow/kubeflow/issues/4889
• [ ] [Profile CRD] Support more than one owner of a profile CR https://github.com/kubeflow/kubeflow/issues/4888
• [ ] [Profile CRD] Support updating the owner of a profile https://github.com/kubeflow/kubeflow/issues/4890

=============== original description

Some users express the interest of an isolation between the cluster admin and cluster user - Cluster admin deploy Kubeflow Pipelines as part of Kubeflow in the cluster; Cluster user can use Kubeflow Pipelines functionalities, without being able to access the control plane.

Here are the steps to support this functionality.

1. Provision control plane in one namespace, and launch argo workflow instances in another
   • provision control plane in kubeflow namespace, and argo job in namespace FOO (parameterization)
   • API server should update the incoming workflow definition to namespace FOO. Sample code that API server modify the workflow
2. Currently all workflows are run under a clusterrole pipeline-runner (definition). And it's specified during compilation (link). Instead, it should run the workflows under a role instead of a clusterrole.
   • change pipeline-runner to role, and specify the namespace during deployment (expose as deployment parameter)
   • API server should update the incoming workflow definition to use pipeline-runner role.
3. Cluster user can access UI through IAP/SimpleAuth endpoint, instead of port-forwarding.

Do you think it would be possible to support non-docker based clusters as well? I'm currently checking out the examples and see that they want to mount the docker.sock into the container. We might achieve the same results when using crictl. WDYT?

Added namespaced pipelines, with UI and API changes, as well as the ability to share pipelines.

Fixes: https://github.com/kubeflow/pipelines/issues/4197

Description of your changes:

• Added a new field in Pipelines table for namespace.
• Uploaded Pipelines are by default namespaced.
• Ability to share Pipelines by selecting "shared" check-mark in the UI.
• Authorization via SubjectAccessReview for Pipelines, PipelinesVersions, and Upload Pipelines endpoints.

Authors: @arllanos @maganaluis

Welcome to Renovate! This is an onboarding PR to help you understand and configure settings before regular Pull Requests begin.

:vertical_traffic_light: To activate Renovate, merge this Pull Request. To disable Renovate, simply close this Pull Request unmerged.

