It's a non trivial, but you could try using haddop/pig. Take a look at contrib/pig in the source.
You could output flat file formats. Aaron On 17 Jul 2010, at 02:09, Sylvain Lebresne wrote: > On Fri, Jul 16, 2010 at 3:53 PM, xavier manach <x...@tekio.org> wrote: >> Hi. >> >> What is the best method for make large extract of datas from cassandra ? >> Extract directly from the SSTables files sound good ? >> If yes, >> There is an API for exploit directly the SSTables files ? >> There is a specification of the SSTables files ? > > Not really, but you can have a look at the sstable2json tool. You may not be > able to afford converting to json, but it shouldn't be too hard to > modify it for your > needs if you are motivated enough. > > -- > Sylvain
