Compile and Run a Kubeflow Pipelines v2 Pipeline with the KFP SDK v2
Author, compile, and submit a Kubeflow Pipelines v2 pipeline to an existing cluster using the KFP SDK v2, then verify the run with kubectl and the Pipelines UI.
Loading…
ReadMeFeed / Community knowledge
#kubeflow-pipelines
Real questions. Useful conversations. Find the people who know your stack.
kubeflowReusing Completed Kubeflow Pipeline Steps with Component Caching
Kubeflow Pipelines can skip expensive steps when inputs and images match, but caching behaves differently across SDK versions and can hide stale results.
From Local Script to Reproducible Workflow: Kubeflow Pipelines v2
A practical pattern for wrapping ML scripts as typed KFP v2 components with pinned images, plus local compile checks and cluster verification steps.
kubernetesWhen Kubeflow Pipelines Earns Its Place (and When a CronJob Wins)
Kubeflow Pipelines pays off when you need caching, lineage, and scheduled retraining on Kubernetes — but a plain CronJob is often the cheaper right answer. Here's how to decide, with a working KFP v2 example.
Adopting Kubeflow Pipelines v2: Portable YAML, Local Testing, and Trade‑offs
Kubeflow Pipelines v2 shifts from imperative Python to a compiled IR, letting teams catch type errors early, test locally, and run the same YAML on any compatible backend—while facing limits on dynamic control flow.
Kubeflow Pipelines and TFJob: When the Extra Controller Earns Its Keep
Wrapping a training step in a TFJob adds a controller, a CRD, and an RBAC surface. Here is how to tell whether your pipeline actually needs it.
Tracking Kubeflow TF Training Runs with MLMD: A Working Setup and Its Pitfalls
Wire a Kubeflow Pipelines TensorFlow training component into MLMD so runs, parameters, and model artifacts are tracked durably — and avoid the ephemeral-storage mistake that silently erases metadata.
tensorflowKubeflow Pipelines and Jupyter Notebooks Ingress authentication and NetworkPolicy interoperability for public exposure prevention
Kubeflow Pipelines and Jupyter Notebooks coexist within a single Kubernetes deployment, each exposing services through Ingress and service-type configurations that govern external access. A core interoperability question emerges from the interaction between Jupyter Notebook Ingress authentication—TLS secrets and OIDC middleware—and the absence of default Net
Kubeflow Pipelines ↔ KServe: Zero‑Downtime Model Migration During Cluster Upgrade
Goal To migrate a small Kubeflow application from one cluster to another while keeping model inference traffic continuous, the migration must avoid any interruption during both the Pipelines control‑plane upgrade and the KServe model revision rollout. Constraints and Uncertainty Argo Workflows, the engine behind Kubeflow Pipelines, does not expose a document