Skip to content

Commit

Permalink
docs: added _Setting up and running a Kubernetes cluster locally with…
Browse files Browse the repository at this point in the history 
… Podman Desktop_

Signed-off-by: Fabrice Flore-Thébault <ffloreth@redhat.com>
  • Loading branch information
@themr0c
themr0c committed Apr 25, 2023
1 parent 23ccff4 commit e6562b2
Show file tree
Hide file tree
Showing 2 changed files with 247 additions and 1 deletion.
240 changes: 240 additions & 0 deletions website/blog/2023-04-19-running-a-local-kubernetes-cluster-with-podman-desktop.md
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,240 @@
---
title: Setting up and running a Kubernetes cluster locally with Podman Desktop
description: Setting up and running a Kubernetes cluster locally with Podman Desktop
slug: running-a-local-kubernetes-cluster-with-podman-desktop
authors: [themr0c]
tags: [podman-desktop, story, kubernetes, kind]
hide_table_of_contents: false
---

<!--truncate-->

## Installing Podman Desktop

Let's assume your workstation has no container engine installed yet.

You need Podman Desktop.
Let's install the software:

1. Go to [Podman Desktop installation documentation](/docs/Installation).
1. Click on your platform name: [Windows](/docs/installation/windows-install), [macOS](/docs/installation/macos-install), or [Linux](/docs/installation/linux-install).
1. Follow the instructions. Stick to the default installation method.
1. Start **Podman Destkop**

Now you have a graphical user interface to control your container engine.

#### Additional resources

- [Installing Podman Desktop on Windows](/docs/installation/windows-install)
- [Installing Podman Desktop on macOS](/docs/installation/macos-install)
- [Installing Podman Desktop on Linux](/docs/installation/linux-install).

## Choosing a container engine

You need a container engine.

Podman Desktop can control various container engines, such as:

- Docker
- Lima
- Podman

Choose the Podman container engine for:

- Added security
- No daemon
- Open source
- Rootless mode (for macOS and Linux: on Windows, it conflicts with Kind)

#### Additional resources

- [Podman website](https://podman.io/)

## Installing and initializing Podman

On Linux, you can install Podman natively.
See: [Installing Podman on Linux](https://podman.io/getting-started/installation#installing-on-linux).

On macOS and Windows, Podman requires to run in a virtual machine.
Podman Desktop helps you installing Podman and initializing a Podman machine:

1. Open Podman Desktop **Dashboard**
1. The **Dashboard** displays _Podman Desktop was not able to find an installation of Podman_.
1. Click on **Install**.
1. Podman Desktop checks the prerequites to install Podman Engine. When necessary, follow the instructions to install prerequisites.
1. Podman displays the dialog: _Podman is not installed on this system, would you like to install Podman?_. Click on **Yes** to install Podman.
1. Click on **Initialize Podman**.

Now you can start working with containers.

You want to deploy your application to a local Kubernetes cluster: now choose your Kubernetes provider.

#### Additional resources

- [Installing Podman on Windows](/docs/Installation/windows-install/installing-podman-with-podman-desktop)
- [Installing Podman on macOS](/docs/installation/macos-install)
- [Installing Podman on Linux](https://podman.io/getting-started/installation#installing-on-linux)

## Choosing a local Kubernetes provider

Podman Desktop can connect to any Kubernetes configured in your `kubeconfig` file.

Podman Desktop can help you run Kind-powered local Kubernetes clusters on a container engine, such as Podman.

Choose Kind.

## Installing Kind as local Kubernetes provider

Podman Desktop helps you [installing the `kind` CLI](/docs/kubernetes/kind/installing-kind):

1. In the status bar, click on **Kind**, and follow the prompts.
1. On Windows, [configure Podman in rootless mode](docs/kubernetes/kind/configuring-podman-for-kind-on-windows)

```shell-session
$ podman system connection default podman-machine-default-root
```

When the `kind` CLI is installed, the status bar doesn't display **Kind**.

