incubator-cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Igor Katkov <ikat...@gmail.com>
Subject data distribution among DataFileDirectories
Date Tue, 29 Sep 2009 19:22:11 GMT
Does cassandra distributes keys evenly among DataFileDirectories?
If one have multiple hard-drives on a single host it makes sense to list
them all there.

I tried with two DataFileDirectories
  <DataFileDirectories>
      <DataFileDirectory>e:/cassandra-4/data</DataFileDirectory>
      <DataFileDirectory>c:/cassandra-4/data</DataFileDirectory>
  </DataFileDirectories>

and it ended up with all the data in the first.

Cassandra does write to both of them, but
1. Writes are not evenly distributed, at one moment there are 15Gb in the
first and 3Gb in the second
2. It continuously migrates data from the second to the first. So once
active writing stops and server is idle, the second DataFileDirectory
becomes empty

I suppose this is an effect from compacting process

Questions:
1. Is it by design?
2. Is there a way to control key distribution, for the cases when
hard-drives are of different capacity?

Mime
View raw message