I am not sure if it is only my impression, but it looks like MongoDB is getting a lot of presentations and video coverage. And the champion seems to be Mike Dirolf (@mdirolf) from 10gen. Below is the freshest presentation he gave on MongoDB at ZPUGDC monthly meetings (a Python group) just a couple of days ago:
the part about MongoDB replication and auto-sharding is really interesting
As a side note (and a question for MongoDB people), I keep hearing for quite a while that auto-sharing is still in alpha, so I am wondering what’s the real reason. Have the current implementation hit a dead end? Or aren’t there enough users asking for it? Or what is it? Mike?
Difference between OODB and document database:
- they use fairly similar concepts
- in OODB you are saving instances, in document databases you are saving data
That basically translates to the fact that data and objects are decoupled and this comes with both pros and cons.
- no concurrency in the stable 1.2
- much better concurrency in 1.3
I will continue to add my notes as I watch the presentation (the video reports 1h20’+).
Update: Here is another confirmation of the thought I’ve expressed at the beginning of the post. Below you can find the presentation Kyle Banker (@Hwaet) gave the other day to ChigacoRuby.