Setting up a local environment

This tutorial gives short instructions on how to set up a development environment for Porch on your local machine. It outlines the steps to get a kind cluster up and running to which a Porch instance running in Visual Studio Code can connect to and interact with. If you are not familiar with how porch works, it is highly recommended that you go through the Starting with Porch tutorial before going through this one.

Note

As your development environment, you can run the code on a remote VM and use the

VS Code Remote SSH plugin to connect to it.

Extra steps for MacOS users

The script the make deployment-config target to generate the deployment files for porch. The scripts called by this make target use recent bash additions. MacOS comes with bash 3.x.x

Install bash 4.x.x or better of bash using homebrew, see this this post for details
Ensure that /opt/homebrew/bin is earlier in your path than /bin and /usr/bin

Note

The changes above permanently change the bash version for all applications and may cause side effects.

Setup the environment automatically

The ./scripts/setup-dev-env.sh setup script automatically builds a porch development environment.

Note

This is only one of many possible ways of building a working porch development environment so feel free to customize it to suit your needs.

The setup script will perform the following steps:

Install a kind cluster. The name of the cluster is read from the PORCH_TEST_CLUSTER environment variable, otherwise it defaults to porch-test. The configuration of the cluster is taken from here.
Install the MetalLB load balancer into the cluster, in order to allow LoadBalancer typed Services to work properly.
Install the Gitea git server into the cluster. This can be used to test porch during development, but it is not used in automated end-to-end tests. Gitea is exposed to the host via port 3000. The GUI is accessible via http://localhost:3000/nephio, or http://172.18.255.200:3000/nephio (username: nephio, password: secret).

Note

If you are using WSL2 (Windows Subsystem for Linux), then Gitea is also accessible from the Windows host via the
http://localhost:3000/nephio URL.
Generate the PKI resources (key pairs and certificates) required for end-to-end tests.
Build the porch CLI binary. The result will be generated as .build/porchctl.

That’s it! If you want to run the steps manually, please use the code of the script as a detailed description.

The setup script is idempotent in the sense that you can rerun it without cleaning up first. This also means that if the script is interrupted for any reason, and you run it again it should effectively continue the process where it left off.

Extra manual steps

Copy the .build/porchctl binary (that was built by the setup script) to somewhere in your $PATH, or add the .build directory to your PATH.

Build and deploy porch

You can build all of porch, and also deploy it into your newly created kind cluster with this command.

make run-in-kind

See more advanced variants of this command in the detailed description of the development process.

Check that everything works as expected

At this point you are basically ready to start developing porch, but before you start it is worth checking that everything works as expected.

Check that the apiservice is ready

kubectl get apiservice v1alpha1.porch.kpt.dev

Sample output:

NAME                     SERVICE            AVAILABLE   AGE
v1alpha1.porch.kpt.dev   porch-system/api   True        18m

Check the porch api-resources

kubectl api-resources | grep porch

Sample output:

packagerevs                                      config.porch.kpt.dev/v1alpha1     true         PackageRev
packagevariants                                  config.porch.kpt.dev/v1alpha1     true         PackageVariant
packagevariantsets                               config.porch.kpt.dev/v1alpha2     true         PackageVariantSet
repositories                                     config.porch.kpt.dev/v1alpha1     true         Repository
packagerevisionresources                         porch.kpt.dev/v1alpha1            true         PackageRevisionResources
packagerevisions                                 porch.kpt.dev/v1alpha1            true         PackageRevision
packages                                         porch.kpt.dev/v1alpha1            true         PorchPackage

Create Repositories using your local Porch server

To connect Porch to Gitea, follow step 7 in the Starting with Porch tutorial to create the repositories in Porch.

You will notice logging messages in VS Code when you run the kubectl apply -f porch-repositories.yaml command.

You can check that your locally running Porch server has created the repositories by running the porchctl command:

porchctl repo get -A

Sample output:

NAME                  TYPE   CONTENT   DEPLOYMENT   READY   ADDRESS
external-blueprints   git    Package   false        True    https://github.com/nephio-project/free5gc-packages.git
management            git    Package   false        True    http://172.18.255.200:3000/nephio/management.git

You can also check the repositories using kubectl.

kubectl get  repositories -n porch-demo

Sample output:

NAME                  TYPE   CONTENT   DEPLOYMENT   READY   ADDRESS
external-blueprints   git    Package   false        True    https://github.com/nephio-project/free5gc-packages.git
management            git    Package   false        True    http://172.18.255.200:3000/nephio/management.git

You now have a locally running Porch (API)server. Happy developing!

Restart from scratch

Sometimes the development cluster gets cluttered and you may experience weird behavior from porch. In this case you might want to restart from scratch, by deleting the development cluster with the following command:

kind delete cluster --name porch-test

and running the setup script again:

./scripts/setup-dev-env.sh

Getting started with actual development

You can find a detailed description of the actual development process here.

Enabling Open Telemetry/Jaeger tracing

Enabling tracing on a Porch deployment

Follow the steps below to enable Open Telemetry/Jaeger tracing on your Porch deployment.

Apply the Porch deployment.yaml manifest for Jaeger.

kubectl apply -f https://raw.githubusercontent.com/nephio-project/porch/refs/heads/main/deployments/tracing/deployment.yaml

Add the environment variable OTEL to the porch-server manifest:

kubectl edit deployment -n porch-system porch-server

env:
- name: OTEL
  value: otel://jaeger-oltp:4317

Set up port forwarding of the Jaeger http port to your local machine:

kubectl port-forward -n porch-system service/jaeger-http 16686

Open the Jaeger UI in your browser at http://localhost:16686

Enable tracing on a local Porch server

Follow the steps below to enable Open Telemetry/Jaeger tracing on a porch server running locally on your machine, such as in VS Code.

Download the Jaeger binary tarball for your local machine architecture from the Jaeger download page and untar the tarball in some suitable directory.
Run Jaeger:

cd jaeger
./jaeger-all-in-one

Configure the Porch server to output Open Telemetry traces:

Set the OTEL environment variable to point at the Jaeger server

In .vscode/launch.json:

"env": {
   ...
   ...
"OTEL": "otel://localhost:4317",
   ...
   ...
}

In a shell:

export OTEL="otel://localhost:4317"

Open the Jaeger UI in your browser at http://localhost:16686
Run the Porch Server.