[elasticsearch] Tweak the 'readinessProbe' command to verify that master

nodes are available.

The behaviour of the `/` endpoint changed[0] between 6.x and 7.x, whereby
previously it would return a HTTP `503` response when the cluster was
blocked, it now returns a HTTP `200` response even if there are no masters
available.

This change updates the behaviour of the `readinessProbe` command during
normal running to verify that the local node is responding and that there
are master nodes available. [1]

The desired behaviour here is that if the data nodes are unable to talk to
their master nodes for whatever reason, then the data nodes will become
`Unready` and therefore be removed from the Service load-balancer until
the master nodes are available again.

Refs:
[0] elastic/elasticsearch#29045
[1] elastic/elasticsearch#34897 (comment)

elasticsearch/templates/statefulset.yaml

@@ -211,8 +211,8 @@ spec:
                 }
 
                 if [ -f "${START_FILE}" ]; then
-                    echo 'Elasticsearch is already running, lets check the node is healthy'
-                    http "/"
+                    echo 'Elasticsearch is already running, lets check the node is healthy and there are master nodes available'
+                    http "/_cluster/health?timeout=0s"
                 else
                     echo 'Waiting for elasticsearch cluster to become cluster to be ready (request params: "{{ .Values.clusterHealthCheckParams }}" )'
                     if http "/_cluster/health?{{ .Values.clusterHealthCheckParams }}" ; then