Detected Package Files

• WORKSPACE (bazel)
• backend/Dockerfile (dockerfile)
• backend/Dockerfile.bazel (dockerfile)
• backend/Dockerfile.cacheserver (dockerfile)
• backend/Dockerfile.persistenceagent (dockerfile)
• backend/Dockerfile.scheduledworkflow (dockerfile)
• backend/Dockerfile.viewercontroller (dockerfile)
• backend/Dockerfile.visualization (dockerfile)
• backend/metadata_writer/Dockerfile (dockerfile)
• backend/src/cache/deployer/Dockerfile (dockerfile)
• components/gcp/container/Dockerfile (dockerfile)
• components/kubeflow/deployer/Dockerfile (dockerfile)
• components/kubeflow/dnntrainer/Dockerfile (dockerfile)
• components/kubeflow/katib-launcher/Dockerfile (dockerfile)
• components/kubeflow/kfserving/Dockerfile (dockerfile)
• components/kubeflow/launcher/Dockerfile (dockerfile)
• components/local/base/Dockerfile (dockerfile)
• components/local/confusion_matrix/Dockerfile (dockerfile)
• components/local/roc/Dockerfile (dockerfile)
• components/sample/keras/train_classifier/Dockerfile (dockerfile)
• contrib/components/openvino/model_convert/containers/Dockerfile (dockerfile)
• contrib/components/openvino/ovms-deployer/containers/Dockerfile (dockerfile)
• contrib/components/openvino/predict/containers/Dockerfile (dockerfile)
• contrib/components/openvino/tf-slim/containers/Dockerfile (dockerfile)
• frontend/Dockerfile (dockerfile)
• manifests/gcp_marketplace/deployer/Dockerfile (dockerfile)
• proxy/Dockerfile (dockerfile)
• samples/contrib/image-captioning-gcp/src/Dockerfile (dockerfile)
• samples/contrib/nvidia-resnet/components/inference_server_launcher/Dockerfile (dockerfile)
• samples/contrib/nvidia-resnet/components/preprocess/Dockerfile (dockerfile)
• samples/contrib/nvidia-resnet/components/train/Dockerfile (dockerfile)
• samples/contrib/nvidia-resnet/components/webapp/Dockerfile (dockerfile)
• samples/contrib/nvidia-resnet/components/webapp_launcher/Dockerfile (dockerfile)
• samples/contrib/nvidia-resnet/pipeline/Dockerfile (dockerfile)
• samples/contrib/versioned-pipeline-ci-samples/helloworld-ci-sample/helloworld/Dockerfile (dockerfile)
• samples/contrib/versioned-pipeline-ci-samples/kaggle-ci-sample/download_dataset/Dockerfile (dockerfile)
• samples/contrib/versioned-pipeline-ci-samples/kaggle-ci-sample/submit_result/Dockerfile (dockerfile)
• samples/contrib/versioned-pipeline-ci-samples/kaggle-ci-sample/train_model/Dockerfile (dockerfile)
• samples/contrib/versioned-pipeline-ci-samples/kaggle-ci-sample/visualize_html/Dockerfile (dockerfile)
• samples/contrib/versioned-pipeline-ci-samples/kaggle-ci-sample/visualize_table/Dockerfile (dockerfile)
• samples/contrib/versioned-pipeline-ci-samples/mnist-ci-sample/tensorboard/Dockerfile (dockerfile)
• samples/contrib/versioned-pipeline-ci-samples/mnist-ci-sample/train/Dockerfile (dockerfile)
• test/api-integration-test/Dockerfile (dockerfile)
• test/frontend-integration-test/Dockerfile (dockerfile)
• test/frontend-integration-test/selenium-standalone-chrome-gcloud-nodejs.Docker/Dockerfile (dockerfile)
• test/imagebuilder/Dockerfile (dockerfile)
• test/images/Dockerfile (dockerfile)
• test/initialization-test/Dockerfile (dockerfile)
• test/sample-test/Dockerfile (dockerfile)
• tools/bazel_builder/Dockerfile (dockerfile)
• go.mod (gomod)
• frontend/mock-backend/package.json (npm)
• frontend/package.json (npm)
• frontend/server/package.json (npm)
• package.json (npm)
• test/frontend-integration-test/package.json (npm)
• frontend/.nvmrc (nvm)
• backend/metadata_writer/requirements.txt (pip_requirements)
• backend/requirements.txt (pip_requirements)
• backend/src/apiserver/visualization/requirements.txt (pip_requirements)
• components/kubeflow/katib-launcher/requirements.txt (pip_requirements)
• contrib/components/openvino/ovms-deployer/containers/requirements.txt (pip_requirements)
• docs/requirements.txt (pip_requirements)
• samples/contrib/azure-samples/databricks-pipelines/requirements.txt (pip_requirements)
• samples/contrib/ibm-samples/ffdl-seldon/source/seldon-pytorch-serving-image/requirements.txt (pip_requirements)
• samples/core/ai_platform/training/requirements.txt (pip_requirements)
• samples/core/container_build/requirements.txt (pip_requirements)
• sdk/python/requirements.txt (pip_requirements)
• test/kfp-functional-test/requirements.txt (pip_requirements)
• test/sample-test/requirements.txt (pip_requirements)
• components/gcp/container/component_sdk/python/setup.py (pip_setup)
• components/kubeflow/dnntrainer/src/setup.py (pip_setup)
• samples/core/ai_platform/training/setup.py (pip_setup)
• sdk/python/setup.py (pip_setup)

Configuration

:abcd: Renovate has detected a custom config for this PR. Feel free to ask for help if you have any doubts and would like it reviewed.

Important: Now that this branch is edited, Renovate can't rebase it from the base branch any more. If you make changes to the base branch that could impact this onboarding PR, please merge them manually.

