Author: jbellis
Date: Sun May 16 22:29:11 2010
New Revision: 944927
URL: http://svn.apache.org/viewvc?rev=944927&view=rev
Log:
add Collections.shuffle(splits) so Hadoop doesn't send all the jobs to the same
nodes at once. patch by Joost Ouwerkerk; reviewed by jbellis for CASSANDRA-1096
Modified:
cassandra/branches/cassandra-0.6/CHANGES.txt
cassandra/branches/cassandra-0.6/src/java/org/apache/cassandra/hadoop/ColumnFamilyInputFormat.java
Modified: cassandra/branches/cassandra-0.6/CHANGES.txt
URL:
http://svn.apache.org/viewvc/cassandra/branches/cassandra-0.6/CHANGES.txt?rev=944927&r1=944926&r2=944927&view=diff
==============================================================================
--- cassandra/branches/cassandra-0.6/CHANGES.txt (original)
+++ cassandra/branches/cassandra-0.6/CHANGES.txt Sun May 16 22:29:11 2010
@@ -16,6 +16,8 @@
* install json2sstable, sstable2json, and sstablekeys to Debian package
* StreamingService.StreamDestinations wouldn't empty itself after streaming
finished (CASSANDRA-1076)
+ * added Collections.shuffle(splits) before returning the splits in
+ ColumnFamilyInputFormat (CASSANDRA-1096)
0.6.1
Modified:
cassandra/branches/cassandra-0.6/src/java/org/apache/cassandra/hadoop/ColumnFamilyInputFormat.java
URL:
http://svn.apache.org/viewvc/cassandra/branches/cassandra-0.6/src/java/org/apache/cassandra/hadoop/ColumnFamilyInputFormat.java?rev=944927&r1=944926&r2=944927&view=diff
==============================================================================
---
cassandra/branches/cassandra-0.6/src/java/org/apache/cassandra/hadoop/ColumnFamilyInputFormat.java
(original)
+++
cassandra/branches/cassandra-0.6/src/java/org/apache/cassandra/hadoop/ColumnFamilyInputFormat.java
Sun May 16 22:29:11 2010
@@ -119,7 +119,7 @@ public class ColumnFamilyInputFormat ext
}
assert splits.size() > 0;
-
+ Collections.shuffle(splits, new Random(System.nanoTime()));
return splits;
}