If The Cloud Were Really The Cloud Then Outages Would Be Rare

Should we go back to calling the cloud hosting and SaaS?

Do we really have a cloud or just offsite Microsoft, Cisco and Linux hosting centers? I'm struck by what seems like a near constant barrage of news about cloud outages and service downtime. From one week to the next it's Google's Gmail or Microsoft's Hotmail unavailability, Rackspace's service credits, Twitter hitting the skids on a near weekly basis, Microsoft's Sidekick data loss, Apple's iTunes servers cratering during upgrades, Blackberry's network downtime, and Salesforce outages. The list goes on and on, outage after outage, and those are just the easy ones to name.

It used to be that an outage meant someone cut into a copper bundle or fiber network somewhere causing local and even regional Internet outages. Our issues seem to rarely be network outages and now are instead cloud and online service outages. Cloud service outages are now worthy of CNN and even local news coverage.

So I have to ask the question, isn't the cloud supposed to be more resilient and expandable than just another offsite data center or hosting solution? What happened to nearly infinite capacity, computing on demand, storage by the drink, virtualized systems that could be brought up and down in a matter of minutes, even seconds? Have we been sold a bill of "cloud seeding" goods, or are we just happy drinkers of cloudy Kool-Aid?

Like everyone else, I've experienced outages such as Salesforce's service or Blackberry's network being down. But the eye opener for me was one weekday morning this spring when Amazon's EC2 service ran out of server capacity and wouldn't let me start up a new virtual instance of my servers because they were out of capacity.Whoops, I thought that wasn't supposed to happen. Wasn't that the whole idea of computing on demand? Never worry about running out of capacity or buying more servers. Just start up another virtual server instance in the "elastic computing" cloud. I guess that cloud's not as elastic as we were led to believe. (See my blog posts Sorry...The EC2 Cloud Is Full, Come Back Later and Will The Cloud Manage To Persist? Lessons From Operating In The Cloud.)

So, back to my question. Are we really going to have a cloud or should we just go back to calling it hosting and SaaS? If the cloud has no more resilience, reliability or capacity to take on dynamic loads than your typical datacenter, then it's not a cloud. It's a datacenter who's exact location is a little cloudy. I understand that the cloud is also about the services it provides to us, but without the capacity to bend, stretch and react to demand, we're not going to be able to fully depend on service availability, evidenced by what typically are pretty weak SLAs from most cloud services today.

I hope we're just in the awkward teenage stage of cloud services where every once in a while things get a bit goofy, until everything has a chance to mature a bit more. I hope so. If we can't solve the services outage and data loss problems of the cloud, it may end up being called the black cloud.

Like this? Here are some of Mitchell's recent posts.

Recent Podcasts:

Mitchell's book recommendations: Also visit Mitchell's other blogs and podcasts:

Visit Microsoft Subnet for more news, blogs, opinion from around the Web. Sign up for the bi-weekly Microsoft newsletter. (Click on News/Microsoft News Alert.)

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.

Copyright © 2009 IDG Communications, Inc.

SD-WAN buyers guide: Key questions to ask vendors (and yourself)