1. Creating a pipeline on ml-pipeline-ui webpage is
saving the pipeline to database mlpipeline, delete a pipeline on ml-ppeline-ui webpage is deleting the record of the pipeline from database mlpipeline.
2. Create a pipeline-run on ml-pipeline-ui webpage is
saving the pipeline-run to database mlpipeline, and then pod workflow-controller of kubeflow pipeline creates a resource workflow based on file pipeline.yaml, on kubernetes cluster.
The backend of kubeflow pipeline is argo workflow, the workflow-controller of kubeflow pipeline is argo workflow controller, who manages creating of a workflow resource on kubernetes cluster, and running components in the workflow on kubernetes cluster.
3. xxx.yaml is
resource definition file of kuberneres cluster.
pipeline.yaml is definition file of kubeflow pipeline, namely argo workflow, which runs on kubernetes cluster.
4. pipeline = workflow = DAG in this context.
5. Artifacts on ml-pipeline-ui webpage are
artifacts stored in database metadb table Artifacts, and the artifacts shown on ml-pipeline-ui webpage are read from datatabase metadb table Artifacts.