## Starting a local Kind-powered Kubernetes cluster

1. Go to **Settings > Resources**
1. In the **Kind** tile, click on the **Create new ...** button.
1. Keep the default options, and click the **Create** button.
1. After successful creation, click on the **Go back to resources** button
1. In the **Podman Desktop** tray, open the **Kubernetes** menu: you can set the context to your Kind cluster: `kind-kind-cluster`.

Now you can start working with containers, and your local Kubernetes cluster.

#### Additional resources

- [Creating a local Kind-powered Kubernetes cluster](docs/kubernetes/kind/creating-a-kind-cluster)

## Choosing an application to deploy

Your choose to run the [Kubernetes documentation example: Deploying PHP Guestbook application with Redis](https://kubernetes.io/docs/tutorials/stateless-application/guestbook/).

https://github.com/praveenkumar/examples/tree/master/guestbook

## Starting the Redis database

The guestbook application uses Redis to store its data.

### Starting the Redis leader

With Podman Desktop, you can prepare the Redis leader image and container on your local container engine, and deploy the results to a Kubernetes pod and service.
This is functionally equivalent to the `redis-leader` deployment that the Kubernetes example propose.

#### Procedure

1. Open **<icon icon="fa-solid fa-cloud" size="lg" /> Images > <icon icon="fa-solid fa-arrow-circle-down" size="lg" /> Pull an image**.
1. **Image to Pull**: type `docker.io/redis:6.0.5`
1. Click **Pull image** to pull the image to your container engine local image registry.
1. Click **Done** to get back to the images list.
1. **<icon icon="fa-solid fa-cloud" size="lg" /> Search images...**: type `redis:6.0.5` to find the image.
1. Click **<icon icon="fa-solid fa-ellipsis-v" size="lg" /> > <icon icon="fa-solid fa-ellipsis-v" size="lg" /> Push image to Kind cluster**.
1. Click **<icon icon="fa-solid fa-play" size="lg" />** to open the **Create a container from image `docker.io/library/redis:6.0.5`** dialog.
1. **Container name**: type `leader`,
1. **Local port for 6379/tcp**: `6379`.
1. Click **<icon icon="fa-solid fa-play" size="lg" /> Start Container** to start the container in your container engine.
1. **<icon icon="fa-solid fa-cloud" size="lg" /> Search containers...**: type `leader` to find the running container.
1. Click **<icon icon="fa-solid fa-stop" size="lg" />** to stop the container, and leave the `6379` port available for the Redis follower container.

1. Click **<icon icon="fa-solid fa-ellipsis-v" size="lg" /> > <icon icon="fa-solid fa-rocket" size="lg" /> Deploy to Kubernetes** to open the **Deploy generated pod to Kubernetes** screen.
1. **Pod Name**: type `redis-leader`.
1. **Use Kubernetes Services**: select **Replace .hostPort exposure on containers by Services. It is the recommended way to expose ports, as a cluster policy may prevent to use hostPort.**

Check warning on line 145 in website/blog/2023-04-19-running-a-local-kubernetes-cluster-with-podman-desktop.md

View workflow job for this annotation

GitHub Actions / runner / vale 

[vale] reported by reviewdog 🐶
[RedHat.TermsWarnings] Consider using 'might (for possiblity)' or 'can (for ability)' rather than 'may' unless updating existing content that uses the term.

Raw Output:
{"message": "[RedHat.TermsWarnings] Consider using 'might (for possiblity)' or 'can (for ability)' rather than 'may' unless updating existing content that uses the term.", "location": {"path": "website/blog/2023-04-19-running-a-local-kubernetes-cluster-with-podman-desktop.md", "range": {"start": {"line": 145, "column": 162}}}, "severity": "WARNING"}
1. **Kubernetes namespaces**: select `guestbook`.
1. Click **<icon icon="fa-solid fa-rocket" size="lg" /> Deploy**.
1. Wait for the pod to reach the state: **Phase: Running**.
1. Click **Done**.

#### Verification

1. Open **<icon icon="fa-solid fa-cubes" size="lg" /> Pods**
1. The pods list has a running `redis-leader` pod.

### Starting the Redis followers

Although the Redis leader is a single Pod, you can make it highly available and meet traffic demands by adding a few Redis followers, or replicas.

