Just a quick note to draw people to an excellent blog post about NoSQL databases and "Partition Tolerance"
. Emun Sirer points out one of the many reasons why "webscale" doesn't mean what you think it does, and how not all new scalable databases are the same. He doesn't say which specific new database is marketing features they don't really have, but I think we can guess, can't we?
I was intrigued by this paragraph from his response to a response to the post you linked:ReplyDelete
Perhaps some view NoSQL as a niche solution that is forever limited to non-demanding applications. I disagree with this and believe that NoSQL has the potential to supplant RDBMSs and capture the bulk of the database market. The "daddy knows best" attitude that RDBMSs bring to data management, the way they strip all information that the developer had about her data and force her to write a declarative specification of what she wants, only to try to then come up with an efficient evaluation plan in the query optimizer, all reflect a klunky aesthetic to system design that the lean and mean NoSQL movement can and will supplant. NoSQL is to RDBMSs what Unix was to Multics. But if we are to see this happen in our lifetimes, we should not hesitate to think big. Why shouldn't it be possible for NoSQL to tackle the big boys and serve as the sole data store for Ticketmaster or EBay or Bank of America or Goldman?
You can't fault the guy for ambition, that's for sure. And if it leads him to build a better non-relational data store, that's a good thing. Note, though, that he defines "NoSQL" as "any database which scales to dozens or hundreds of servers" in another of his posts, so from that perspective, he's probably right.