Rackspace outages teach cloud lessons

Clarifying issues surrounding this emerging security architecture

Back-to-back power outages at cloud service provider Rackspace offer some lessons that cloud providers and their customers can benefit from.

For customers it speaks to the necessity of checking out service providers’ infrastructure for redundancy. Rackspace says it had backup generators and batteries in place, but failures of one or the other resulted in outages to some customers anyway.
Given that this type of outage does happen, customers should carefully weigh exactly what services they entrust to the public cloud. Applications and data that are critical to running their businesses should either not be placed there or there should be a separate backup that kicks in when the cloud fails.

Keeping applications out of the cloud and hosting them privately presumes that businesses can serve up better reliability than a cloud service provider can. Doing so can be more costly than using a cloud, and customers have to perform a risk assessment to decide if it makes sense.

To continue reading, register here and become an Insider. You'll get free access to premium content from CIO, Computerworld, CSO, InfoWorld, and Network World. See more Insider content or sign in.

Back-to-back power outages at cloud service provider Rackspace offer some lessons that cloud providers and their customers can benefit from.

For customers it speaks to the necessity of checking out service providers’ infrastructure for redundancy. Rackspace says it had backup generators and batteries in place, but failures of one or the other resulted in outages to some customers anyway.
Given that this type of outage does happen, customers should carefully weigh exactly what services they entrust to the public cloud. Applications and data that are critical to running their businesses should either not be placed there or there should be a separate backup that kicks in when the cloud fails.

Keeping applications out of the cloud and hosting them privately presumes that businesses can serve up better reliability than a cloud service provider can. Doing so can be more costly than using a cloud, and customers have to perform a risk assessment to decide if it makes sense.

Cloud providers should take note of the immediately apparent impact on Rackspace. Its stock dropped 4% after one of the outages. The public learned of the outage, which has to have an impact on the company’s reputation. The company won’t know for some time whether it loses customers over the outages, but that is something for other providers to watch.

Rackspace showed good sense in providing details of what happened, the extent of the outages and the steps it was taking to prevent recurrences. All of these measures can go a long way toward restoring confidence in the company. People are willing to accept failures if they can understand the causes and determine that they were not caused by recklessness.

Read more about security in Network World's Security section.

Tim Greene is senior editor at Network World.