kubeflow
/
pipelines
mirror of https://github.com/kubeflow/pipelines.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
pipelines/manifests/gcp_marketplace
History
Yuan Gong d79071c0be chore(release): bumped version to 1.4.1 2021-02-25 23:50:18 +00:00
..
chart/kubeflow-pipelines chore(release): bumped version to 1.4.1 2021-02-25 23:50:18 +00:00
deployer fix issue of creating default bucket (#3626) 2020-04-26 23:05:28 -07:00
hack feat: GCP marketplace - preserve install config in configmap and secret (#4471) 2020-09-06 22:39:41 -07:00
test feat: GCP marketplace - preserve install config in configmap and secret (#4471) 2020-09-06 22:39:41 -07:00
OWNERS chore: update OWNERS 2020-11-09 10:18:18 +08:00
README.md doc(gcp): recommend KFP standalone over marketplace CLI installation (#4508) 2020-09-16 21:20:45 -07:00
cli.md doc(gcp): recommend KFP standalone over marketplace CLI installation (#4508) 2020-09-16 21:20:45 -07:00
guide.md Add BigQuery option (#3097) 2020-02-26 23:03:20 -08:00
schema.yaml chore(release): bumped version to 1.4.1 2021-02-25 23:50:18 +00:00

README.md

Kubeflow Pipelines for GKE Marketplace

Using the Google Cloud Platform Marketplace

Get up and running with a few clicks! Install this Kubeflow Pipelines app to a Google Kubernetes Engine cluster using Google Cloud Marketplace. Follow the on-screen instructions and Google Cloud AI Platform Pipelines documentation.

Using the command line

For users, please refer to Kubeflow Pipelines Standalone on how to install via commandline using kustomize manifests. The installation is almost the same as AI Platform Pipelines.

Refer to the Installation Options for Kubeflow Pipelines doc for all installation options and their feature comparison.

Developement guide

Only for developing AI Platform Pipelines purpose, we have a CLI installation guide for AI Platform Pipelines. It's not suitable for end users. The tool "mpdev" is for Kubeflow Pipeline developers.

This section details how to test your changes before submit codes.

  1. Code changes and locally committed

  2. Build

gcloud builds submit --config=.cloudbuild.yaml --substitutions=COMMIT_SHA="$(git rev-parse HEAD)" --project=ml-pipeline-test

gcr.io/$PROJECT_ID/hosted/$COMMIT_SHA/ contains the binaries.

  1. Auto-test (Install & Uninstall)

MM_VER is major minor version parsed from VERSION file which is on major.minor.patch version format.

MM_VER=$(cat VERSION | sed -e "s#[^0-9]*\([0-9]*\)[.]\([0-9]*\)[.]\([0-9]*\)#\1.\2#")
gcloud builds submit --config=test/cloudbuild/mkp_verify.yaml --substitutions=COMMIT_SHA="$(git rev-parse HEAD)",_DEPLOYER_VERSION=$MM_VER --project=ml-pipeline-test
  1. Manual-test (Install with advanced parameters and don't uninstall)

Make sure your kubectl can connect to a target test cluster.

APP_INSTANCE_NAME=<yours>
NAMESPACE=<yours> # Make sure you already created the namespace
MANAGEDSTORAGE=true # True means use CloudSQL + Minio-GCS; False means use in-cluster PVC + MySQL.
CLOUDSQL=<yours> # Format like project_id:zone:cloudsql_instance_name
PROJECTID=<yours> # This field will be removed after Marketplace can pass in the project ID
mpdev install  --deployer=gcr.io/ml-pipeline-test/hosted/$(git rev-parse HEAD)/deployer:$MM_VER \
    --parameters='{"name": "'$APP_INSTANCE_NAME'", "namespace": "'$NAMESPACE'", "managedstorage.enabled": '$MANAGEDSTORAGE', "managedstorage.cloudsqlInstanceConnectionName": "'$CLOUDSQL'"}'