9 Spectacular Cloud Computing Fails

Over the past 18 months, IBM has secured 1,200 patents on cloud computing, including about 400 in the first half of 2015 alone.

One is about scaling down a virtual machine as its traffic recedes, another deploys sensitive data to a secure server, and a third creates snapshots of virtual machines for rapid recovery in the event of a failed workload.

These examples don't necessarily bring to mind a sense of blinding brilliance or original innovation, but these cloud operations can be patented. For those who conceive of the cloud as an environment based on public standards with many shared elements, the grant of these patents isn't entirely reassuring.

IBM, one of the world's largest patent-holders, year after year, and a leading patent acquirer, is busy applying the skills of its stable of patent lawyers in order to gain more cloud patents.

[Want to learn more about the purpose of the patent system? See Patent Debate Spawns Stanford Study.]

We had hoped this round of computing would leave behind highly proprietary components and fiercely fought-over intellectual property. It was going to be based on common-denominator components, Internet and Web service standards, etc., so minimizing the need for proprietary moving parts.

What happened to the sharing economy? Is it here, but won't be based on much sharing when it comes to cloud infrastructure?

Granted IBM, with its Bluemix developer platform-as-a-service or its Softlayer cloud data centers, might have specific inventions and innovations that IBM wishes to protect. But the nature of IBM's patents cited doesn't appear to be particular to Bluemix, Softlayer, or any other IBM-originated service.

Rather, they appear to be more general-purpose cloud processes or operations that many service providers already use or would implement if left free to do so. IBM's patent library might therefore be used to demand royalty payments from other service providers or, perhaps equally likely, be used to ward off claims made against IBM by aggressive patent-holders among the service providers. 

Whatever the intent, these patents illustrate how the cloud, even though it's conceived of as a shared environment following public standards, may be subject to some of the same intellectual property disputes and patent trolling as earlier, more directly proprietary environments.

Are we at the beginning of the cloud patent wars? Let's hope not.

But if you think it's common knowledge that a cloud workload can be managed "elastically" to meet demand -- scaled up to meet increasing traffic, then scaled down again -- it may come as a surprise to you that IBM holds a patent on allowing a virtual machine to scale down again as high priority traffic diminishes.

The US Patent and Trademark Office has seen fit to issue Patent 8,949,415, which allows a cloud service to provide a virtual machine with fewer resources after it is determined to have fallen "below a threshold … of high priority network traffic." The idea is that as the traffic diminishes, the cloud service has a right to conserve network bandwidth "by performing a network disablement action with respect to network traffic to and from the VM, without disabling the VM." (In other words, to scale back the bandwidth of a virtual machine, while allowing it to continue running.) This patent was applied for Jan. 4, 2012, and issued Feb. 3, 2015, to three IBMers: Kelly Abuelsaad of Poughkeepsie, N.Y.; Lisa DeLuca in San Francisco; and Soobaek Jang, of Hamden, Conn.

Scaling back a virtual machine to match a decline in high priority traffic. It's not exactly the first time this idea has occurred to a cloud operator, is it? Should IBM own a patent on it? You'll have to judge its distinct and original properties for yourself in "Activity-Based Virtual Machine Availability in a Networked Computing Environment."

"This can help reduce costs charged by a cloud service provider for the amount of data transferred over a network to and from the VM," explains IBM in its summary of the patent.

Granted, that's a simplified explanation, but how about Patent no. 8,984,132:  "System and Method for Supporting Secure Application Deployment in a Cloud." Parts of the invention "identify aspects of a software application that use secure data and deploy the secured data to one or more secure servers." Hmm. This is a patentable process because it's unique to IBM? "Communication between the secure servers and publicly available servers may be managed using secure communications," the description continues. Now there's a breakthrough.

Then there's Patent 9,015,164 for the high availability of virtual servers. "Aspects of the disclosed invention enable a cloud environment to take snapshots of virtual machines, which can then be used for recovery purposes." Really? Snapshotting is patentable in the cloud? It's widely used in many ways in storage systems, including cloud storage systems. Why would IBM be able to get a patent on that? Read "High Availability for Cloud Servers" and perhaps you can explain it to me.

How many of these cloud patents should the Patent Office examiners be issuing?

Exactly how deep is the Patent Office's cloud expertise, anyway?

Is it as deep as its touch screen expertise, which led to its award of all those patents to Apple on the iPhone, even though prior art seemed to indicate Apple didn't invent very many of the touch screen's features? I hope the Patent Office will do better by the cloud in terms of keeping it out of one vendor's hands.

Meanwhile, IBM continues to rack up patent applications and lock up awards on cloud processes.