Slashdot Mirror

← Back to Stories (view on slashdot.org)

Mainframe Techies Are A Dying Breed

Posted by timothy on from the why-in-my-day dept.

dipfan writes "Great piece in today's Financial Times on the surprising survival of mainframes - but the problem in the US is finding experienced techies to run them: "55 per cent were over 50, compared with fewer than 10 per cent of those with Unix or Windows NT server skills." Cobol programers, still needed for legacy applications, are mostly in their 40s. Help is on the way, though, thanks to IBM's use of Linux, which "freshens the labor pool" according to the article." (See also this earlier post on the mainframe-operator labor pool.)

34 of 483 comments (clear)

  1. Employers' fault... by darken9999 · · Score: 5, Insightful
    Maybe if employers wouldn't require every employee to have such mass amounts of experience, there would be a few younger admins around. You know, almost like a junior admin... "Well, he knows how to admin a system, so we can teach him the specifics."

    I think being a mainframe admin would be a blast (maybe I just don't know better), but in my eight years of sysadmin work, I've never touched a mainframe. Every job posting I recall coming across required previous experience.

    1. Re:Employers' fault... by Xzzy · · Score: 4, Insightful

      > Every job posting I recall coming across required
      > previous experience.

      Takes a long time, but it will become a self-resolving problem. The existing "old guard" will eventually die out (either literally or via retirement) and create a demand in the market.

      This will either cause companies to lower their standards or discard the old mainframes.

      It would make good business sense to address the problem before it reaches critical mass (ie, so much of the old guard is gone that there's no way to train newbies), but if the Y2K problem was any indication, foresight isn't a prerequisite for running a business.

    2. Re:Employers' fault... by Eskarel · · Score: 4, Insightful
      Well I would agree with the fact that it's the employers fault, but disagree on what it is thay are doing which causes it.

      I would of course agree with the frustration tech workers seeking employment right out of college, and workers in general for that matter. It's hard to get experience when everyone wants it to give you a job, but that's not really the problem when it comes to network and sysadmin positions.

      It's perfectly sensible to require people who run something as complicated as a mainframe, or even network administration, or a half a dozen other things require strong experience. However if these companies want to have people available to work on these systems in the future they also have to provide opportunities for people to gain this experience without having to rely on them for full administration. That is to say companies should be hiring more PFY's so that they can train the next generation of administrators through real life experience.

      Employers don't want to do this of course because it involves having an extra employee, but they would be much better off in the long term if they had people who had real experience.

      Of course an additional problem with this sort of thing is unrealistic pay expectations among tech workers in general. No one is going to hire a PFY for 60,000 a year, but there are still many people on slashdot who believe that 50,000 is a ridiculously low salary for a full time job. So while a lot of it is employers being cheap, it's also somewhat us being unrealistic.

    3. Re:Employers' fault... by captain_craptacular · · Score: 3, Insightful

      What school did you go to? I graduated 2 years ago and it was ALL about C/C++ and there was a required class in assembler (MIPS).

      I was under the (possibly mistaken) impression that most CS schools were harping on C/C++ because if you knew them, you could learn almost any language quickly because >50% of them are based on C, use C syntax, use C++ object constructs, etc...

      --
      They who would give up an essential liberty for temporary security, deserve neither liberty nor security
    4. Re:Employers' fault... by sql*kitten · · Score: 1, Insightful

      Exactly. It's also very much a chicken-and-egg problem to get into the mainframe world, becuase the barriers to entry are much higher from the standpoint of working on the technologies at home to get that all-important First Job. It's easy to build a $500 linux server or buy a $1000 used ultrasparc sun machine to learn some unix and unix coding on, but ... how are you going to learn mainframe stuff?

      In this economy? IBM can hoover up entire graduating classes with the promise of steady jobs then send them to Mainframe School. What're you gonna do, sit round for 6 months, 12 months, longer, working in Starbucks and waiting for a hot dotcom job, or go work for Big Blue with its health and dental and matching 401k and bottomless pockets? It's really not the problem the scaremongers are saying it is. Any of the big systems integrators - Accenture, CSC, EDS, Unisys - could do it, and they will, as soon as it's necessary to their businesses.

