Oh, and btw, that's not 2 billion indexed rows per-node, that's 2 billion rows that have the *same* index value per-node (each index value is a separate row in the index).
On Fri, Mar 8, 2013 at 8:29 AM, Sylvain Lebresne <sylv...@datastax.com>wrote: > Not exactly. Each replica only indexes the rows that it is a replica for. > So there is a limitation currently indeed, but it is of 2 billion indexed > rows per-node. > > > On Thu, Mar 7, 2013 at 8:05 PM, Edward Sargisson <esa...@pobox.com> wrote: > >> Hi, >> Please correct me if this statement is wrong. >> >> Secondary indexes are limited to indexing 2 billion rows - because they >> turn a row into a column and C* has a limit of 2 billion columns. >> >> Cheers, >> Edward >> > >
