docs: local test (#247)

* docs: new file explaining local test

Author: kekcleader

docs/contributing.md

@@ -25,7 +25,13 @@ For this you need to export the `KUBEBUILDER_ASSETS` environment variable:
 ```shell
 KUBEBUILDER_ASSETS=$(pwd)/bin/k8s/$DIR_WITH_ASSETS
 # where $DIR_WITH_ASSETS is the directory that contains binaries for your OS.
+o
 ```
+
+### Live Local Test
+
+See [Local Test](local_test.md).
+
 ### See Test Coverage
 
 You can check the coverage as HTML report:

docs/local_test.md

@@ -0,0 +1,76 @@
+# Test Locally
+
+**Warning!** This test is for those who have access to `helm-charts-priv`.
+
+## Run and check cluster
+
+1. Create and run a cluster if it is not running yet.
+
+```shell
+git clone https://github.com/openmfp/helm-charts-priv.git
+cd helm-charts-priv
+task local-setup
+```
+
+2. Verify that the cluster is running.
+
+Run k9s, go to `:pods`. All pods must have a status of "Running".
+It may take some time before they are all ready. The possible issues may be insufficient RAM and/or CPU cores. In this case, increase the limits in Docker settings.
+
+3. In k9s, go to `:pods`, then open pod `kubernetes-graphql-gateway-...`.
+
+Open container `kubernetes-graphql-gateway-gateway` to see the logs.
+The logs must contain more than a single line (with "Starting server...").
+If you see only this single line, the problem might be in the container called "kubernetes-graphql-gateway-listener".
+
+Note the `IMAGE` column, corresponding to the two `kubernetes-...` container. It contains the name and the currently used version of the build, i.e.
+```
+ghcr.io/openmfp/kubernetes-graphql-gateway:v0.75.1
+```
+
+4. Build the Docker image:
+```shell
+task docker
+```
+
+5. Tag the newly built image with the version used in local-setup -- that image is going to be replaced with the one built on step 4.
+```shell
+docker tag ghcr.io/openmfp/kubernetes-graphql-gateway:latest ghcr.io/openmfp/kubernetes-graphql-gateway:v0.75.1
+```
+Use the name you and version got from the `IMAGE` column on step 3. Leave the version number unchanged.
+
+6. Check your cluster name:
+```shell
+kind get clusters
+```
+In this example, the cluster name is `openmfp`.
+
+7. Load the new image into your kind cluster:
+***Docker-based kind:***
+```shell
+kind load docker-image ghcr.io/openmfp/kubernetes-graphql-gateway:v0.75.1 -n openmfp
+```
+The argument `-n openmfp` is to change the default value of the cluster name, which is `kind`.
+
+***Podman-based kind:***
+- Pull (or build) the image locally with Podman:
+```shell
+podman pull ghcr.io/openmfp/kubernetes-graphql-gateway:v0.75.1
+```
+
+- Save it to a tarball in OCI-archive format:
+```shell
+podman save --format oci-archive ghcr.io/openmfp/kubernetes-graphql-gateway:v0.75.1 -o kubernetes-graphql-gateway_v0.75.1.tar
+```
+
+- Load the tarball into your Podman-backed kind cluster:
+```shell
+kind load image-archive kubernetes-graphql-gateway_v0.75.1.tar -n openmfp
+```
+
+8. In k9s, go to `:pods` and delete the pod (not the container) called `kubernetes-graphql-gateway-...`.
+
+Kubernetes will immediately recreate the pod -- but this time it will use the new version of the build.
+
+9. Once the pod is recreated, go to [https://portal.dev.local:8443](https://portal.dev.local:8443)
+and check if everything works fine.