  2. Huh? Stuffing FUD in there or what? by stratjakt · · Score: 5, Insightful

    "55 per cent were over 50, compared with fewer than 10 per cent of those with Unix or Windows NT server skills." Cobol programers, still needed for legacy applications, are mostly in their 40s. Help is on the way, though, thanks to IBM's use of Linux, which "freshens the labor pool" according to the article."


    How does linux freshen the mainframe labor pool, and not the Unix/Windows NT pool?

    Linux ain't System/36 or MPE or any other mainframe OS. And show me one linux app that's written in COBOL. (The language exists, but I've never seen it put to use).

    This is a self correcting problem. A good admin/coder can pick up mainframe stuff when he needs to. All the 50+ year olds are still working the jobs they got when they were 30. When they die off/retire, younger folks will pick it up.

    I mean, hell, I picked up enough about MPE and FORTRAN and COBOL to do my job inside of a week. And I got competent with S/36 and RPG at my last job.

    It aint rocket science. It's like a skilled machinist learning to shoe horses.
    --
    I don't need no instructions to know how to rock!!!!
    1. Re:Huh? Stuffing FUD in there or what? by TopShelf · · Score: 5, Insightful

      This more a symptom of employers concentrating on specific experience more than talented personnel. A fundamental skill that the vast majority of IS professionals have is the ability to LEARN and ADAPT. Unfortunately there's no buzzword that can signify this on a resume, so it gets ignored.

      --
      Stop by my site where I write about ERP systems & more
  3. So, the admins are old. by PhxBlue · · Score: 5, Insightful

    What's the problem, here? If the 50-year-old programmer is the only one who knows jack about mainframes, hire the 50-year-old programmer. Don't whine about not having enough qualified programmers, when what you really want is just-out-of-college programmers that you can bully into working for you at half the salary of someone with real experience.

    --
    !#@%*)anks for hanging up the phone, dear.
    1. Re:So, the admins are old. by PhxBlue · · Score: 2, Insightful

      I think the general idea is that the 50-something programmer will most likely wish to retire soon.

      I'm old enough to remember when companies kept their employees on long enough to offer them retirement plans. . . and I'm only 28. So if the 50-something wants to retire "soon"--which is probably on the order of 10-15 years--that still gives you several years where you'll have his expertise available not only to do the job but to train newer and younger programmers as well. There's really no good reason not to hire a 50-something, from a long-term economic standpoint; but then, no one's accused American companies of being able to see beyond the next fiscal quarter.

      --
      !#@%*)anks for hanging up the phone, dear.
    2. Re:So, the admins are old. by Anonymous+Codger · · Score: 3, Insightful

      50-year-old programmers aren't interested in innovation? We just want to clock time and go home? What a total load of bigotted BS. Attitudes like this are the reason older programmers have such a hard time finding work. I will look forward with pleasure to the day you celebrate your 50th birthday in your shipping crate home because the next generation of ignorant, biased young whippersnapper managers won't hire you.

      --
      No sig? Sigh...
  4. training? by Anonymous Coward · · Score: 1, Insightful

    So, what is the problem with hiring people with relaed experience and training them up on mainframes, assuming they don't mind pidgeon holing their careers? Training someone shouldn't be that hard, no? If no-one is going to train people in niche technologies isn't it obvious that there will be a shortage?

  5. main frame techies by pigscanfly.ca · · Score: 3, Insightful

    I know a main frames adminstrator . Depending on what you mean by main frames , the newer unix based ones I wouldnt mind adminstering . The problem is that there are a whole wack of old crappy mainframes which are running legacy applications that very few people understanding sitting around . Now if there was somewhere to actually learn about how to handel those I would probably take the course ; but as it stands now most info systems degrees dont deel much with legacy applications . Maybe a college degree in legacy code / computing in addition to a BSC would be interesting (of course colleges would have to higher old qualified people) . An alternative would be "just read the manual" ; however if I "just read the manual" most places wont consider me comptenet (nor should they there are tones of undocumented "features") . What is really needed (if we are going to keep on using this legacy systems without relapcing them) is for a tech publisher to gather up a bunch of mainframe adminstrators and document all the undocument features in the older generation (and newer ones as well) of mainframes .

