Python Kopf Operator Source-to-Image (s2i)

Source-to-Image for python kopf based operators.

Usage

This Python Kopf operator source-to-image can be used to build python-base operators using OpenShift source-to-image or Dockerfile based builds. This s2i image may also be used with odo to do in-cluster operator development.

Kopf Peering CustomResourceDefinitions

Unless Kopf standalone mode is used, peering between Kopf operators should be enabled. The Kopf peering custom resource definitions are required for peering and can be installed with:

oc apply -f https://raw.githubusercontent.com/redhat-cop/containers-quickstarts/master/build-s2i-python-kopf/peering-crds.yaml

Kopf Options

Kopf command line options are configured using environment variables. The file kopf-opt.sh may be used to set options at compile time. These are used to build up the command line for kopf run. The following environment variables are supported:

Option	Default	Description
KOPF_NAMESPACED	false	Set to "true" to run kopf with --namespace option to restrict watch to specific namespace.
KOPF_NAMESPACE	none	Namespace for kopf watch. When run in cluster this defaults to the operator pod namespace.
KOPF_STANDALONE	false	Set to "true" to run kopf with --standalone option. This prevents kopf from peering and talking to other kopf operators.
KOPF_PRIORITY	dynamic	Operator priority in relation to other kopf operators. The default when running containerized is to base the priority on the pod IP or otherwise ${RANDOM} if the IP address cannot be determined.
KOPF_PEERING	none	Kopf custom peering using --peering option.
KOPF_OPTIONS	none	Command line options for kopf run to support features not covered by other configuration variables.

Python Requirements

Python requirements can be specified in the requirements.txt file. These will be processed and installed to the container image during source-to-image build using Python3 pip.

Developing with odo

This source-to-image project supports odo for in-cluster operator development.

Installing odo

To use this image with odo you can create an image stream in the openshift namespace. This will make python-kopf appear as a valid project type for odo create.

oc create -f imagestream.yaml

Python Kopf Operator Examples

Simple Example

OpenShift Template Deployer

Deploying

The source-to-image run will configure Kopf to present TCP port 8080 for liveness probes.

Building this Source-to-Image Base Image

OpenShift

An OpenShift build template is provided for building this container image.

Create events console build configuration:

oc process --local -f build-template.yaml | oc apply -f -

Then either build the python-kopf-s2i image from git source:

oc start-build python-kopf-s2i

Or build the python-kopf-s2i image from local source:

oc start-build python-kopf-s2i --from-dir=../ -F

Dockerfile

This image may also be built with the provided Dockerfile.

Using podman:

sudo podman build --tag python-kopf-s2i:latest -f ./Dockerfile

Using docker:

sudo docker build -t python-kopf-s2i:latest .

Testing s2i build with the Provided Example

To verify the source-to-image base image is functional you will want to build an operator image to verify functionality. A simple operator example is provided in the examples/ directory.

OpenShift

If testing the source-to-image build in an OpenShift cluster then you may use the example build template provided.

oc process --local -f examples/kopf-simple/build-template.yaml \
-p FROM_IMAGE=$(oc get is python-kopf-s2i -o jsonpath="{.status.tags[?(@.tag=='latest')].items[0].dockerImageReference}") \
| oc apply -f -

Then either build the example operator from git source:

oc start-build kopf-simple -F

Or build the example from local source:

oc start-build kopf-simple --from-dir=../ -F

Description for how to deploy and test this example can be found in Simple Example.

Dockerfile

This example image may also be built with the provided Dockerfile. The provided configuration builds from the base image quay.io/redhat-cop/python-kopf-s2i:latest. In order to test your build of the s2i image you will need a modified version of the Dockerfile:

sed 's/^FROM .*/FROM python-kopf-s2i:latest/' example/Dockerfile >example/Dockerfile.custom

Build using podman:

sudo podman build --tag python-kopf-s2i-example:latest -f ./example/Dockerfile.custom

Using docker:

sudo docker build -t python-kopf-s2i-example:latest -f ./example/Dockerfile.custom example/