What to Do When the Cloud Comes Crashing Down

By Mike Elgan  |  Posted 2015-09-23 Print this article Print
Cloud Crash

These failures were caused by a flaw in Fujitsu's system for handing off to the backup power supply.


Nest, which is owned by Google, reported on Sept. 7 the unavailability of the cloud services that support both their Nest Thermostat and Dropcam camera products. Users were unable to use the cloud services for the thermostats or see the video for their Dropcams during a three-hour outage. It was the second major outage in a single week.

The outages for Dropcam were a serious concern for some, as the cameras are often used for security or as baby monitors. It also raised the question of whether using cloud-based cameras for security is really a good idea.

Cloud Service Crashes Are Always a Concern

Because all these services are so significant, and because these outages happened within the span of a month, it's worth taking a moment to consider this strange place we've come to.

Cloud computing has become so commonplace that we find ourselves using services that we can forget are cloud services. When people enjoy the proper, automatically set temperature in their homes with a Nest thermostat, for example, they tend not to think: "I'm doing cloud computing." The recent outage serves as a reminder.

I think there are three takeaways from all this.

1. There's nothing magical about the cloud. "The cloud" is just somebody else's computers located somewhere else. All the problems that exist in one's own data-center can exist within the cloud services.

2. The cloud comes with a certain degree of helplessness. One benefit of the cloud is detachment. A large number of people are sweating bullets every day to keep cloud services up and running as best they can, but when things go sideways, there's usually nothing that subscribers can do about fixing it. This brings me to my most important point.

3. It's important to always have a cloud Plan B. That could be an alternative cloud service. It could be localized access of one's data in a non-cloud service. Or, as in the case of the Skype and Twitter outages, the alternative could be using a different facet of the service.

For Skype, using the Web version was an instant way to continue using the service. For Twitter, a third-party client would have been a good alternative to not using Twitter at all during the outage.

In the meantime, let's cross our fingers and hope that the past month isn't an indication of the reliability of the cloud in general. It's probably just a coincidence that all these big outages happened within a month of each other.

Either way, it's always best to make sure you're ready with a Plan B for each cloud service you're using.


Submit a Comment

Loading Comments...
Manage your Newsletters: Login   Register My Newsletters

Rocket Fuel