  6. Yeah right ... by SuperDuG · · Score: 2, Insightful
    At my school for the undergrad degree plan of "Computer Science Business" you _HAVE_ to learn COBOL. I hated every minute of it because it's a pain in the ass and antiquated and outdated. We have an entire department dedicated to the maintenance of IBM MVS 390 systems that basically run the entire school. Everything from registration to classes to payroll is handled through a slew of cobol programs and frontends all designed almost 20 years ago.

    Basically the "Computer Science Business" degree plan is designed to make cobol monkeys for either the school, statefarm, kraft, or caterpillar who still rely heavily on cobol for day-to-day operations. What's the catch? In less than 10 years all the formentioned companies will be converted to either a .NET or Java platform to control all their operations. COBOL's last major reworking was done 18 years ago, it's time to switch to something new.

    I hate cobol and I always will, if I ever see an VSAM or coding paragraph again I'll probably freak. I'd rather work at McDonalds than be a COBOL monkey. I don't think I'm alone with my views either, as this article proves. These systems are old, prone to crashes, and not supported by level one support anywhere. They have heavy maitenance price tags and it's for this reason that it is more economical for these companies to completely rewrite their systems. IBM Running on Linux will NOT save COBOL, it's a dead language, just some people still speak it.

    Death to cobol you worthless language.

    --
    Ignore the "p2p is theft" trolls, they're just uninformed
    1. Re:Yeah right ... by YrWrstNtmr · · Score: 2, Insightful

      In less than 10 years all the formentioned companies will be converted to either a...

      That's what they said 10-20 years ago, when those systems were new.
      "Don't worry about that pesky 2 digit year thing. These systems will not be around that long."

  7. We are not dead, we just by Archfeld · · Score: 4, Insightful

    moved into more lucrative positions. Match my current salary and I'll go back to hexdump processing, IMS MTO, CICS batch, MVS/TSO, JES3/2, VM, REXX, DOS/VSE you name it. I've been a mainframe/mid-range support in nearly every environment around, I can even roll a VTAM sub-area :)
    But M$ exchange cluster design and management pays MUCH better.

    --
    errr....umm...*whooosh* *whoosh* Is this thing on ?
  8. "mostly in their 40s" -- oh no! by fmaxwell · · Score: 4, Insightful

    Cobol programers, still needed for legacy applications, are mostly in their 40s.

    Oh no! People in their 40s will only want to remain in the work force for another 20 years or so. What will the companies do then? Train people? Not in the U.S.! All employees must be hired with all needed skills. We wouldn't want to spend money training them because that investment would be wasted when we laid them off and shipped their job over to India.

    Nobody gets upset that most CEOs are in their 50s. No one is concerned that corporate attorneys are usually over 40. You don't see a panic because the average charter boat captain is in his 40s.

    Working in the computer field is like living the movie Logan's Run. Once you are out of your twenties, everyone from management to your fellow tech workers thinks your time is over.

    Or is it simpler than that? Maybe companies realize that they can underpay and overwork young, naive, single people but that people in their 40's with experience, families, and responsibilities will expect fair pay, benefits, and working conditions.

    1. Re:"mostly in their 40s" -- oh no! by Ian_Bailey · · Score: 2, Insightful

      The difference is that with CEOs, there are a whole bunch of executives waiting to climb the corporate ladder, just a litle bit younger. This progresses all the way down to the interns, fresh out of college/university.

      Law firms have lots of younger junior partners just itching for a step up, and paralegals and other staff behind them.

      But who's waiting to take over for the COBOL programmers? No one now, no one coming soon, and no one in the forseeable future. That's the problem.

    2. Re:"mostly in their 40s" -- oh no! by fmaxwell · · Score: 3, Insightful

      But who's waiting to take over for the COBOL programmers? No one now, no one coming soon, and no one in the forseeable future. That's the problem.

      If COBOL programmers were paid as handsomely as executives and corporate attorneys, you would find plenty of people eager to learn COBOL.

      This is not some big mystery. Why would someone in the IT field go that direction when COBOL programmers are paid more poorly, have less opportunity for advancement, and generally do less rewarding work? It's like paying teachers $20K per year and wondering why more people aren't getting teaching degrees.

      The pay is the problem.

    3. Re:"mostly in their 40s" -- oh no! by Moose4 · · Score: 3, Insightful

      Mod this guy up.

      Where I work, their idea of "retraining" us COBOL jockeys was to send us to MCSD 5-day cram refresher classes--never mind that none of us had coded VB6 before in our lives--and then expect us to get an MCSD certification, without EVER using the stuff in the wild. Once they realized that we couldn't do that, all training was withdrawn. Anything we want we have to get on our own, which is OK, that's the way things go--but we won't be given any opportunity to actually use the stuff unless we take a 50% pay cut to go to work as a "junior" programmer somewhere else.

      Meanwhile, we see MCSDs, .NET and VB6/COM+ specialists, brought in off the street at higher pay grades and more money than 20-year veteran employee COBOL programmers. Unless they're on H-1Bs, in which case they get brought in at $20k under what even we're making.

      If somebody is smart enough to be a good COBOL programmer for 15 years, they can learn new tools, even ones radically different. But instead companies will throw their older workers away--even though they're the ones that know the business processes--and bring in younger ones that they can work 90 hours a week and/or underpay.

      --
      "Settle down, Beavis. We've got an experiment to do."
    4. Re:"mostly in their 40s" -- oh no! by fmaxwell · · Score: 2, Insightful

      I cannot help but notice that people are in a company for 15-20 years, but still do not find themselves in a position to make these decisions.

      Yet they feel comfortable complaining about the decisions made by the ones who are in that position.

      If management is "comfortable" criticizing the work of those people who have chosen a career in software engineering, why shouldn't software engineers criticize the work (i.e., decisions) of those who have chosen a career in management? Many software engineers are in that field because they enjoy the work. They don't aspire to be managers. Criticizing them because they have not become managers is invalid.

  9. Don' t Go There by DeltaOne18 · · Score: 2, Insightful

    I used to be a COBOL programmer for an insurance firm. It wasn't by choice. I started out on a Java web app team and got transfered. I must say that there was a general lack of good technical knowledge about mainframe programming at the company (it didn't help they laided off some of the best guys) which made it even hard to do my job. Developing on the mainframe is much different, I find it more mondane and boring, then working with modern PCs and OO languages.

    I am sure some people like it but I hated it and had to leave the company to get away from it since no one would transfer to my position.

    --Kurt A web developer's weblog

  10. A Non Issue by tealover · · Score: 3, Insightful

    There are always a level of IT employees who didn't go to school and get a CS degree. It may be a clerical worker trying to move up. A painter trying to hop on the bandwagon. For many of them, they don't really know the technology out there.

    Employers target these people and train them. I know. I was one of them.

    I went to a school called Chubb in New Jersey, which is run by the Chubb Insurance company. It was originally an inhouse training development center for Chubb so they could train new employees on their mainframe systems. It got very popular and they opened it up to outside companies to make a few bucks. It has gotten very popular and is located in several states now.

    The companies who need mainframe workers know about schools like Chubb. The only thing that has changed at Chubb over the years as it became less of a Chubb training center is that they have to cater to the people who do know about current technology, so they also offer non-mainframe curriculum. But as far as I know (haven't been there in 10 years), mainframe is still their bread and butter.

    --
    -- You see, there would be these conclusions that you could jump to
  11. India to the rescue? by Tablizer · · Score: 3, Insightful

    I espect India to set up mainframe training centers and train hundreds of thousands in COBOL, JCL, etc.

    They have a habit of showing up at our doors for that kind of thing, whether we need them or not.