With Podman Desktop, you can prepare the Redis follower image and container on your local container engine, and deploy the results to Kubernetes pods and services.
This is functionally equal to the `redis-follower` deployment that the Kubernetes example propose.

#### Procedure

1. Open **<icon icon="fa-solid fa-cloud" size="lg" /> Images > <icon icon="fa-solid fa-arrow-circle-down" size="lg" /> Pull an image**.
1. **Image to Pull**: type `gcr.io/google_samples/gb-redis-follower:v2`
1. Click **Pull image** to pull the image to your container engine local image registry.
1. Click **Done** to get back to the images list.
1. **<icon icon="fa-solid fa-cloud" size="lg" /> Search images...**: type `gb-redis-follower:v2` to find the image.

1. Click **<icon icon="fa-solid fa-ellipsis-v" size="lg" /> > <icon icon="fa-solid fa-ellipsis-v" size="lg" /> Push image to Kind cluster**.
1. Click **<icon icon="fa-solid fa-play" size="lg" />** to open the **Create a container from image docker.io/library/redis:6.0.5** dialog.

Check failure on line 172 in website/blog/2023-04-19-running-a-local-kubernetes-cluster-with-podman-desktop.md

View workflow job for this annotation

GitHub Actions / runner / vale 

[vale] reported by reviewdog 🐶
[RedHat.CaseSensitiveTerms] Use 'Redis' rather than 'redis'.

Raw Output:
{"message": "[RedHat.CaseSensitiveTerms] Use 'Redis' rather than 'redis'.", "location": {"path": "website/blog/2023-04-19-running-a-local-kubernetes-cluster-with-podman-desktop.md", "range": {"start": {"line": 172, "column": 119}}}, "severity": "ERROR"}
1. **Container name**: type `follower`,
1. **Local port for 6379/tcp**: `6379`.
1. Click **<icon icon="fa-solid fa-play" size="lg" /> Start Container** to start the container in your container engine.
1. **<icon icon="fa-solid fa-cloud" size="lg" /> Search containers...**: type `follower` to find the running container.
1. Click **<icon icon="fa-solid fa-stop" size="lg" />** to stop the container: you don't need it to run in the container engine.
1. Click **<icon icon="fa-solid fa-ellipsis-v" size="lg" /> > <icon icon="fa-solid fa-rocket" size="lg" /> Deploy to Kubernetes** to open the **Deploy generated pod to Kubernetes** screen.
1. **Pod Name**: type `redis-follower`.
1. **Use Kubernetes Services**: select **Replace .hostPort exposure on containers by Services. It is the recommended way to expose ports, as a cluster policy may prevent to use hostPort.**

Check warning on line 180 in website/blog/2023-04-19-running-a-local-kubernetes-cluster-with-podman-desktop.md

View workflow job for this annotation

GitHub Actions / runner / vale 

[vale] reported by reviewdog 🐶
[RedHat.TermsWarnings] Consider using 'might (for possiblity)' or 'can (for ability)' rather than 'may' unless updating existing content that uses the term.

Raw Output:
{"message": "[RedHat.TermsWarnings] Consider using 'might (for possiblity)' or 'can (for ability)' rather than 'may' unless updating existing content that uses the term.", "location": {"path": "website/blog/2023-04-19-running-a-local-kubernetes-cluster-with-podman-desktop.md", "range": {"start": {"line": 180, "column": 162}}}, "severity": "WARNING"}
1. **Kubernetes namespaces**: select `guestbook`.
1. Click **<icon icon="fa-solid fa-rocket" size="lg" /> Deploy**.
1. Wait for the pod to reach the state: **Phase: Running**.
1. Click **Done**.
1. To add replicas, repeat the last step with another **Pod Name** value.

#### Verification

1. Open **<icon icon="fa-solid fa-cubes" size="lg" /> Pods**
1. The pods list has a running `redis-follower` pod.

#### Known issue

Currently, you need a workaround to increase the replicas with Podman Desktop. The team in working on an impromvement:

- [Create a deployment rather than a pod](https://github.com/containers/podman-desktop/issues/1323)
- [Configure `replicas` count](https://github.com/containers/podman-desktop/issues/2210)

## Starting and exposing the Guestbook frontend

Now that you have the Redis storage of your guestbook up and running, start the guestbook web servers.
Like the Redis followers, the frontend is deployed using Kubernetes pods and services.

