This is true, but you'd seem to have a very flat schema where consistent holds within, not between (apart from unique keys) records, and where concurrency and transactionality is required at record level.
Anyway, just a thought. Barry On Dec 8, 2013 1:13 PM, "Michael Brunnbauer" <[email protected]> wrote: > > Hello Barry, > > On Sun, Dec 08, 2013 at 12:03:36PM +0000, Barry Norton wrote: > > Actually, I'd go one step further and ask that, since you have a fixed > > record-oriented schema and don't need to query a lot of joins as far as I > > can see (only foaf:knows?) > > The foaf:knows relations are not stored in the RDB. > > >, do you need a relational DBMS even? > > Couldn't you easily put the whole record into a NoSQL DBMS or, since you > > concentrate on FTS, just index them in Lucene/Solr? > > Yes - that would be possible and is an option for the future. But we have > much > more experience with RDBs and the point where we run into scaling problems > with the RDB is far away. > > One feature of a RDB that we currently do not use but that could be useful > are > transactions: Inconsistencies are improbable but possible as several > separate > components modify the database simutanously. Avoiding inconsistencies in > a NoSQL DBMS is much harder as I recall. > > Regards, > > Michael Brunnbauer > > -- > ++ Michael Brunnbauer > ++ netEstate GmbH > ++ Geisenhausener Straße 11a > ++ 81379 München > ++ Tel +49 89 32 19 77 80 > ++ Fax +49 89 32 19 77 89 > ++ E-Mail [email protected] > ++ http://www.netestate.de/ > ++ > ++ Sitz: München, HRB Nr.142452 (Handelsregister B München) > ++ USt-IdNr. DE221033342 > ++ Geschäftsführer: Michael Brunnbauer, Franz Brunnbauer > ++ Prokurist: Dipl. Kfm. (Univ.) Markus Hendel >
