Recent versions of Cassandra (I noticed it in RC1, which is what I am using) add a very cool "estimateKeys" JMX operation for each column family. I have a quick question: Is the value an estimate for the number of keys on a particular node or for the entire cluster?
I have a 5 node cluster (RF=2) and am getting wildly different estimates from different nodes. For example, one says ~50 million and another ~75 million. Dan Hendry (403) 660-2297
