Tektoncd-operator

FAQ

click here

Configurations

Override Images:

Following environment variables allows overriding images of pipelines, trigger and task addons components

1. IMAGE_PIPELINES_<DEPLOYMENT-IMAGE-NAME> e.g. IMAGE_PIPELINES_WEBHOOK It allows overriding pipelines or triggers deployment images. Note, *_PIPELINE_* will override images in pipelines manifest only. Same way IMAGE__TRIGGERS_*
2. IMAGE_PIPELINES__ARG_<DEPLOYMENT-IMAGE-ARG-NAME> e.g. IMAGE_PIPELINES_ARG_NOP It allows overriding pipelines or triggers deployment images of containers args. Note, *_PIPELINE_ARG_* will override images in pipelines manifest only. Same way IMAGE__TRIGGERS_ARG_*
3. IMAGE_ADDONS_<STEP-NAME> e.g. IMAGE_ADDONS_PUSH It allows overriding ClusterTask addons steps images.
4. IMAGE_ADDONS_PARAM_<NAME> e.g. IMAGE_ADDONS_PARAM_BUILDER It allows overriding ClusterTask addons params images. the _PARAM_ will replace the value of Task.Spec.Param.

Note: IF IMAGE NAME, IMAGE ARGUMENT, STEP NAME AND PARAMETER NAME HAS "-" IN IT, THEN PLEASE SUBSTITUTE IT BY "_"

Note: BE CAUTIOUS WHILE SUBSTITUTING THE IMAGES. FOR INSTANCE, IF DEPLOYMENT HAS MORE THAN ONE CONTAINER AND CONTAINER HAS SAME NAME AS DEFINED IN THE OVERRIDE IMAGE CONFIGURATION, THEN THIS COULD RESULT IN UNWATED IMAGE SUBSTITUTATION

Dev env

Checkout your fork

The Go tools require that you clone the repository to the src/github.com/openshift/tektoncd-pipeline-operator directory in your GOPATH.

To check out this repository:

1. Create your own fork of this repo
2. Clone it to your machine:

mkdir -p ${GOPATH}/src/github.com/openshift
cd ${GOPATH}/src/github.com/openshift
git clone git@github.com:${YOUR_GITHUB_USERNAME}/tektoncd-pipeline-operator.git
cd tektoncd-pipeline-operator
git remote add upstream git@github.com:tektoncd/tektoncd-pipeline-operator.git
git remote set-url --push upstream no_push

Prerequisites

You must install these tools:

1. go: The language Tektoncd-pipeline-operator is built in
2. git: For source control
3. kubectl: For interacting with your kube cluster
4. operator-sdk
5. opm
6. yamllint

Install Minikube

Create minikube instance

minikube start -p mk-tekton \
 --cpus=4 --memory=8192 --kubernetes-version=v1.12.0 \
 --extra-config=apiserver.enable-admission-plugins="LimitRanger,NamespaceExists,NamespaceLifecycle,ResourceQuota,ServiceAccount,DefaultStorageClass,MutatingAdmissionWebhook"  \
 --extra-config=apiserver.service-node-port-range=80-32767

Set the shell environment up for the container runtime

eval $(minikube docker-env -p mk-tekton)

Development build

Development Guide

Running tests

Install OLM

Clone OLM repository (into go path)

git clone git@github.com:operator-framework/operator-lifecycle-manager.git \
          $GOPATH/src/github.com/operator-framework/

Install OLM

Ensure minikube is installed and docker env is set see above

cd $GOPATH/src/github.com/operator-framework/operator-lifecycle-manager

GO111MODULE=on NO_MINIKUBE=true make run-local

NOTE: NO_MINIKUBE=true: we don't want to start a new minikube instance while installing OLM

Launch web console

Open a new terminal

cd $GOPATH/src/github.com/operator-framework/operator-lifecycle-manager

./scripts/run_console_local.sh

Deploy openshift-pipelines-operator on minikube for testing

1. Change directory to ${GOPATH}/src/github.com/openshift/tektoncd-pipeline-operator

2. Create openshift-operators namespace

   kubectl create namespace openshift-operators

3. Apply operator crd

   kubectl apply -f deploy/crds/*_crd.yaml

4. Deploy the operator

   kubectl apply -f deploy/ -n openshift-operators

5. Install pipeline by creating an Install CR

   kubectl apply -f deploy/crds/*_cr.yaml

Deploy openshift-pipelines-operator using CatalogSource on OLM

1. Install minikube see above

2. Install olm see above

3. Create openshift-operators namespace

   kubectl create namespace openshift-operators

4. Generate local catalog source

   NAMESPACE=operators ./scripts/olm_catalog.sh > olm/openshift-pipelines-operator.resources.yaml

5. Add local catalog source

   kubectl apply -f olm/openshift-pipelines-operator.resources.yaml

   Once the CatalogSource has been applied, you should find it under Catalog > Operator Management of the web console

6. Subscribe to Openshift Pipelines Operator

   1. Open web console

   2. Select openshift-operators namespace

   3. Select Catalog > Operator Management

   4. Select Catalog > Operator Management > Operator Catalogs

   5. Scroll down to Openshift Pipelines Operator under Openshift Pipelines Operator Registry

      NOTE: it will take a few minutes to appear after applying the catalogsource

   6. Click Create Subscription button

      1. ensure namespace in yaml is openshift-operator e.g.

         sample subscription

           apiVersion: operators.coreos.com/v1alpha1
           kind: Subscription
           metadata:
             generateName: openshift-pipelines-operator-
             namespace: openshift-operators
           spec:
             source: openshift-pipelines-operator-registry
             sourceNamespace: openshift-operators
             name: openshift-pipelines-operator
             startingCSV: openshift-pipelines-operator.v0.3.1
             channel: alpha

      2. Click Create button at the bottom

7. Verify operator is installed successfully

   1. Select Catalog > Installed operators
   2. Look for Status InstallSucceeded

8. Install Tektoncd-Pipeline by creating an install CR

   1. Select Catalog > Developer Catalog, you should find Openshift Pipelines Install

   2. Click on it and it should show the Operator Details Panel

   3. Click on Create which show an example as below

      example

          apiVersion: tekton.dev/v1alpha1
          kind: Install
          metadata:
          name: pipelines-install
          namespace: openshift-pipelines-operator
          spec: {}

      NOTE: This will install Openshift Pipeline resources in Tekton-Pipelines Namespace

   4. Verify that the pipeline is installed

      1. Ensure pipeline pods are running

         kubectl get all -n tekton-pipelines

      2. Ensure pipeline crds exist

         kubectl get crds | grep tekton

         should show

         clustertasks.tekton.dev
         installs.tekton.dev
         pipelineresources.tekton.dev
         pipelineruns.tekton.dev
         pipelines.tekton.dev
         taskruns.tekton.dev
         tasks.tekton.dev

   NOTE: Now TektonCD Pipelines can be created and run

End to End workflow

This section explains how to test changes to the operator by executing the entire end-to-end workflow of edit, test, build, package, etc...

It assumes you have already followed install minikube and OLM.

Generate new image, CSV

1. Make changes to the operator
2. Test operator locally with operator-sdk run --local
3. Build operator image operator-sdk build <imagename:tag>
4. Update image reference in deploy/operator.yaml
5. Update image reference in CSV deploy/olm-catalog/openshift-pipelines-operator/0.3.1/openshift-pipelines-operator.v0.3.1.clusterserviceversion.yaml

Update Local CatalogSource

1. 1. Build local catalog source localOperators

   NAMESPACE=operators ./scripts/olm_catalog.sh > olm/openshift-pipelines-operator.resources.yaml