November 10 2009 1:15 PM EST
Found this on Slashdot... Nice exposure, Jonathan!
Off to read it and the Slashdot thread!
Very cool, and an intriguing article to boot! :)
November 10 2009 1:50 PM EST
I am curious as to more examples, specifically, where relational databases fail... Not saying I have ever dealt with 50 TB of data, but I've dealt with 1 TB, and there was nothing there that couldn't have scaled with proportional cash and staffing... And that was using Visual Foxpro as the front end to a DBF file and SQL Server hybrid environment.
My big issue with "No SQL" is -- where is the business logic if not in a reliable data schema? Not to say rows of variable size can't still have a schema to them, but row/columns have been around for decades for a reason -- they map to business rules. Add programming and you've got process flow. A NoSQL approach would appear to put more logic in the programming, no? Not that there is anything wrong with that, but it is just exchanging one devil for another where segmentation of logic is still paramount.
Though, the article was more about scalability and uptime. So that takes me back to my first point, wondering where, exactly, conventional relational systems fail?
This thread is closed to new posts.
However, you are welcome to reference it
from a new thread; link this with the html
<a href="/bboard/q-and-a-fetch-msg.tcl?msg_id=002vLU">Check out this article by the Creator himself! *s</a>