    --
    Table-ized A.I.
  12. Re:Legacy by override11 · · Score: 3, Insightful

    while legacy has something to do with it, the 100% uptime (with voluntary IPL's) of our iSeries mainframe is very compelling.

    Here's the scenario: A hdd fails, the system automatically calls IBM and a tech is dispatched the same day. I get paged, and meet a tech at the front door.

    IBM Tech
    I heard you have a drive failure here

    Me
    I do??

    IBM Tech
    No problem, I have a drive right here, it will only take a second to swap it out

    He swaps out said drive, zero down time, and nary a performance hit because a hot spare came online. You have got to love that kind of service and uptime, and just plain reliability.

    --
    No I didnt spell check this post...
  13. Re:Legacy by BitwizeGHC · · Score: 4, Insightful

    Bjarne Stroustrup has been known to observe that the primary difference between "legacy" systems and the systems replacing them is that the legacy system works and scales.

    A case can be built for the verity of that assertion as applied to the mainframe situation.

    --
    N4st0r, trixx0r h0bb1tz0rz! Th3y st0l3 0ur pr3c10uzz!
  14. Learn by doing it... by dark-br · · Score: 2, Insightful

    The problem is that no one teaches mainframe operations in schools, you basically need to learn by being dropped into it - and not screwing up everything. Fewer and fewer businesses are willing to invest in promising new talent to learn these legacy systems, but their own mainframe gurus are retiring or dying off - so eventually this corporations will 'bleed out' skill-wise.

    And no, the mainframe cannot be replaced by a client-server solution. I listened to this moron chant throughout school - mainframes are not dead. REALITY CHECK - there are just some applications where a mainframe makes more sense. Mainframes can handle enormous amounts of data without having to break it up for a cluster, or without being bogged down with I/O like most client-server type solutions. Mainframes are great when you need to handle databases with tons of information in it - and you need to consistantly dig through it. Most machines cannot handle it, and will buckle. Mainframes almost never buckle, unless you are testing new stuff on them (naughty newbie - that's what a test LPAR is for) or you do funky things to them.

  15. please mod parent down by theuglykid · · Score: 2, Insightful

    I didn't gain any experience from school, because most schools don't seem to value mainframes anymore. This is probably because server vendors and their software vendors are more aggressively seeking institutions of higher education. These vendors seem to have more throw away money and more "progressive" marketing strategies by getting students "hooked" on to their own products(MS is really good at this game) early in their educations. This is great because some tof the graduates of today will be the managers of tomorrow and will hold the purse strings of their IT departments. WHat do you think a manager will purchase if given an opportunity? The tools (s)he's already familiar with. What will a forward thinking manager purchase when faced with a need to upgrade the system? Some will survey what the prospective employees are already familiar with (possibly to cut training costs). I gained my experience by never turning down an opportunity to work (and thereby learn). My future is secure.

    I am a 25 year old programmer who spends 96% of the time working on OS/390 mainframes using JCL and MVS COBOL. Any other time is divided between Java and VB for special apps.

    The team I work with (5 of us, total) is officially dubbed the "Legacy" team. Our total IT department is comprised of roughly 80 employees (so you can see how few are able to do or want to do what we can). I am the youngest on my team by 12 years. I would guess that the average age of our team members is 45 (not including me in the calc). The great thing is, because I am willing to work and I lack the offensive attitude of the parent comment, I make BANK.

    I fear for you, but I don't fear you.

  16. This Article Sets Up Future H1B visa increase by Cryofan · · Score: 3, Insightful

    This article is likely a setup article for other articles which will eventually oh-so-delicately suggest that more H1B programmers are needed from India because they supposedly still have the "old" technology, and we desperately need those old Indian skills, so therefore best that we increase the h1b programmer quota.

    Some things never change......

    --
    eat shiat and bark at the moon
  17. Re:Or... by dogfart · · Score: 2, Insightful
    And what do you use for the OS?

    My understanding is that some 20-30 year old predecessor to MVS (or OS/390 or z/OS) is in the public domain, but obtaining more recent versions is very expensive (unless you use a pirated version - anyone know a z/OS warez site?).

