Cloud Gloating

virginia_storm
The deluge of negative press started within hours of the June 29th Amazon outage. The articles ranged from a general negativity around the readiness of the cloud, to competitors using the event to promote their services, to a general disparagement of companies relying so heavily on AWS (e.g. Netflix, Pinterest). A faintly detectable theme behind some of the voices is what I call cloud gloating: a satisfaction at seeing a company and technology so publicly fail.

We’ve seen this before: most notably after the April 2011 AWS outage. Amazon specifically and the cloud generally have many enemies: competitors looking for market share, IT professionals fearing for the future, journalists looking for a story. But beneath all the negativity is a simple and undeniable truth: technology can and will fail. From the first failure of the ENIAC to the latest outage at Amazon, no vendor or technology can achieve complete reliability. With complexity comes risk, with risk comes the potential for failure, with potential failure comes failure.

I am not an apologist for Amazon. In fact, I harbor some minor resentment over the incident management and communication of the event. But I do feel the need to say that cloud gloating does not promote any of our long term interests. In spite of this outage, the cloud remains more reliable than most alternative solutions. It is far superior to corporate IT in reliability, performance and security for many use cases today, and will be for many more in the future. I also am confident that Amazon will learn from this incident in the same way the April event resulted in the improved reliability and performance of EBS.

So for those of you doing the cloud gloating, I can say only this: be humble, since you may be one incident away from losing that smirk on your face. For the rest of us: let’s pick ourselves up, dust ourselves off, and continue the long and steady march into cloud.


The Cloud Ambulance Chaser Award goes to Whatsyourprice.com, who managed to use this outage to gain outsized publicity for its move of 10 instances (yes, you read that right) off of AWS.