You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Steeve Morin (JIRA)" <ji...@apache.org> on 2013/01/15 17:30:12 UTC

[jira] [Created] (CASSANDRA-5161) Slow secondary index performance when using VNodes

Steeve Morin created CASSANDRA-5161:
---------------------------------------

             Summary: Slow secondary index performance when using VNodes
                 Key: CASSANDRA-5161
                 URL: https://issues.apache.org/jira/browse/CASSANDRA-5161
             Project: Cassandra
          Issue Type: Bug
          Components: Core
    Affects Versions: 1.2.0
            Reporter: Steeve Morin


When having a big number of vnodes (such as 256/node advised by http://www.datastax.com/dev/blog/upgrading-an-existing-cluster-to-vnodes), using secondary indices becomes really slow because Cassandra has to search for the index in all the nodes.

This results in very slow queries (in the order of 1-2s).

For an exemple, see this tracing with a 3 physical nodes cluster with 256 vnodes on each.
The keyspace has a RF=3.

https://gist.github.com/16f510fc349584966a56

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira