update doc

doc/source/configuration.rst

@@ -92,10 +92,12 @@ Sizing and tunning
 Roughly, Elassandra write throughput should be half the cassandra write throughput if you index all columns. If you only index a subset of columns, performance would be better. 
 
 Recommanded production setting for Apache cassandra and Elasticsearch can be applied to Elassandra :
+
 * Configure less than half the total memory of your server and up to 30.5Gb. Minimum recommended DRAM for production deployments is 32Gb.
 * Increase number of Elassandra node or use index partitionning to keep shards size below 50Gb.
 * Avoid huge wide rows, locking on a wide row can dramatically affect performance.
 * During indexing, if you don't need search, disable refresh (default is every second). 
+
 .. code::
 
    PUT /<my_index>/_settings -d '{ "index" : { "refresh_interval" : "-1" } }'