What to Expect

With your current configuration, Renovate will create 57 Pull Requests:

:children_crossing: Branch creation will be limited to maximum 2 per hour, so it doesn't swamp any CI resources or spam the project. See docs for prhourlylimit for details.

:question: Got questions? Check out Renovate's Docs, particularly the Getting Started section. If you need any further assistance then you can also request help here.

This PR has been generated by WhiteSource Renovate. View repository job log here.

/kind bug

What steps did you take and what happened: Enabled authentication with Azure AD on AKS and installing Kubeflow with kfctl_istio_dex.v1.1.0.yaml but skipping the dex from the manifest as Azure AD is an OIDC provider. The load balancer is exposed over https with TLS 1.3 self-signed cert.

OIDC Auth Service Configuration:

• client_id=XXXX
• oidc_provider=https://login.microsoftonline.com/XXXX/v2.0
• oidc_redirect_uri=https://XXXX/login/oidc
• oidc_auth_url=https://login.microsoftonline.com/XXXX/oauth2/v2.0/authorize
• application_secret=XXXX
• skip_auth_uri=
• namespace=istio-system
• userid-header=kubeflow-userid
• userid-prefix=

Issue When using KFP client to upload the pipeline (client.pipeline_uploads.upload_pipeline()) with below client config throws an error.

client = kfp.Client(host='https://<LoadBalancer IP Address>/pipeline', existing_token=<token>)

Error HTTPSConnectionPool(host='<Host_IP>', port=443): Max retries exceeded with url: /pipeline/apis/v1beta1/pipelines/upload?name=local_exp-6714175b-6d59-40d0-9019-5b4ee58dc483 (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: self signed certificate (_ssl.c:1076)')))

Is there a way to override cert verification?

or

When using KFP client to upload the pipeline (client.pipeline_uploads.upload_pipeline()) with below client config redirects to google auth error.

client = kfp.Client(host='https://<LoadBalancer IP Address>/pipeline ,client_id=<client_id>, other_client_id=<client_id>,other_client_secret=<application_secret>,namespace='kfauth')

Environment:

• Kubeflow version: v1.1.O
• kfctl version: kfctl_v1.1.0-0-g9a3621e_linux.tar.gz
• kfp version: 1.0.1
• python version: 3.6.8
• kfp-server-api version: 1.0.1
• Kubernetes platform: Azure Kubernetes Service
• Kubernetes version: 1.17.11

CC: @Bobgy

Description of your changes: Remove Bazel from the api generations, this is part of https://github.com/kubeflow/pipelines/issues/3250. The suggested solution is based upon work from: https://github.com/kubeflow/pipelines/pull/4393.

The suggested solution uses docker to make it simpler for users to not have to install all the necessary tools and environments locally, but not sure it is the best solution. Post this as an early draft in order to discuss possible solutions.

Checklist:

• [x] The title for your pull request (PR) should follow our title convention. Learn more about the pull request title convention used in this repository.

• [ ] Do you want this pull request (PR) cherry-picked into the current release branch?

  Learn more about cherry-picking updates into the release branch.

We are trying to finalize Kubeflow 0.7 by end of month.

Which version of KFP should be shipped in 0.7.

We are currently shipping KFP 0.1.23. It looks like this is about 1 month old. It looks like there was a fairly recent release 0.1.31 https://github.com/kubeflow/pipelines/releases

@IronPan @jessiezcc Should we ship 0.1.31 in 0.7? Are there additional improvements that we would like to ship in 0.7? If so do we have an ETA for when they will land?

Motivation

I am running a kubeflow pipeline deployment with my custom helm chart and a minio s3 gateway to my custom bucket. This bucket has a different name from the default one in kfp, hence I need some way to parameterize the s3 artifact configs.

