The load contains duplicate data which is created due to compaction. Run 'cleanup' command with nodetool to those big nodes and you should see the load drop to the actual usage.
- Garo On Thu, Nov 4, 2010 at 11:08 AM, Mark Zitnik <mark.zit...@gmail.com> wrote: > Hi All, > > I'm having a problem in spreading data across the cluster. > > my replication factor is 3, please advice why there is a big difference > between 10.11.40.239 and 10.11.40.161. > > Thanks > > Address Status Load > Range Ring > > 10.11.40.173 Up 220.58 MB > 5488375922431206513329053672302981681 |<--| > 10.11.40.248 Up 79.03 MB > 21834979508285328193838231992126539188 | ^ > 10.11.40.157 Up 79.48 MB > 26827287656215870496305624629770425714 v | > 10.11.40.159 Up 70.28 MB > 47864683020405054074376652399555662273 | ^ > 10.11.40.160 Up 71.45 MB > 52127237413510596835396555869135415868 v | > 10.11.40.171 Up 70.59 MB > 58686864619218705531024294466626882735 | ^ > 10.11.40.241 Up 126.81 MB > 97920742222079324486086191292998090436 v | > 10.11.40.239 Up 434.14 MB > 143147210934849238702354354634298482182 | ^ > 10.11.40.161 Up 288.68 MB > 155148009923170935299284525080785995290 |-->| > >
