In this video Scott Laningham of IBM developerWorks interviews Scott Chate from Corent Technology about their use of Amazon's cloud for building SaaS solutions and how WebSphere, DB2 and pureXML play a useful role. A couple of months ago Scott C. also participated in a DB2 Chat with the Lab webcast on Cloud Computing, a recording of which can be found on ChannelDB2. And if you want to see things in action, check out this screencast of how you can quickly whip out a SaaS app in minutes using SaaSFactory and DB2 AMIs.
Aug 3 and 4 were holidays for us, so while clearing up my Inbox I was surprised to find a thank you note from Ken North from a couple of days ago. Ken recently did a comprehensive write-up titled Databases in the Cloud: Elysian Fields or Briar Patch about the plethora of choices available for storing and managing data in the cloud. Ken covers everything from simple key-value pair data stores to industrial strength databases like DB2. Something that Ken writes and I very much agree with:
The SQL database has survived every paradigm shift critics said would be the death of SQL, including object-oriented programming (OOP), online analytical processing (OLAP), Internet computing and the World Wide Web. Some have suggested SQL platforms are not sufficiently scalable for large workloads or data volumes, but there's ample evidence to the contrary. The UPS shipping system central database processes 59 million transactions per hour. It has a table that contains more than 42 billion rows has achieved a peak workload of more than 1 billion SQL statements per hour with IBM DB2.
And now with a paradigm shift to Cloud Computing, you can be sure that SQL databases will evolve, thrive, and become even more relevant for large scale data managment. How they will evolve to meet the new demands of cloud computing will be a topic for a new blog post, but one aspect that I feel is pretty important in the evolution of relational database technology is dynamic elasticity. Being able to scale a database from a few to many virtual nodes in the cloud on the fly, and then back down to a few nodes when the peak demand subsists. We're already seeing technology that is making database elasticity possible - for example xkoto Gridscale and DB2 ... watch this video to hear David Tung of xkoto talk about it.