With random partitioner, the rows are sorted by the hashes of the keys, so for all intents and purposes, not sorted.

This comment below really is talking about how columns are sorted, and yes when time uuids are used, they are sorted by the time component, as a time uuids start with the time component and then adds various randomness bits.

On 03/07/2012 01:51 AM, Tamar Fraenkel wrote:
Hi!
I am currently experimenting with Cassandra 1.0.7, but while reading http://www.datastax.com/dev/blog/schema-in-cassandra-1-1  something caught my eye:
"Cassandra orders version 1 UUIDs by their time component"
Is this true?
If I have for example USER_CF where key is randomly generated java.util.UUID (UUID.randomUUID()), will the rows be sorted by the generation time?
I use random partitioner if that makes any difference.
Thanks,



Tamar Fraenkel 
Senior Software Engineer, TOK Media 





Reply via email to