IBM SoftLayer Aims to Bring the Cloud to the Mainframe

Mike Vizard
Slide Show

Data Center Applications and the Cloud: What You Need to Know

For years, IBM has been trying to couple the distributed x86 and RISC-based systems to the mainframe, with mixed success. While customers clearly need to manage different classes of workloads on each of those platforms, not many felt compelled to acquire them all from IBM. In fact, the inability to execute that strategy is one of the reasons that IBM is opting to sell its x86 server business to Lenovo.

But with its acquisition of SoftLayer, IBM is gearing up to try a new approach. At the IBM Pulse 2014 conference this week, IBM announced that SoftLayer will be making available IBM Power servers in its cloud for workloads such as Big Data analytics that are optimized for RISC-based servers alongside the x86 servers from Super Micro that SoftLayer currently uses. The next logical step, says IBM SoftLayer CEO Lance Crosby, is to take pods of servers made up of both x86 and IBM Power servers and make them available as services that can be easily invoked by a mainframe application.

In the parlance of SoftLayer, a pod is a collection of about 150 racks of servers consisting of anywhere between 4,000 to 5,000 nodes that can be dynamically applied to any workload. Crosby envisions making those pods first available as a service alongside the mainframe systems that IBM hosts on behalf of its customers, and then also giving customers the option to deploy a pod alongside a mainframe running on their own site.

Crosby says he’s lobbying within IBM to make the mainframe itself another element of the SoftLayer cloud. But given the complexity of the licensing terms under which IBM provides access to mainframes, many questions would need to be sorted out. For all of its efforts to shift to services, the IBM mainframe, along with all the products and services, still accounts in one form or another for the bulk of IBM revenues. As such, tinkering with the mainframe licensing model is not taken lightly within IBM.


At the same time, however, Crosby says many of the traditional approaches IT vendors have used for decades are now simply archaic in the era of the cloud. In fact, Crosby says the time may have come to re-evaluate how data centers are actually constructed. For example, 18-in. racks full of two- to four-socket servers are a throwback to another time. Crosby says in the future the entire building where the data center resides may one day be a bus that cloud service providers simply plug processors into as needed.

While Crosby doesn’t think that every workload will eventually find its way on to a public cloud, he does note that the economies of scale of public cloud services are becoming too good to ignore. Right now, SoftLayer requires about one administrator per 1,000 servers, a ratio that is only going to improve over time.

What will ultimately differentiate SoftLayer from other services, says Crosby, is not so much the cost of delivering those services, but rather the amount of control SoftLayer gives IT organizations over those resources via thousands of application programming interfaces. By the end of the year, IBM SoftLayer will have 28 global data centers in place that are all connected via a common private network that allows customers to treat all the components of those data centers as one private virtual data center. For Crosby, all those components are essentially ingredients that IT organizations can compose at will in ways that no one inside or out of the cloud can predict today.



Add Comment      Leave a comment on this blog post

Post a comment

 

 

 

 


(Maximum characters: 1200). You have 1200 characters left.

 

 

Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.


 

Resource centers

Business Intelligence

Business performance information for strategic and operational decision-making

SOA

SOA uses interoperable services grouped around business processes to ease data integration

Data Warehousing

Data warehousing helps companies make sense of their operational data