VISITOR ANALYSIS
Referrer No referring link
Host Name ec2-184-73-42-139.compute-1.amazonaws.com
IP Address 184.73.42.139 [Label IP Address]
Country United States
Region Louisiana
City Denham Springs
ISP Amazon.com
Returning Visits 0
Visit Length 0 seconds
VISITOR SYSTEM SPECS
Browser Default Browser 0
Operating System unknown
Resolution Unknown
Javascript Disabled
Navigation Path
Date Time Type WebPage
17th June 2010 15:52:19 Page View No referring link

And another hit from a few days ago:

VISITOR ANALYSIS
Referrer No referring link
Host Name
IP Address 174.129.185.245 [Label IP Address]
Country United States
Region Washington
City Seattle
ISP Amazon.com
Returning Visits 0
Visit Length 0 seconds
VISITOR SYSTEM SPECS
Browser Research Projects 0
Operating System unknown
Resolution Unknown
Javascript Disabled

But did he really? Here’s the quotation from the article: “Selipsky defined an availability zone as two physical locations that would not go down under the same disaster scenario. ‘It’s not axiomatic,’ he said, but broadly true that Availability Zones are located in different data centers.”

Not axiomatic but broadly true? How do we parse that statement? It sounds to me as though at least two of the AZs are in the same data center, or else they would just say definitively that they’re all in separate facilities, instead of invoking these linguistic hedges. I’ve tracked Amazon’s statements on AZs, and this seems consistent with past statements that emphasize that each zone has dedicated infrastructure, rather than their physical location

http://searchcloudcomputing.techtarget.com/news/article/0,289142,sid201_gci1359572,00.html

Amazon has announced S3 services in Europe, where regulatory issues on data storage limit use of out-of-market backup and DR services, but I haven’t yet seen an EC2 zones for Europe. I recall that there were requests for this when they announced the three East Coast zones.

The tougher question: are those three us-east zones in three separate physical data centers, or has Amazon segmented multiple pods within one data center? The Amazon Web Services blog says that “each zone is designed in such a way that it is insulated from failures which might affect other zones within the region. By running your application across multiple zones within a region you can protect yourself from zone-level failures.”

There’s slightly more detail in the developer documentation: “Each availability zone runs on its own physically distinct, independent infrastructure, and is engineered to be highly reliable. Common points of failures like generators and cooling equipment are not shared across availability zones, and availability zones are designed to be independent with failure modes like fires and flooding.”

Hmmm … more info, but doesn’t necessarily rule out pods with dedicated generators and cooling. Do their partners know any more? Apparently even RIghtscale doesn’t. From their blog: “I don’t have inside information about the location of their facilities, but I imagine some may be in New York and others may be in Virginia, so the distance between zones may be considerable, thus translating into some network latency.”

That would align with today’s announcement that there are CloudFront nodes in both Ashburn and Newark. One possibility: the three east coast zones are split between the two data centers. That would provide geographic diversity, but also explain why Amazon doesn’t just say “all zoines are in different data centers.”

Not sure if that helped, but that’s what we know.