Mainframe management gets its swagger.
Back in the old days kids followed in the footsteps of their
parents and their parent's parents. A proud tradition of craftsmanship, skills
and knowledge continuing across the generations—weavers, bakers, potters,
bankers. But never (oddly) Mainframe Systems Programmers. Indeed, so loath have
been multiple generations of twentysomething IT graduates to code Assembler and
JCL, and to delve into the minutia of z/OS administration, that all that stands
between here and the final death of the IBM mainframe market is a thin greying
line of aging expertise. Granddad might be eyeing a permanent move to the
potting shed, but for now he's still got a PTF to apply and
an SMP/E install to finish.
The green thumbs of retirement have had to wait for just one more day of typing
in front of the green screen.
In all seriousness, the aging of the workforce that keeps
the world's mainframes running is now recognised as a serious problem. Despite
oft-quoted claims of the platform's death, the mainframe continues to be the
computing engine that powers much of the economy's basic processing functions.
Withdraw cash from an ATM, use that cash to buy produce from a major retailer,
and later swipe your credit card to refuel your car and with every transaction
you've probably cycled through the processing engine of a mainframe sitting
quietly in the background. Despite the efforts of the UNIX/Linux/Windows
vendors over the years to swap all that mainframe workload to a range of
distributed computing architectures no real credible alternative has emerged
for the bulk of the high end processing needs served by the 'frame. So we are
left with a lot of business critical processing taking place on a platform that
is managed by a workforce scheduled to retire en masse within the coming five
to ten years.
The challenge now is to make what is seen as a legacy platform
manageable by a new generation of systems programmers—though it might be more
accurate to say that the first challenge to be solved is getting a new
generation even interested in working with the 'frame to begin with. In the
eyes of infrastructure vendor CA these two problems are intimately connected.
The tips of CA's roots are firmly anchored deep in the
mainframe utility and management markets—so much so that arguably it has
become inextricably linked in the eyes of Wall Street and customers alike as "a
boring legacy platform vendor." CA is the largest mainframe software ISV,
second only to IBM in share of the mainframe wallet. Of course CA does a whole
lot more than mainframe software, from distributed systems and network
management to security management to project and portfolio governance. Indeed
CA has been so focussed on building capability in such markets for much of the
last fifteen years that they have allowed themselves to take the foot off the mainframe
accelerator, whilst turning their eyes away from the road ahead.
CA is, of late, a vendor with a mission to not only revive
their own mainframe business, but along the way to do no less than help
resurrect the overall fortunes of the platform. Under the leadership of EVP
Chris O'Malley, CA's mainframe business unit has put great emphasis and effort
toward entirely modernising the manner in which the platform is managed and
maintained. CA's mainframe ambitions were announced in November 2008, however
announcements aren't delivery. In the intervening months O'Malley's team have
made good the first stages of their roadmap promises and after much development
effort are now busy working with a reasonably wide customer base to beta test
an entirely new way of installing and maintaining software on the platform.
CA's Mainframe 2.0 initiative has thus far delivered (in
beta) an entirely GUI driven way of performing the esoteric task of installing
software packages, driving the (IBM) proprietary SMP/E install process. This
author speaks from experience in describing SMP/E as a detail oriented task entirely
devoid of drag, drop and click. "Intuitive" is not a word one would
naturally use to describe the process by which almost all software is installed
on the mainframe platform.
CA's Mainframe Software Manager tool uses GUIs (served natively from the 'frame) and
software interface constructs that will be familiar to anyone who has ever installed software on a
PC or Mac platform. In doing so, the skillset required to manage the platform
has been dramatically reduced, while CA's (not independently verified)
benchmarks indicate that the time taken to install a product is also dramatically
reduced when their tool is used, compared to a native SMP/E process.
Meanwhile CA has gone on a recruitment drive of their own to
woo recent graduates to work on the development of the Mainframe 2.0
initiative. O'Malley and his team have logged up an unenviable CO2
footprint in an attempt to build a buzz amongst the existing mainframe community
as well as in the ranks of soon to be IT graduates. Perhaps it is also a
statement about the state of the job market, but CA reports that a
surprisingly high number of students have crowded into O'Malley's university recruitment
sessions over the last six months or so.
Turning around the fortunes of the platform is not without
significant challenges however. Arguably this is an effort that software vendor
CA can never hope to pull off without the support of IBM, even while the New
York vendor attempts to take software marketshare away from Big Blue. Achieving
the balancing act of co-opetition between CA and IBM will be a delicate task as
the strategic goals are balanced against the hard headed tactical sales targets
that drive field sales force behaviour.
Meanwhile CA is only at the first stage of delivering
against its multi year and multi step development roadmap for Mainframe 2.0.
The evidence of CA's efforts so far is promising and existing mainframe
customers would do well to pay attention to CA's strategy and execution,
especially if the aging workforce problem is one that is real and increasingly
urgent to those customers. CA may yet make the mainframe so interesting to the
next generation of IT workers that the platform will stop being one that is "Whotevva
- FAIL", and start being a...well...whatever it is that you young people today
call something that is right-on and groovy.
attended CA's Industry Analyst symposium in Ottawa. The author worked for CA
until 2007, and today holds no financial interest in the vendor.