Is the clock ticking on mainframe expertise?
27 Apr 2010, Andrew Charlesworth, Computing
http://www.computing.co.uk/ctg/analysis/1832081/is-clock-ticking-mainframe-expertise
Earlier this month, in a report entitled Ensuring You Have Mainframe Skills Through 2020, analyst Gartner warned of an impending mainframe skills gap. A generation of experienced IT staff are approaching retirement just as the mainframe is experiencing a renaissance.
The world’s top 25 banks run on mainframes and there are more Cobol transactions in a day than Google hits, according to some estimates. IBM, a mainframe market leader, profits handsomely from its computing behemoths: last year’s revenue from System z was up 19 per cent on 2008.
Mainframes are highly efficient. For example, a System z can replace 1,500 x86 servers but consumes 15 per cent of the power. Bank of New Zealand, for example, is replacing 200 Sun servers with a five-CPU System z.
However, not everyone agrees that the industry faces a mainframe skills issue. After all, some so-called skills crises are often over-egged by head-hunters and trainers, and mainframe staff have been laid off owing to recession and offshoring.
“Valued mainframers, far from being lost to normal retirement, are in reality being lost to redundancy and forced into early retirement,” said IT worker John Williams, in response to Computing’s article about the Gartner report on 13 April.
Other IT workers have contacted Computing, angered by what they see as a systematic “ship-it-abroad” policy for IT jobs and puzzled by the incongruity of a mainframe skills gap.
One points to a job advertised for a software licence controller in south-east England commanding a salary of £40,000 but carrying a caveat: “This position will initially be UK based (1 to 2 years) and then based in one of our offices in India. The salary will then reflect the Indian salary scale.”
Even IBM sees no immediate skills crisis. “I’m not getting any feedback that there is a significant mainframe skills gap,” said Colin Grocock, IBM mainframe business development manager.
Furthermore, mainframe vendors such as IBM and Unisys and software developers such as CA and Micro Focus are simplifying mainframe development by automating operations, and training a steady stream of graduates.
However, no matter how simplified the programming tools or automated the systems, experienced staff are still needed to maintain critical legacy apps.
“The deep knowledge of 20-year-old application environments is essential,” said Grocock. “There is a big difference between deep business application knowledge and programming language skills.”
For example, at mail order apparel firm JD Williams a key member of the mainframe team retired last December.
“He was an expert in a 25-year-old Customer Information Control System application,” said Mike Madden, the firm’s development services manager. “He wasn’t the only one who knew that system, but he was the man you’d go to if you had a problem.”
So arguably the skills gap is not in entry-level skills, but the deep knowledge of business applications built up by experienced staff over 30 years.
“How come we lay off people with these skills? The industry has almost stripped itself bare,” said Darin Edmunds, head of education services at Micro Focus. “We have to start teaching core essentials of mainframe computing, not just programming by rote, and educate people to apply their training to make meaningful business solutions. That only comes with experience and knowledge of best practice.”
Indeed, it was these deep-level skills that were highlighted in Gartner’s report.
“A bright and motivated 23-year old will not be ready in week one to step in and replace personnel with intimate and detailed knowledge based on 30 years of experience,” reads the report.
Furthermore, developing for mainframes requires different thinking compared with, say, Windows or Linux systems. Mainframes use procedural language where control resides in the program, not user interaction.
And in organisations where applications have been patched on patch over 30 years, code can be optimised but there still needs to be someone who understands the overall structure of the application.
“It would be naïve to think that you could replace that person with software,” said Thomas Leitner, senior European vice president of mainframe solutions at CA. “Our idea is to free up that person from routine admin tasks so they can focus on optimising the code and passing on their knowledge.”
Demand for experienced staff with business nous is likely to outstrip supply, warned Gartner vice president Mike Chuba: “There will be a call for this deep knowledge, as vendors and the various consultants, system integrators and outsourcers are all in need.”
That might not affect organisations today, but they need to plan for the future and ensure that skills are transferred.
“We do not have anyone else who is critical to the mainframe due to retire immediately, but in five years there will be a steady stream, so we have to keep on top of it,” said JD Williams’ Madden.
