Repository navigation
Update logic in sync.py to match the effort in upstream #822
Description
Context
In kubeflow/community-distribution#3283, sync.py was updated to work with the change to seaweedfs for S3 storage.
The main change was to allow multi-user isolation in pipelines, as pipelines artifacts are now stored in separate buckets instead of the one at KFP_DEFAULT_PIPELINE_ROOT. The update also introduced changes to the creation of Deployments, as it doesn't create Deployments for ml-pipeline-visualizationserver and ml-pipeline-ui-artifact in the user namespace.
The diff between the 2 changes can be found here.
We should investigate how we want to modify the file to fit with our plan for S3 storage.
Activity
syncronize-issues-to-jira commentedon Jan 16, 2026
Thank you for reporting your feedback to us!
The internal ticket has been created: https://warthogs.atlassian.net/browse/KF-8359.
mvlassis commentedon Jan 20, 2026
Exploration
Handling of
ml-pipeline-uideployments in the user namespacesync.pywas updated in chore: Seaweedfs as Minio replacement kubeflow/pipelines#11965 to conditionally create theml-pipeline-uideployments in the user namespace as well as the respective service, see https://github.com/kubeflow/pipelines/blob/bc89227f26afb8c5601581c2b7b5634ce017728d/manifests/kustomize/base/installs/multi-user/pipelines-profile-controller/sync.py#L230.ARTIFACTS_PROXY_ENABLED.frontendimage in thekfp-uiconditionally uses this environment variable to redirect requests to theml-pipeline-uiservice in the user deployment.ARTIFACTS_PROXY_ENABLEDwas changed tofalsein chore: Seaweedfs as Minio replacement kubeflow/pipelines#11965. Comments in https://github.com/kubeflow/manifests/blob/1a12e1be42f69ad4a3d374dc3e9cb2da6eb26706/applications/pipeline/upstream/base/installs/generic/pipeline-install-config.yaml#L97-L104 explain that this change was made for security reasons, the reasoning behind the change being still unclear.Miscellaneous
kfp-profile-controllerfrom the current Python one toalpine/k8s, see this PR.