Label updater + documentation fix (#93)

* Adjusting pod label update logic to not track label as updated at detected update fails.

* Emphasize in documentation that node available resources must be meet by broker resource requirements.

Author: bczoma

README.md

@@ -37,7 +37,7 @@ For other PubSub+ Software Event Broker configurations or sizes, refer to the [P
 
 ### 1. Get a Kubernetes environment
 
-Follow your Kubernetes provider's instructions ([other options available here](https://kubernetes.io/docs/setup/)). Ensure you meet [minimum CPU, Memory and Storage requirements](docs/PubSubPlusK8SDeployment.md#cpu-and-memory-requirements) for the targeted PubSub+ Software Event Broker configuration size.
+Follow your Kubernetes provider's instructions ([other options available here](https://kubernetes.io/docs/setup/)). Ensure you meet [minimum CPU, Memory and Storage requirements](docs/PubSubPlusK8SDeployment.md#cpu-and-memory-requirements) for the targeted PubSub+ Software Event Broker configuration size. Important: the broker resource requirements refer to available resources on a [Kubernetes node](https://kubernetes.io/docs/concepts/scheduling-eviction/kube-scheduler/#kube-scheduler).
 > Note: If using [MiniKube](https://kubernetes.io/docs/setup/learning-environment/minikube/), use `minikube start` with specifying the options `--memory` and `--cpu` to assign adequate resources to the MiniKube VM. The recommended memory is 1GB plus the minimum requirements of your event broker.
 
 Also have the `kubectl` tool [installed](https://kubernetes.io/docs/tasks/tools/install-kubectl/) locally.
@@ -51,7 +51,7 @@ kubectl get nodes
 ### 2. Install and configure Helm
 
 Follow the [Helm Installation notes of your target release](https://github.com/helm/helm/releases) for your platform.
-Note that Helm is transitioning from v2 to v3. Many deployments still use v2. The event broker can be deployed using either version, however concurrent use of v2 and v3 from the same command-line environment is not supported.
+Note that Helm is transitioning from v2 to v3. Some deployments still use v2. The event broker can be deployed using either version, however concurrent use of v2 and v3 from the same command-line environment is not supported.
 
 On Linux a simple option to set up the latest stable release is to run:
 
@@ -100,7 +100,7 @@ Helm is configured properly if the command `helm version` returns no error.
 
 (Click on the arrow to open instructions for Helm v2 or v3)
 
-<details><summary><b>Instructions using Helm v2</b></summary>
+<details><summary><b>Install using Helm v2</b></summary>
 <p>
 
 a) Create a Solace PubSub+ Software Event Broker deployment for development purposes using `pubsubplus-dev`. It requires a minimum of 1 CPU and 3.6 GB of memory be available to the event broker pod.
@@ -123,7 +123,7 @@ helm install --name my-release solacecharts/pubsubplus-ha
 </p>
 </details>
 
-<details><summary><b>Instructions using Helm v3</b></summary>
+<details><summary><b>Install using Helm v3</b></summary>
 <p>
 
 a) Create a Solace PubSub+ Software Event Broker deployment for development purposes using `pubsubplus-dev`. It requires a minimum of 1 CPU and 2 GB of memory available to the event broker pod.

pubsubplus/templates/solaceConfigMap.yaml

@@ -237,14 +237,14 @@ data:
 
     # Function to set Kubernetes metadata labels
     set_label () {
+      # Params: $2 label set value; $3 file where internal tracking of label value stored
       #Prevent overdriving Kubernetes infra, don't set activity state to same as previous state
       previous_state=`cat $3`
       if [ "${2}" = "${previous_state}" ]; then
-        #echo "`date` INFO: ${APP}-Current and Previous state match (${2}), not updating label"
+        #echo "`date` INFO: ${APP}-Current and Previous state match (${2}), not updating pod label"
         :
       else
-        echo "`date` INFO: ${APP}-Updating label from `cat ${3}` to ${2}"
-        echo ${2} > ${3}
+        echo "`date` INFO: ${APP}-Updating pod label using K8s API from `cat ${3}` to ${2}"
         echo "[{\"op\": \"add\", \"path\": \"/metadata/labels/${1}\", \"value\": \"${2}\" }]" > /tmp/patch_label.json
         K8S=https://kubernetes.default.svc.cluster.local:$KUBERNETES_SERVICE_PORT
         KUBE_TOKEN=$(</var/run/secrets/kubernetes.io/serviceaccount/token)
@@ -254,18 +254,18 @@ data:
             --request PATCH --data "$(cat /tmp/patch_label.json)" \
             -H "Authorization: Bearer $KUBE_TOKEN" -H "Content-Type:application/json-patch+json" \
             $K8S/api/v1/namespaces/$NAMESPACE/pods/$HOSTNAME ; then
-          # Fall back to alternative method to update label
+          # Label update didn't work this way, fall back to alternative legacy method to update label
           if ! curl -sSk --output /dev/null -H "Authorization: Bearer $KUBE_TOKEN" --request PATCH --data "$(cat /tmp/patch_label.json)" \
             -H "Content-Type:application/json-patch+json" \
             https://$KUBERNETES_SERVICE_HOST:$KUBERNETES_PORT_443_TCP_PORT/api/v1/namespaces/$STATEFULSET_NAMESPACE/pods/$HOSTNAME ; then
             echo "`date` ERROR: ${APP}-Unable to update pod label, check access from pod to K8s API or RBAC authorization" >&2
             exit 1
           fi
-          echo "`date` INFO: ${APP}-Failed to update label from ${3} to ${2}, retrying"
         fi
+        echo ${2} > ${3} # Track latest value of label
       fi
     }
-    # note that there are no re-tries here, if check fails then return not ready.
+    # Main logic: note that there are no re-tries here, if check fails then return not ready.
     APP=`basename "$0"`
     state_file=/tmp/activity_state
     if [ ! -f ${state_file} ]; then  # State file not found, creating