Status

• Waiting for Review

Features

• kfp can now declare a custom artifact location inside a pipeline or containerop.

from kfp import dsl
from kubernetes.client.models import V1SecretKeySelector


@dsl.pipeline( name='foo', description='hello world')
def foo_pipeline(namespace: str):

    # configures artifact location
    artifact_location = dsl.ArtifactLocation.s3(
                            bucket="foobar",
                            endpoint="minio-service.%s:9000" % namespace,  # parameterized namespace
                            insecure=True,
                            access_key_secret=V1SecretKeySelector(name="minio", key="accesskey"),
                            secret_key_secret={"name": "minio", "key": "secretkey"}  # accepts dict also
    )

    # set pipeline level artifact location
    conf = dsl.get_pipeline_conf().set_artifact_location(artifact_location)
    
    # use pipeline level artifact location (i.e. minio-service)
    op1 = dsl.ContainerOp(name='foo', image='bash:latest')

    # use containerop level artifact location (i.e. aws)
    op2 = dsl.ContainerOp(
                        name='foo', 
                        image='bash:latest',
                        # configures artifact location
                        artifact_location=dsl.ArtifactLocation.s3(
                            bucket="foobar",
                            endpoint="s3.amazonaws.com",
                            insecure=False,
                            access_key_secret=V1SecretKeySelector(name="s3-secret", key="accesskey"),
                            secret_key_secret=V1SecretKeySelector(name="s3-secret", key="secretkey"))
    )

TLDR changes

• argo-models is now a dependency in setup.py (argo v2.2.1)
• Added static class ArtifactLocation
  • to help generate artifact location for s3
  • to help generate artifact for workflow templates
• Updated PipelineConf to support artifact location
• Updated k8s helper and related, to support openapi objects (I accidentally used openapi generator instead of swagger codegen for argo-models)
• Added unit test for ArtifactLocation
• Fixed unit test for kfp.aws (Found that it has a bug, and was not imported into the unit test)

This change is 

Bumps json5 from 2.1.1 to 2.2.3.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Feature Area

/area sdk

What feature would you like to see?

Protocol buffer filtering for the finished_at field in runs.

What is the use case or pain point?

We are looking to archive runs that finished over 90 days ago. While we have a workaround (see below), if we have a filter (like you do currently for the created_at field), it would mean that our request payloads are substantially smaller.

Is there a workaround currently?

Our current approach is to get all non-archived runs (using a predicate filter), then iterate over the result to compare the datetime of the finished_at field to the datetime from 90 days ago. We then archive those runs that completed 90 days ago.

Love this idea? Give it a 👍.

Issue The create_endpoint, delete_endpoint and deploy_model aiplatform components currently use the google_cloud_pipeline_components.container.v1.* scripts, which fail with the error below (example for create_endpoint)

Error while finding module specification for 'google_cloud_pipeline_components.container.v1.endpoint.create_endpoint.launcher' (ModuleNotFoundError: No module named 'google_cloud_pipeline_components.container.v1')

Description of your changes
This PR changes the entrypoint of the create_endpoint, delete_endpoint and deploy_model components from python3 -u -m google_cloud_pipeline_components.container.v1.endpoint.create_endpoint.launcher to python3 -u -m google_cloud_pipeline_components.container.aiplatform.remote_runner --cls_name Endpoint --method_name create

Adapted component inputs to comply with the imput requested by remote_runner:

• Removed outdated parameters, as they are not supported by new versions of the Vertex AI Python SDK and can break execution.

• Added new optional method parameters supported by the Vertex AI Python SDK.

The deploy_model component now uses Model.deploy instead of Endpoint.deploy, thus the deploy_model folder has been moved from the endpoint folder to the model folder.

Modified all relevant tests to comply with the new component definition.

Checklist:

• [ ] The title for your pull request (PR) should follow our title convention. Learn more about the pull request title convention used in this repository.

1. Enable cloning a recurring run with same runtimeConfig and run-trigger from the original recurring run.
2. Enable user change runtimeConfig and run-trigger from cloned recurring run.

Description of your changes:

#8502 details most of the changes and reasoning in a design document.

Removing most of the aws-helper code that pertains to grabbing the credentials from the ec2 instance metadata as this code is already handled as part of the credentialProviderChain that is being imported here.

