Werner_Vogels: Good morning or afternoon depending on where you're located. Thanks for joining. I'm looking forward to answering as many of your questions as possible in the next hour.
Peter K: Where do you see Amazon fitting in or in what roles for the future of Information Everywhere availability. What are some of the future tracks and R&D the company might take?
Werner_Vogels: At Amazon we believe that we can offer the data in the Amazon platform through Amazon e-commerce service to allow access from any application. The mixing and mashing of different services to build new businesses is the economic trend of the future and EC2 (Elastic Compute Cloud) and S3 (Simple Storage Service) also fit in to this. Go read the push and pull paper by John Hagel and Seely Brown on this.
Nick: Are you seeing interest/adoption of EC2 (Elastic Compute Cloud) and S3 (Simple Storage Service) in 'traditional' high performance and grid computing areas? If yes, what areas are taking advantage of those services?
Werner_Vogels: We see adoption of Amazon EC2 and Amazon S3 in a broad spectrum of companies. One particular area is the traditional high performance computing field such as financial industry, pharmaceuticals and movie rendering.
Ironman: I have been curious about the Amazon EC2 business model. Can you give us an idea of the success of that model and the types of tools you have used to capture usage information and allocate costs? How applicable are those tools to internal enterprise cost allocation?
Werner_Vogels: EC2 and S3 were technologies that were developed for Amazon internally first and as such they have proven their success in an enterprise environment. We have an extensive internal monitoring and scheduling system that optimizes EC2 usage but it is very early and very research focused.
EC2_Fan: Do you plan to offer high availability and SLAs around your S3 and EC2 services?
Werner_Vogels: Regarding high availability, S3 and EC2 are built on the same technology that Amazon.com uses. With respect to SLAs, this is something we hear a lot from developers and we hope to have something to share with you soon. One more point on SLA's. They need to be meaningful in an Internet context. We are diving deep on making sure that is the case.
Moderator-Keith: Here's a pre-submitted question: Any Harry Potter VII stories? What tricks did you use to keep systems handling massive worldwide orders and deliveries?
Werner_Vogels: We often talk about how well Amazon scales, but most of those discussions are how Amazon provides rock-solid performance with respect to the number of customers, merchants, visitors, etc. HP demonstrates another side of Amazon that scales really well and this is shipments and delivery processing. To handle the additional processing that the HP scale requires we use EC2 to dynamically acquire and release capacity and our systems are built to exploit these elastic techniques.
jason_watkins: Smaller Web sites experience similar pains to larger Web sites, but we don't have the engineering or operational resources to build our own distributed computing frameworks. Things like EC2 help, but the market is really underdeveloped. Do you have any ideas for approaches that smaller companies could use to gain some of the benefits of Amazon's in-house technology, but with less daunting staff requirements?