Merge lp:~bac/charmworld/update-rebuild-docs into lp:charmworld

[Brad Crittenden]

Update for rebuilding ES in a production or production-like environment.
R=rharding

Update for rebuilding ES in a production or production-like environment.

[Richard Harding (rharding)]

LGTM

review: Approve

[Juju Gui Bot (juju-gui-bot)]

review: Approve (continuous-integration)

=== modified file 'docs/rebuild-elasticsearch.rst'
--- docs/rebuild-elasticsearch.rst	2014-04-13 19:21:23 +0000
+++ docs/rebuild-elasticsearch.rst	2015-01-12 20:53:51 +0000
@@ -9,18 +9,31 @@
 First, move aside the entire contents of the ES database (substituting
 the current date -- in ISO 8601 format of course -- for "DATE").
 
+On jc-charmworld/0:
+    $ sudo /srv/deploymgr/charmworld stop
+
 On jc-elasticsearch/0:
 
     $ /etc/init.d/elasticsearch stop
     $ mv /var/lib/elasticsearch /var/lib/elasticsearch.DATE
     $ /etc/init.d/elasticsearch start
+    # Confirm an empty ES database.
+    $ curl -XGET 'http://localhost:9200/_status?pretty=true'
 
-Then on jc-charmworld/0:
+On jc-charmworld/0:
 
     $ sudo su - webops_deploy
     $ cd ~/charmworld
     $ INI=production.ini ./bin/es-update
 
+On jc-elasticsearch/0:
+
+    # Confirm a charmworld schema exists.
+    $ curl -XGET 'http://localhost:9200/_status?pretty=true'
+    
+On jc-charmworld/0:
+    $ sudo /srv/deploymgr/charmworld start
+
 The above will only take a few minutes (~5) to complete.  Next, wait for
 a set of enqueue/ingest jobs to run.  Once run, the ES database will be
 back in order.