I did this because we needed a Jena on Cassandra at work and I got permission to open source the code.
As far as benchmark, no benchmarking was done. SDB is only in maintenance. Moving to a column store is a rather different beast than relational model so I would expect that native column store would be more efficient. But I have not tested. Claude On Sat, Jan 7, 2017 at 3:14 PM, Nicolas Paris <nipari...@gmail.com> wrote: > Le 05/58/2017 ą 21:58, Claude Warren écrivait : > > I posted this on the Dev list but ment to post it here. Sorry for the > > x-post. > > > > Jena on Cassandra (https://github.com/Claudenw/jena-on-cassandra) now > has a > > complete Graph implementation and passes all the contract tests. > > > > It is now possible to use a Cassandra cluster as the storage layer for > > Jena. > > Hello Claude, > > This is a very interesting project. Why did you choose cassandra ? Have > you made some benchmarking before going with this distributed database ? > I have been say Aple uses 100k cassandra server cluster. > FYI postgresql do have a FDW (a way to CRUD over Cassandra cluster). > I wonder if the old SDB -> postgresql -> cassandra would work too > > > > > > HOWEVER -- > > > > > > 1. This is very early code on not ready for prime time. > > 2. There is no assembler so it will not be possible to use with Fuseki > > yet. > > 3. There are no SPARQL custom execution strategies for performance. > > > > Please feel free to try the code and let me know what you think. > > > > Claude > > > > -- > > I like: Like Like - The likeliest place on the web > > <http://like-like.xenei.com> > > LinkedIn: http://www.linkedin.com/in/claudewarren > -- I like: Like Like - The likeliest place on the web <http://like-like.xenei.com> LinkedIn: http://www.linkedin.com/in/claudewarren
