On 8/22/10 12:00 AM, Wayne wrote:
Due to compaction being so expensive in terms of disk resources, does it
make more sense to have 2 data volumes instead of one? We have 4 data
disks in raid 0, would this make more sense to be 2 x 2 disks in raid 0?
That way the reader and writer I assume would always be a different set
of spindles?
Cassandra supports multiple data directories via specifying multiple
DataFileDirectory items in the DataFileDirectories block. Unfortunately
due to the way that compaction currently works, this often results in
one or the other drive being much hotter than the other and lots of
inter-disk shuffling of data.
https://issues.apache.org/jira/browse/CASSANDRA-897
Is a JIRA which refers to these issues. I am not aware of anyone who has
successfully improved their overall performance via multiple data
directories. I am, however, aware of people who have encountered
negative consequences as a result of doing so. As such, I do not
recommend it.
=Rob