What Scenarios and Architectures for BIG + Fast Data in the Cloud?

[Adam Bloom]

What key scenarios (or use cases) have have you seen or dealt with for BIG AND FAST data in the cloud?

And, what key architecture approaches have you seen at the data layer for these scenarios?

For example, live market feeds on wall-street have had large data sets and required fast access...and internet search is all about big data and real-time search makes the processing need very fast.  These systems have very specific needs and architectures.  Could you share what you have run across?

We are hoping people in the group can share links or direct, real-world information about these types of solutions to become a decent resource on the topic.

We've searched around for such a resource (list of scenarios and architectures for BIG/FAST needs) and haven't found a great one...please share if you know of one.

The more detail, the better....Thank You!

Here are a few example links on the topic where interesting use cases and architectures are shared:

http://www.zdnet.com/blog/big-data/fast-data-hits-the-big-data-fast-lane/309 

http://highscalability.com/blog/2011/11/29/datasift-architecture-realtime-datamining-at-120000-tweets-p.html 

http://instagram-engineering.tumblr.com/post/13649370142/what-powers-instagram-hundreds-of-instances-dozens-of 

http://blogs.vmware.com/vfabric/2012/08/how-big-data-meets-fast-data-in-the-cloud.html (disclosure - we also worked on this article)