Re: Lies and Urban Myths
Amazon absolutely are on that list - they provide many different data centers for failover and continuity and provide multiple tools to enable that. Yes, it's more expensive to deploy this way and, yes, it's more complicated. The blame for the failures of applications using AWS is 100% on the customers who cheaped out and didn't build resiliency into their deployment architectures.
Amazon have some lessons to learn about their own status and notification tools and making sure they operate during an outage, but overall this has been a valuable lesson in identifying those companies that are too stupid or miserly to use the proper tools available to them. Plus some truly shocking deployment decisions - web fonts stored on S3? Config files for a mouse? Complete failure of security systems when one AWS node goes down? FFS, are all your developers 19?