Testing our production-ready pipeline

Congratulations on setting up the production pipeline! Next, we will test its robustness. One great way to do this is to create a new release and observe and study whether the production pipeline successfully deploys the model to production (in the production Kubernetes cluster setup containing the pipeline). Follow these steps to test the pipeline:

1. First, create a new release, go to the Pipelines | Releases section, select your previously created pipeline (for example, Port Weather ML Pipeline), and click on the Create Release button at the top right-hand side of the screen to initiate a new release, as shown here:

   Figure 10.11 – Create a new release

2. Select the artifacts you would like to deploy in the pipeline (for example, Learn_MLOps repo, _scaler, and support-vector-classifier model and select their versions. Version 1 is recommended for testing PROD deployments for the first time), and click on the Create button at the top...