Author: jbellis
Date: Tue Sep 28 19:45:21 2010
New Revision: 1002339
URL: http://svn.apache.org/viewvc?rev=1002339&view=rev
Log:
update storage-conf to more-strongly encourage setting InitialToken, and to
reference Operations wiki page
Modified:
cassandra/branches/cassandra-0.6/conf/storage-conf.xml
Modified: cassandra/branches/cassandra-0.6/conf/storage-conf.xml
URL:
http://svn.apache.org/viewvc/cassandra/branches/cassandra-0.6/conf/storage-conf.xml?rev=1002339&r1=1002338&r2=1002339&view=diff
==============================================================================
--- cassandra/branches/cassandra-0.6/conf/storage-conf.xml (original)
+++ cassandra/branches/cassandra-0.6/conf/storage-conf.xml Tue Sep 28 19:45:21
2010
@@ -186,16 +186,16 @@
<Partitioner>org.apache.cassandra.dht.RandomPartitioner</Partitioner>
<!--
- ~ If you are using an order-preserving partitioner and you know your key
- ~ distribution, you can specify the token for this node to use. (Keys
- ~ are sent to the node with the "closest" token, so distributing your
- ~ tokens equally along the key distribution space will spread keys
- ~ evenly across your cluster.) This setting is only checked the first
- ~ time a node is started.
-
- ~ This can also be useful with RandomPartitioner to force equal spacing
- ~ of tokens around the hash space, especially for clusters with a small
- ~ number of nodes.
+ ~ You should always specify InitialToken when setting up a production
+ ~ cluster for the first time, and often when adding capacity later.
+ ~ The principle is that each node should be given an equal slice of
+ ~ the token ring; see http://wiki.apache.org/cassandra/Operations
+ ~ for more details.
+ ~
+ ~ If blank, Cassandra will request a token bisecting the range of
+ ~ the heaviest-loaded existing node. If there is no load information
+ ~ available, such as is the case with a new cluster, it will pick
+ ~ a random token, which will lead to hot spots.
-->
<InitialToken></InitialToken>