The guestbook app uses a PHP frontend.
It is configured to communicate with either the Redis follower or leader Services, depending on whether the request is a read or a write.
The frontend exposes a JSON interface, and serves a jQuery-Ajax-based UX.

With Podman Desktop, you can prepare the Guestbook frontend image and container on your local container engine, and deploy the results to Kubernetes pods and services.
This is functionally equal to the `frontend` deployment that the Kubernetes example propose.

#### Procedure

1. Open **<icon icon="fa-solid fa-cloud" size="lg" /> Images > <icon icon="fa-solid fa-arrow-circle-down" size="lg" /> Pull an image**.
1. **Image to Pull**: type `gcr.io/google_samples/gb-frontend:v5`
1. Click **Pull image** to pull the image to your container engine local image registry.
1. Wait for the pull to complete.
1. Click **Done** to get back to the images list.
1. **<icon icon="fa-solid fa-cloud" size="lg" /> Search images...**: type `gb-frontend:v5` to find the image.

1. Click **<icon icon="fa-solid fa-ellipsis-v" size="lg" /> > <icon icon="fa-solid fa-ellipsis-v" size="lg" /> Push image to Kind cluster**.
1. Click **<icon icon="fa-solid fa-play" size="lg" />** to open the **Create a container from image `gcr.io/google_samples/gb-frontend:v5`** dialog.
1. **Container name**: type `frontend`,
1. **Local port for 80/tcp**: `9000`.
1. Click **<icon icon="fa-solid fa-play" size="lg" /> Start Container** to start the container in your container engine.
1. **<icon icon="fa-solid fa-cloud" size="lg" /> Search containers...**: type `frontend` to find the running container.
1. Click **<icon icon="fa-solid fa-stop" size="lg" />** to stop the container: you don't need it to run in the container engine.
1. Click **<icon icon="fa-solid fa-ellipsis-v" size="lg" /> > <icon icon="fa-solid fa-rocket" size="lg" /> Deploy to Kubernetes** to open the **Deploy generated pod to Kubernetes** screen.
1. **Pod Name**: type `frontend`.
1. **Use Kubernetes Services**: select **Replace .hostPort exposure on containers by Services. It is the recommended way to expose ports, as a cluster policy may prevent to use hostPort.**

Check warning on line 229 in website/blog/2023-04-19-running-a-local-kubernetes-cluster-with-podman-desktop.md

View workflow job for this annotation

GitHub Actions / runner / vale 

[vale] reported by reviewdog 🐶
[RedHat.TermsWarnings] Consider using 'might (for possiblity)' or 'can (for ability)' rather than 'may' unless updating existing content that uses the term.

Raw Output:
{"message": "[RedHat.TermsWarnings] Consider using 'might (for possiblity)' or 'can (for ability)' rather than 'may' unless updating existing content that uses the term.", "location": {"path": "website/blog/2023-04-19-running-a-local-kubernetes-cluster-with-podman-desktop.md", "range": {"start": {"line": 229, "column": 162}}}, "severity": "WARNING"}
1. **Kubernetes namespaces**: select `default`.
1. Click **<icon icon="fa-solid fa-rocket" size="lg" /> Deploy**.
1. Wait for the pod to reach the state: **Phase: Running**.
1. Click **Done**.
1. FIXME: Expose the frontend.

#### Verification

1. Open **<icon icon="fa-solid fa-cubes" size="lg" /> Pods**
1. The pods list has a running `frontend` pod.
1. FIXME: access the application.
8 changes: 7 additions & 1 deletion website/blog/authors.yml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -20,4 +20,10 @@ deboer:
name: Tim deBoer
title: Architect
url: https://github.com/deboer-tim
image_url: https://github.com/deboer-tim.png
image_url: https://github.com/deboer-tim.png

themr0c:
name: Fabrice Flore-Thebault
title: Technical writer
url: https://github.com/themr0c
image_url: https://github.com/themr0c.png

0 comments on commit e6562b2

Please sign in to comment.