Use configuration template for single-instance ES deployment.

Until now it's only been used for clustered deployment,
but it works fine in single-instance mode, too.

playbooks/roles/elasticsearch/tasks/main.yml

 ---
 
 # elasticsearch
-# 
+#
 # Dependencies:
 #
 #   * common
 #   * oraclejdk
-# 
+#
 # Example play:
-#  
+#
 # This role can be used to do a single-server or clustered
 # installation of the elasticsearch service.  When a cluster
 # is being installed, there are two important things that
@@ -57,20 +57,17 @@
 - name: update elasticsearch defaults
   template: >
     src=etc/default/elasticsearch.j2 dest=/etc/default/elasticsearch
-  when: ELASTICSEARCH_CLUSTERED
-    
+
 - name: drop the elasticsearch config
   template: >
     src=edx/etc/elasticsearch/elasticsearch.yml.j2 dest={{ elasticsearch_cfg_dir }}/elasticsearch.yml
     mode=0744
-  when: ELASTICSEARCH_CLUSTERED
 
 - name: drop the elasticsearch logging config
   template: >
     src=edx/etc/elasticsearch/logging.yml.j2 dest={{ elasticsearch_cfg_dir }}/logging.yml
     mode=0744
-  when: ELASTICSEARCH_CLUSTERED  
-    
+
 - name: Ensure elasticsearch is enabled and started
   service: name=elasticsearch state=started enabled=yes