On Tue, Dec 15, 2009 at 12:17 PM, Baz <[email protected]> wrote: > Interesting articles Vince. If I did some of those queries in a relational > db, I'd be fired on the spot :) Do you know if these queries > happen concurrently or sequentially behind the scenes? Given, that it's in > the cloud of infinite resources, and they were run concurrently, the only > big issue could be cost, and at least not performance. Well also the fact > that there is a limit of 30 subqueries per request, and to do an IN, a > subquery is run for each term (then merged) meaning you can't have more than > 30 terms in an IN statement. > > The other thing that comes into play here is that since the datastore isn't relational, trying to use it as if it where probably isn't the right approach. I run into this a lot as I've been using CouchDB. It's a completely different way of storing data, so reworking applications that use the datastore to adhere to a more document-centric approach will likely yield better results, not to mention fit better within the limitations of the datastore.
Just some additional food for thought. -- Matthew Woodward [email protected] http://mpwoodward.posterous.com identi.ca/Twitter: @mpwoodward Please do not send me proprietary file formats such as Word, PowerPoint, etc. as attachments. http://www.gnu.org/philosophy/no-word-attachments.html -- Open BlueDragon Public Mailing List http://www.openbluedragon.org/ http://twitter.com/OpenBlueDragon mailing list - http://groups.google.com/group/openbd?hl=en !! save a network - please trim replies before posting !!
