Addressed comments and added instruction to make the ha-example more standalone

Signed-off-by: Thomas Cooper <code@tomcooper.dev>

Author: tomncooper

ha-example/README.md

@@ -87,22 +87,26 @@ spec:
    image: quay.io/streamshub/flink-sql-runner:0.2.0
    flinkVersion: v2_0
    flinkConfiguration:
-    # job manager HA settings
-    execution.checkpointing.interval: 1min
-    state.checkpoints.dir: s3://test/cp
+   # job manager HA settings
+   execution.checkpointing.interval: 1min
+   state.checkpoints.dir: s3://test/cp
 ```
 The settings above will checkpoint the Task state every 1 minute under the s3 path provided.
 
 ## Example: Making the `recommendation-app` fault-tolerant and highly available
 
 Here, we will use the [recommendation-app](../recommendation-app) as an example to demonstrate the job manager HA.
-
+We are installing several additional components for this example so you may need a larger K8s cluster (more CPU and memory for you minikube deployment), than for the other examples.
 
 1. If you haven't already, install cert-manager (this creates cert-manager in a namespace called `cert-manager`):
    ```
    kubectl create -f https://github.com/jetstack/cert-manager/releases/download/v1.17.2/cert-manager.yaml
    kubectl wait deployment --all  --for=condition=Available=True --timeout=300s -n cert-manager
    ```
+1. Create the `flink` namespace:
+   ```shell
+   kubectl create namespace flink
+   ```
 1. Installing Flink Kubernetes Operator with leader election enabled like this:
    ```shell
    helm install flink-kubernetes-operator flink-operator-repo/flink-kubernetes-operator \
@@ -115,6 +119,17 @@ Here, we will use the [recommendation-app](../recommendation-app) as an example
    -n flink
    ```
 1. Follow the [guide](minio-install/README.md) to deploy and create a local S3 compatible storage service using minio and add a bucket named `test`.
+1. Deploy Apicurio, Strimzi and Kafka as per the instructions in the [README](../README.md#installing-apache-kafka-apache-flink-and-apicurio-registry).
+1. Deploy the data generator application to generate data for the recommendation app:
+   ```shell
+   kubectl -n flink apply -f ../recommendation-app/data-generator.yaml
+   ```
+   This will create a deployment that generates data and sends it to the Kafka topic `flink.recommendation.products`.
+1. Create a ConfigMap that holds product inventory data in CSV format.
+    ```shell
+    kubectl create configmap product-inventory --from-file ../recommendation-app/productInventory.csv -n flink
+    ```
+   The ConfigMap will be volume mounted to the recommendation-app pods.
 1. Deploy the `FlinkDeployment` CR with HA configured
    ```shell
    kubectl -n flink apply -f recommendation-app-HA/flink-deployment-ha.yaml

ha-example/minio-install/README.md

@@ -3,9 +3,10 @@
 This folder contains an example minio deployment yaml and the instructions are based on the [minio documentation](https://min.io/docs/minio/kubernetes/upstream/index.html).
 
 1. Deploy the minio with default configurations:
-   ```
+   ```shell
    kubectl -n flink apply -f minio.yaml
    ```
+   This will create the minio deployment and service for the minio API endpoint (that Flink will use to store state).
 1. Expose the minio WebUI:
    ```shell
    kubectl -n flink expose deployment minio --name=minio-ui --type=NodePort --port=9090
@@ -22,10 +23,6 @@ This folder contains an example minio deployment yaml and the instructions are b
 1. Create a bucket via MinIO WebUI. To do this, open the address found above and login with username: `minioadmin` and password: `minioadmin`, then go to `Buckets` -> `Create Bucket` to create a bucket called `test`.
 1. Monitor the files in the bucket.
    Click on the `Object Browser` to view the files in the buckets.
-1. Expose the minio API:
-   ```shell
-   kubectl -n flink expose deployment minio --name=minio-api --type=ClusterIP --port=9000
-   ```
 
 After minio is deployed and bucket is created, the flink configuration can be set like this in the `FlinkDeployment` CR:
 ```yaml

ha-example/minio-install/minio.yaml

@@ -30,3 +30,18 @@ spec:
           hostPath: # MinIO generally recommends using locally-attached volumes
             path: /mnt/disk1/data # Specify a path to a local drive or volume on the Kubernetes worker node
             type: DirectoryOrCreate # The path to the last directory must exist
+---
+apiVersion: v1
+kind: Service
+metadata:
+  labels:
+    app: minio
+  name: minio-api
+spec:
+  type: ClusterIP
+  selector:
+    app: minio
+  ports:
+    - port: 9000
+      protocol: TCP
+      targetPort: 9000