Checklist:

• [ ] The title for your pull request (PR) should follow our title convention. Learn more about the pull request title convention used in this repository.

Environment

• KFP version: N/A

• KFP SDK version: 2.0.0b10

• All dependencies version:

kfp                      2.0.0b10
kfp-pipeline-spec        0.1.17
kfp-server-api           2.0.0a6

Steps to reproduce

In version 1.x.x it was possible to compile components that uses PipelineTaskFinalStatus, e.g using:

from kfp.v2.dsl import component, PipelineTaskFinalStatus, pipeline, ExitHandler
from kfp.components import load_component_from_file


@component(output_component_file='example.yaml')
def example(status: PipelineTaskFinalStatus):
    print(status)


loaded_component = load_component_from_file('example.yaml')

@pipeline
def example_pipeline():
    with ExitHandler(loaded_component()):
        pass

In version 2.0.0bx we expected this to still be possible, e.g with

from kfp.dsl import component, PipelineTaskFinalStatus, pipeline, ExitHandler
from kfp.compiler import Compiler
from kfp.components import load_component_from_file


@component
def example(status: PipelineTaskFinalStatus):
    print(status)


Compiler().compile(example, 'example.yaml')

loaded_component = load_component_from_file('example.yaml')

@pipeline
def example_pipeline():
    with ExitHandler(loaded_component()):
        pass

However, this is no longer possible for what seems like a variety of reasons. First of all the sdk explicitly forbids it in

https://github.com/kubeflow/pipelines/blob/fdf3ee7b68b2293d08e14c389b0dab9a57854e2a/sdk/python/kfp/compiler/pipeline_spec_builder.py#L333-L338

I tried commenting out the above check, which does enable us to compile valid component yaml without further changes:

# PIPELINE DEFINITION
# Name: example
# Inputs:
#    status: dict
components:
  comp-example:
    executorLabel: exec-example
    inputDefinitions:
      parameters:
        status:
          isOptional: true
          parameterType: STRUCT
deploymentSpec:
  executors:
    exec-example:
      container:
        args:
        - --executor_input
        - '{{$}}'
        - --function_to_execute
        - example
        command:
        - sh
        - -c
        - "\nif ! [ -x \"$(command -v pip)\" ]; then\n    python3 -m ensurepip ||\
          \ python3 -m ensurepip --user || apt-get install python3-pip\nfi\n\nPIP_DISABLE_PIP_VERSION_CHECK=1\
          \ python3 -m pip install --quiet     --no-warn-script-location 'kfp==2.0.0-beta.10'\
          \ && \"$0\" \"$@\"\n"
        - sh
        - -ec
        - 'program_path=$(mktemp -d)

          printf "%s" "$0" > "$program_path/ephemeral_component.py"

          python3 -m kfp.components.executor_main                         --component_module_path                         "$program_path/ephemeral_component.py"                         "$@"

          '
        - "\nimport kfp\nfrom kfp import dsl\nfrom kfp.dsl import *\nfrom typing import\
          \ *\n\ndef example(status: PipelineTaskFinalStatus):\n    print(status)\n\
          \n"
        image: python:3.7
pipelineInfo:
  name: example
root:
  dag:
    tasks:
      example:
        cachingOptions:
          enableCache: true
        componentRef:
          name: comp-example
        inputs:
          parameters:
            status:
              componentInputParameter: status
        taskInfo:
          name: example
  inputDefinitions:
    parameters:
      status:
        isOptional: true
        parameterType: STRUCT
schemaVersion: 2.1.0
sdkVersion: kfp-2.0.0-beta.10

Unfortunately, the input type of the status parameter is converted to STRUCT in the process, so loading in the component yaml results in a component that fails at runtime because the status argument is not provided by the backend as expected.

Expected result

It should be possible to compile components that use PipelineTaskFinalStatus for feature parity with version 1.x.x

Materials and Reference

Impacted by this bug? Give it a 👍.