    And what you will need is more than the OS, there are other utilities that are also licensed by IBM and cost a fortune.

    You might be able to learn a bit of JCL and some basic TSO commands on Hercules + an ancient mainframe OS, but you would not be qualified to do more than the most elementary tasks on any modern production mainframe.

    If IBM was smart, they would release a z/OS for Linux package, including basic compilers and utilities, just for hobbyists and students.

    Is anyone at IBM listening?

    --

    "dope will get you through times of no money better than money will get you through times of no dope"

  18. Locomotives are still around by Ra5pu7in · · Score: 5, Insightful

    Locomotives / freight trains are still used regularly. They serve a need that cannot be met with automobiles or even 18-wheelers. For Joe Sixpack and his family, an automobile is definitely a more efficient way to cross the country. For ABC Florist who relies on fresh cuttings, locomotives take too long - trucks are better. But for XYZ Furniture ordering fifty sofas, twenty-five coffee tables, one thousand various lamps, etc., it would take a large number of trucks (each having a driver to pay) vs. twelve cars in a freight train (one driver to pay).

    There is a use for mainframes in particular industries - personal computers and servers aren't the be-all end-all answer to every computing need.

    --
    I was taking one day at a time, but then several days got together and ambushed me. (from a Rhymes with Orange comic)
  19. From the Devil's IT Dictionary by Frankus · · Score: 3, Insightful

    mainframe n. An obsolete device still used by thousands of obsolete companies serving billions of obsolete customers and making huge obsolete profits for their obsolete shareholders. And this year's run twice as fast as last year's.

  20. Re:Or... by ebyrob · · Score: 2, Insightful

    perl, vi, emacs or anything you'd expect on a nix box ... tcp/ip may be a possibility

    What no vi??! You've got to be joking. I've yet to meet a platform without at least a couple crappy clones. Next you'll tell me regular expressions are not available and you're using a C compiler without ANSI support.

    Seriously though, any system not supporting the tools you mention would seem halfway dead already. I'd figure the only thing such an environment would be good for is jumping to something more useful. Perhaps in some cases it really would be more feasible to consider emulation instead of porting and/or rewriting. But if the question of getting onto something a little more mainstream isn't even being asked, that sounds like a recipe for disaster. (ie: someone else will do it for you, and gain a huge competitive advantage)

    You want experts on Rexx, JCL, RACF/ACF2 and such? Train them *after* you hire them. Additional languages (and platforms) should be easy enough to pick up for any halfway decent programmer or sys admin.

    For the record:
    z/OS has POSIX 2 support.
    os370 blech! Keep it away!
    MVS has at least POSIX 1 support.
    etc? not sure...

  21. I'm a fugitive from the law of averages? by alumshubby · · Score: 2, Insightful

    I'm an early-40s guy who's retraining to be a programmer (been a tech writer), and I'd like to break into COBOL programming -- mainly because around here at least, it looks like the road less traveled.

    --
    "How many light bulbs does it take to change a person?" --BMcC-->
  22. Re:60,000 too high? by lysium · · Score: 2, Insightful
    I beg to differ. The defacto trainiee is the Help Desk technician and the desktop technician. People who are beginning to know the fundamentals of computer/information systems. Many organizations will not even let IT/MIS staff log onto a server until they have demonstrated responsibile, intelligent actions. No one is going to allow trainees to operate production servers.

    Now the junior admin is still necessary, but the "junior" is something of a misnomer; The job is hardly entry-level now. Another factor: the job pool is still overflowing with the gold-digging inept. So many untechnical (at heart) people are masquarading as IT workers, that it becomes necessary to offer more money simply to attract a level of competence.

    Or these are just clever rationalizations. Funny, you speak of paltry living wages as a positive thing. You see, the corporate people on top of the pyramid will still be making "New York wages" while the techies join the rest of the wage slaves making "Missouri wages." So yes, the companies cannot afford to employ everyone at exaggerated salary levels: in the future, the executives will simply make certain to keep it only for themselves. This suspicion will always make me advocate higher wages for skilled labor.

    --
    Together, we will drive the rats from the tundra.