Reader comments
© Incisive Media Investments Limited 2012, Published by Incisive Financial Publishing Limited, Haymarket House, 28-29 Haymarket, London SW1Y 4RX, are companies registered in England and Wales with company registration numbers 04252091 & 04252093
Yes, migrating mainframes to the cloud does take skills (in mainframe AND cloud)
Statements like "Mainframes are highly efficient. For example, a System z can replace 1,500 x86 servers but consumes 15 per cent of the power. Bank of New Zealand, for example, is replacing 200 Sun servers with a five-CPU System z" are VERY misleading.
Anyhow, much of the IT computing does still run on mainframe. The business agility of a mainframe is very low while in cloud it is high.
The schools do not teach people mainframes anyhow. Though, skills are really acquired over 30 years.
IBM, itself (I spent almost 25 years there starting with mainframes and ending in cloud) has a skills problem. They will not publicly admit that.
In my opinion, migrating mainframes to the cloud is an opportunity worth billions and billions of dollars.
Posted by: Derik Pereira 28 Apr 2010
A solution to the dilemma faced by major organisations
I frequently have conversations with IT Managers who have suffered great loss. Loss that is of key dependencies (those individuals who seem to work 24x7 who can answer the question and resolve the problem almost instantly) and critical knowledge vital to IT teams. Not the ?how to do? something kind of knowledge but the important ?why? knowledge, built up over many years and forged out of trial and error, the love and hate relationship with systems, applications and ever changing business requirements ? the blood, sweat and tears of the IT world.
These are the people who are at the heart of an IT team ? those who provide the answer every time, day or night. They know the history of why, when and where and can point you to a line of code in a legacy app, defend obscure mainframe traits, or give you the name of the developer who came up with the fix or solution. In fact the very generation of experienced IT staff approaching retirement as described in Andrew Charlesworth's article - Is the clock ticking on mainframe expertise?
Interestingly this is not a new phenomenon and is not just prevalent in the mainframe environment. For a number of years IT managers have been considering the risk posed by people rather than technology single points of failure and key people dependencies within their teams. Some have been very pro-active in grasping the nettle and have endeavoured to de-risk their environments, with varying success dependent on the rigor placed around the methodology employed. The issue has always been one of priority vs the drive for business as usual. A programme started internally with best intentions rarely succeeds in getting to completion when the priority call is made around project delivery vs any kind of training initiative.
Working with these IT SMEs (Subject Matter Experts) for the past 10 years, particularly in the banking sector, has provided a real insight into the challenges faced by organisations as they work to embed a knowledge sharing culture, and has proved that this is not an insurmountable problem. Managers have been making choices which ensure that this experience is cascaded into teams to build best practice, that knowledge is retained when an individual exits or retires from the organisation, or where mergers and acquisitions necessitate rapid knowledge transfer to integrate IT teams.
However this is not an area where universities can provide a learning module which will magically transform graduates, because the crux of the problem is internally based focussed on ?why? rather than ?how? expertise, something which can only be developed over time in the context of organisational understanding. This therefore precludes any formulaic approach to understanding how to resolve problems and the thinking behind that resolution, as those decisions are based on year?s of accumulated expertise which could never be pre-empted in the context of learning outside of a particular job role.
It is a truth that a new recruit or long term colleague could shadow an SME for any period of time, and of course would observe enough to be able to replay a given situation, but would still not understand what that SME had drawn on by way of expertise to deliver that resolution. It is that tacit knowledge which has to be scoped and shared to deliver resource flexibility and benefit within IT teams.
It therefore doesn?t matter how talented, bright or enthusiastic personnel are if an organisation cannot offer them the internal training opportunities which expose them to real subject matter experts in the context of a structured programme which develops their understanding of ?why? and allows them over a period of time to demonstrate a depth of understanding which can be applied and then validated by the expert.
We at Answer Consulting (www.answerconsulting.com), formally known as UPCO, using our Mindswap programme assist organisations to do this day by day with real success. We provide the solution to the dilemma faced by many organisations and deliver a structured and measurable approach to knowledge sharing and transfer which works. We are the answer to that problem - you only had to ask.
Posted by: Lindy Wainwright, UPCO Programme Manager for Knowledge Transfer and Principal Consultant 05 May 2010