LegitimateSounding.com

When needing two or more fairly disparate systems to work together seamlessly, having complete flexibility at the database level can be a blessing.

Take for instance the problem of a ten year old legacy system hosting millions of accounts, and an up to date content management system that needs complete access to that data as if it were its own. You can manage multiple systems with complicated triggers, methods for moving data around, expensive joins, funky stored procedures, hacks to the code, or you can simply use a writable view.

SQL is everywhere. Believe it or not there are legacy relational "schema-with" databases filled with data all over the internet. Chances are even your own office has at least one SQL database lurking in a closet somewhere.

So, how do you leverage your existing "schema-with" databases and still be able to use the power of Map/Reduce? Introducing MR SQL: A Map/Reduce Front-End to SQL.

Often times, I don't get to get my hands dirty at work. Not being one to let myself atrophy, I keep my eyes out for new and exciting things to catch my fancy, and spend hours and hours writing new code: usually reinventing the wheel, often times poking and prodding, just trying to figure out what I'm going to do with what I find.

One of the projects that caught my eye a bit over a year ago was CouchDB, a RESTful document storage engine, that happens to have Map/Reduce support. Being the database freak that I am, I started thinking about all of the projects I've worked on in the past that could have been improved with a document model over pseudo-relational databases. So many came to mind, and I was excited about the flexibility of CouchDB; so useful for so many things, especially with strong data analysis abilities via map and reduce.