Getting started on Amazon's Cloud?

I may have a very large-scale computational task coming up in the next few months (Intensive processing of ~400 GB of data, probably in the few-thousands of CPU-hours). I am likely to do this either on a cluster at my school or Amazon's cloud computing service. It's a machine learning task, so it will be generally more of a scientific nature than software/business.

I have done more than trivial parallel programming. What are good resources to learn about Amazon's cloud service and learn how to effectively harness 开发者_Go百科it so I don't waste my money? More generally, what are good cloud computing resources?

---

If you're just getting started, I would research EC2 and S3 - they're pretty much the intro-level services that Amazon offers. They have a free tier if you're making a new account, but if you want to mess around with a few machines and not worry about racking up a big bill while you're still testing your code, Eucalyptus offers an API-compatible version to Amazon's EC2 and S3 that you can mess around with. They normally offer software you can install on your own boxes, but it looks like in your case you may not want to install it on your own boxes, so you can use their Community Cloud to test everything out for free.

---

Some of the useful stuff that get you started with Amazon Web Services:

http://paulstamatiou.com/how-to-getting-started-with-amazon-ec2

Some interesting facts on EC2 & SQS - for building a scalable application:

http://sqs-public-images.s3.amazonaws.com/Building_Scalabale_EC2_applications_with_SQS2.pdf

A whole bunch of useful information for EC2, EBS, S3, SQS,...

http://kenneth.kufluk.com/blog/2010/03/getting-started-with-amazon-ec2/