Thanks Jonathan. Usually on such projects I work from the trunk. The current piece we've been working on uses the lucene on cassandra (lucandra) code, which supports v.4 of cassandra. Jake i believe is working on upgrading.
Our company is currently in an experimental stage trying different methods for indexing large amounts of documents. For the most part lucandra is working great (quick setup, good query times), but the initial load is killing us. I've been talking with a few co-workers to maybe try indexing using the traditional method of using supercolumns in cassandra as did facebook (that way we can use the bleeding edge cassandra code). I was curious if anyone has done benchmarks on insert using the old method vs using the Cassandra-336 multi-key batch code? -- View this message in context: http://n2.nabble.com/Status-of-CASSANDRA-336-tp4468535p4468652.html Sent from the cassandra-user@incubator.apache.org mailing list archive at Nabble.com.
