In-Depth
Behind IBM’s Mainframe Market Push
The best thing IBM can do to help secure the mainframe’s future is to roll up its sleeves and get to work, says one prominent Big Bluer
When IBM Corp. trumpeted a new $100 million mainframe offensive earlier this month, some Big Iron pros seemed curiously underwhelmed. It isn’t that they question Big Blue’s commitment to the mainframe, nor do they doubt IBM’s intention to spend the sum it has promised in the ways it has promised—i.e., to help drive mainframe usability, manageability, and programmability (see http://www.esj.com/enterprise/article.aspx?EditorialsID=2212).
Some mainframers say throwing money at a problem isn’t always (or even mostly) the best way to resolve it.
“The mainframe is still sick, if not dying. There is a lot of work to do to return it to viability,” said one mainframer, a system operator in a small Big-Iron shop. Pricing, for example, remains this IT pro’s biggest bugaboo: his shop doesn’t have zSeries Application Assist (zAAP) or zSeries Integrated Information Processor (zIIP) engines—nor does it have any need for one. What he needs, he says, is for IBM to make the mainframe a more affordable proposition for traditional customers running traditional workloads.
Bob Hoey, worldwide vice-president of System z sales for IBM, agrees—to a point. Spending $100 million by itself will do little to secure the mainframe’s future, he concedes. Truth is, Hoey maintains, there isn’t any single silver bullet that will help vouchsafe the mainframe’s future. The best that can be done, he says, is to roll up one’s sleeves and get to work.
Hoey argues that IBM’s $100 million mainframe offensive is a case of doing just that. Far from being a silver bullet, it’s earmarked to address a single business problem. “There’s been a lot of talk in the industry about a year 2007 problem. Most analysts will tell you that by the year 2007, about 50 percent of those people with mainframe skills will be retirement-eligible. So there’s been a tremendous focus on the part of IBM to ensure that we bring the next generation of mainframe skills to the marketplace.”
IBM has several similar initiatives, including its zSeries Academic Initiative (see http://esj.com/Enterprise/article.aspx?EditorialsID=1583). That effort primarily targets college students in an effort to get them interested in and working on Big Iron. This month’s effort, on the other hand, focuses primarily on IT pros in non-traditional mainframe environments.
“What IBM decided to do in this announcement is to put a stake in the ground and say, ‘We’re committed on a five-year journey in reducing the amount of complexity in managing a mainframe compute environment.’ This is a statement that applies equally to the other IBM middleware products that run on the mainframe platform,” Hoey explains.
The goal, he says, is to accelerate the rate at which programmers and operators attain proficiency. “Instead of the two to three years that it might take today in order for a systems programmer to get proficient on the platform, we hope we can compress this down to six months or less. We hope that by doing this, we think we can convince them to train their current systems programmers on to the mainframe and consider it as an option or an alterative to RISC and Intel.”
Will Simplification Introduce Vulnerabilities?
There’s a danger here, too. After all, one of the mainframe’s strengths is an airtight security model which inherently says that whatever isn’t expressly permitted is reflexively denied. The opposite has traditionally been the case on Windows systems, and yet the latter are typically lauded as hallmarks of usability, manageability, and programmability—in other words, of the very trifecta IBM is trying to nail on the Big Iron front.
Isn’t there a danger that as Big Blue tries to simplify mainframe administration and software development—in part by redesigning interfaces and automating (via Windows “wizard”-like prompts) common tasks—the mainframe itself will become prone to some of the same vulnerabilities or exploits that have plagued Windows systems?
Hoey categorically rejects this possibility. “We are going to ensure that we continue to do everything required such that the mainframe’s vulnerability is not increased. The mainframe is recognized as the most secure platform on the planet. It’s got industrial-strength, bullet-proof security that we’ve written into the access control, authentication control, and general higher-level security features of the platform, and we are not going to in any way, shape, or form dilute or reduce the effectiveness of the security features,” he maintains.
On the other hand, Hoey argues, security administration in Big Iron environments could almost certainly benefit from an ease-of-use infusion. “The security administration on the mainframe today is probably exponentially more complex than it is in a Windows environment. We will never, in my opinion, have a systems management environment that is as easy as or easier to use than a Windows environment, and one reason for that is that we just have much more code in a z/OS environment than you would in a Windows system,” he contends. “For example, we have more than one million lines of error-correction code written into z/OS. That code does not exist in Unix or Windows systems.”
Elsewhere, Hoey says, IBM is making good progress on another of its Big Iron training initiatives. “We announced [an initiative] to train an additional 20,000 mainframe skills by the year 2010. In fact, we’re well ahead of that, so we’ll probably bump that number up,” he notes.
Hoey says he’s optimistic—even downright bullish—about the mainframe’s future. At the same time, he stresses, he isn’t deluding himself. A host of problems need to be addressed, including software licensing. As far as he sees it, however, training the next generation of zSeries pros is the most pressing mainframe problem, and training involves not just inculcating competency, but proficiency and skill, too. That won’t be easy, he acknowledges. “It’s going to be a lot of work on the part of IBM to ensure that the next generation of mainframe computer users are not just competent but proficient.
“We don’t think there’s an easy answer to this problem. There’s no silver bullet. So we’re investing wherever we see opportunity. Not only do we have this academic computing initiative, [but] we’re also working with ISVs to enable more of the ISV software on the platform. We’re also working with a number of our existing installed accounts to put in place the appropriate education and training. We’re betting we will be successful, but it will not be an easy task.”