Back at Oracle Open World 2008, Oracle gave some lip service to how they would get into cloud computing... in case you are not familiar with the term, "cloud computing" is a way of designing your systems so that your data resources (and sometimes your services) behave as if they are "in the internet cloud." Its a combination of a service-oriented architecture, software-as-a-service, and storage-as-a-service. Developers love it, but system administrators are still a bit weary...

Basically, you rent the computational power and storage you need, and only pay for what you use. In theory you can rely on your provider -- such as Google or Amazon.com -- to take care of backups for you. Its a great idea for startups (Twitter does it) and mid-sized companies, so they can keep costs down, while still leaving room to grow. For large companies with their own dedicated data centers, cloud computing makes less sense for production software... but its usually a great idea for development and testing.

Anyway... I was curious how Oracle's "Cloud" strategy would develop... and I was pleasantly surprised to find some recent collaboration between Amazon and Oracle. They put together some Best Practices for Oracle In The Cloud, which I found on Justin Kestylyn's blog:

I really like the idea of encrypting database backups, and storing them in the cloud. That's an excellent idea, for pretty much anybody... and it is supported back to Oracle Database 9i. Check out the Cloud Backup Whitepaper for more info...

I also really see the value for using the Amazon cloud for the persistence layer for archives. The Oracle Universal Online Archive could be a real killer app, but proving its value will need about a Terabyte of storage, just to do a proof-of-concept. Unfortunately, that's not exactly something you can run on a VM Ware virtual machine... but you could do it as an Amazon Machine Image (AMI).

I wouldn't be surprised if we saw more and more archiving solutions that use Amazon's Cloud for persistence...