Incorrect hostname for Pipeline SDK

[ianonavy]

When following Connect the Pipelines SDK to Kubeflow Pipelines on Kubeflow 1.6.1 in the subsection Full kubeflow (from inside cluster), I discovered the hard way that Kubernetes RBAC authorization is fulfilled at the API service, and not the UI service. I believe this documentation may be wrong:

import kfp

# the namespace in which you deployed Kubeflow Pipelines
namespace = "kubeflow"

# the KF_PIPELINES_SA_TOKEN_PATH environment variable is used when no `path` is set
# the default KF_PIPELINES_SA_TOKEN_PATH is /var/run/secrets/kubeflow/pipelines/token
credentials = kfp.auth.ServiceAccountTokenVolumeCredentials(path=None)

client = kfp.Client(host=f"http://ml-pipeline-ui.{namespace}", credentials=credentials)

print(client.list_experiments())

What I got to work was the following:

import kfp

# the namespace in which you deployed Kubeflow Pipelines
namespace = "kubeflow"

# the KF_PIPELINES_SA_TOKEN_PATH environment variable is used when no `path` is set
# the default KF_PIPELINES_SA_TOKEN_PATH is /var/run/secrets/kubeflow/pipelines/token
credentials = kfp.auth.ServiceAccountTokenVolumeCredentials(path=None)

client = kfp.Client(host=f"http://ml-pipeline.{namespace}:8888", credentials=credentials)

print(client.list_experiments())

I am not certain whether the documentation is wrong or we just have a weird installation. If someone can confirm that the documentation is indeed misleading, then I am happy to contribute a fix.

[varodrig]

@kubeflow/wg-pipeline-leads to provide feedback

[varodrig]

/area pipelines