If that wasn't surprising enough, 74 percent believe the mainframe will have some role in a cloud computing initiative, and 70 percent will sustain or extend their current mainframe environment into the cloud. Finally, 82 percent of respondents said they intend to use the mainframe in the future either as much or more than today."First of all, you have to consider who sponsored the study, and it's clearly somebody with a mainframe dog in the cloud computing hunt. Second, I'm not sure too many people in the emerging world of cloud computing were declaring that the mainframe was being displace by the cloud. Thus, none of this surprised me.
Whatever technology comes along, we assume that's going to be it for the traditional mainframe computers that have served us faithfully for the last 30 years. It has not happened yet, which goes to the point made in the survey, but mainframe-based-computing has evolved with all other technologies over time.
The fact of the matter is, there are no good paths (meaning approaches and tools) for getting from the mainframe to the cloud. Thus, it's not that we don't want to consider cloud computing in its private, public, or hybrid forms as an option to finally getting rid of the mainframe computers. It's just that the cost of doing so is largely prohibitive and the tools required are nonexistent.
Those focused on cloud computing these days are looking at providing SaaS alternatives to traditional packaged enterprise applications, such as ERP and CRM, or building new applications in the cloud on PaaS and IaaS providers. Few have looped back to look at the more "traditional" systems as contenders for cloud computing migration as of yet.
I suspect that if we used a time machine to go five years into the future, we'll find that few legacy or mainframe applications have moved anywhere. But, not for the reliability and efficiency reasons cited in the survey. It's just too costly and too much of a pain in the neck to do so.
Where I do disagree with many in the mainframe space is that mainframe computing will continue to be a platform for new applications. When you do a cost comparison with open-source options around these days, or platforms that happen to exist in the clouds, it's clear that it's very cost inefficient to build these applications on mainframe computers, generally speaking.
I suspect that those in the world of mainframes are sweating a bit over the hype-driven manner in which cloud computing has come onto the scene. Like other so called revolutions that promised to change the landscape of the data center, mainframes will be largely untouched in the short term.A recent study found that "79 percent of IT organizations consider the mainframe to be an essential component of their cloud computing strategy."... Yet, there are no good paths (meaning approaches and tools) for getting from the mainframe to the cloud...