More than 20 years ago, Amazon began its journey to become the best online and offline retail player in the world. Ten years ago, Amazon Web Services began its journey to become the number one public cloud provider. And this week, Amazon suffers a breakdown that slows down the hard-won success of years of work. Its sales site and cloud collapse under the weight of the 2018 Prime Day event.
Overwhelmed by the load of millions of enthusiastic buyers, Amazon may have known yesterday its worst day in terms of operations. Instead of shopping online, platform customers are faced with errors that prevent them from validating their order.
Amazon did not say why this situation occurs. One sarcastic user suggested that “Amazon.com should not follow the recommendations of the Well-Architected Framework Guide, which is the title of the manual published by Amazon on how to build cloud applications.
An alternative is possible
Amazon uses the services of AWS since 2011. As explained Jon Jenkins, then head of engineering at Amazon, “We started to think in 2009 the way to use [AWS] to Amazon.com .. November 10, 2010 is the day we turned off the latest physical Web server serving Amazon.com. ”
Unfortunately, even with years of experience, AWS did not rise to the challenge this week. Amazon needs to repair its cloud. My own analysis, based on what I’ve seen, is Amazon’s front-end caching and AWS Elastic Load Balancing (ELB) have not been up to the task of dealing with these millions of simultaneous buyers.
You can mitigate problems with Amazon by using the Amazon Smile web alternative interface . This site is for users who wish to donate a portion of their sales to a charity of their choice. In addition, if you have a direct link to a product, such as those listed below, this should work as well. An Amazon spokesperson told ZDNet: “Some customers have difficulties making their purchases, and we are working to solve this problem quickly.”
Prime Day is one of Amazon’s biggest events of the year and Amazon has blown it up
On AWS, when I tried to access the AWS Management Console , guess what I found? Go ahead, guess. Yes, it’s true: the same error page as Amazon’s. “We are currently experiencing intermittent errors when using login credentials to the AWS Management Console,” Amazon admitted yesterday. The AWS console is used to control AWS resources. Amazon also claimed that the underlying AWS services and console connections using IAM users are functioning normally.
Since then, Amazon has updated the message. It now states, “We have identified the root cause of intermittent errors on access to the AWS Management Console when using login credentials, and we continue to work toward a complete resolution of the problem.”
In the end, Amazon will make everything work again. This is not the real problem. The real deal is that Prime Day is one of Amazon’s biggest events of the year and that Amazon blew it up.
As my colleague Jason Perlow sarcastically put it , “If only Amazon had a huge amount of resiliently scalable data centers with georedundancy, if only!”
This will not be enough for loyal Amazon or AWS customers to move elsewhere to purchase cloud products and services. But he points out in red ink that even Amazon can make huge technological mistakes.