Charles > Reading the abase website, Cassandra sounds like an amazing tool. > Although, a few more features than we are interested in, their philosophy > sounds... sound. However, I am skeptical of their claim: > > "Generating statistics for very large and heavily used tables with > Oracle's analyze command can take anything up to several hours depending > on the size and usage of the table being analyzed. With Cassandra, those > statistics can be calculated within minutes." > > How can you do that without sampling? And once you start sampling, you > introduce the possibility of error. On large tables there's a very easy way to do it without sampling: parallelization. Cheers, Chris -- //www.freelists.org/webpage/oracle-l