back to article How do we train the next generation of data centre wranglers?

There's a world of difference between what the average IT person does today and what they'd have done ten or fifteen years ago. So where's it going? What will we need to teach the next generation of data centre staff? The shifting focus The big difference between yesteryear and now is the growth in managed hosted services. “ …

  1. Anonymous Coward
    Anonymous Coward

    Technology evolves like religions.

    The first generation knows why they do things - because they had evolved the rituals from their own thoughts about the way things work.

    Most of the next generation just uses the same rituals with only an inkling of "why".

    The generations after that have no idea of the "why". All they get is an introductory "history" lecture before they are plunged into the "important" details of how things are done now.

    I have met network designers who are trained in the use of manufacturers' configuration tools and how to crank the handle. Then they ask me a question that is puzzling them - like "what does the prefix mega actually mean?". Binary arithmetic is a mystery to them too - never mind what phrases like "Manchester Encoding" might mean.

    1. Bob Wheeler

      "Manchester Encoding"

      Oh my, Oh my, tip of the hat for that one. Manchester Encoding was the best, it only lost out due to the greed of some company that shall remain nameless.

    2. IanCa

      who designs & builds the clouds?

      those network designers who turn the handle - they aren't designers, they are monkeys.

      proper network designers understand all the layers - physical up to application.So many people go on about this xxxaaS, "it will all be in the cloud". But a cloud is not just a fluffy thing - its some actual stuff that someone has to make work - which have lots of layers, lots of real tin in DCs, and lots of complexity - they may have lots of nice gui drag and drop me a new server in the UI, but underneath there's a lot of stuff going down that someone has to understand. Because at some point it will not work and have to be fixed. or tuned / optimised because it doesn't go as fast/far as promised. Vendors cannot support all this by themselves, you need IT bods who understand servers, networks, storage, OS, physicals , to build a cloud - all the same stuff.

      the problem set (and the work that it generates) has not gone away its just moved slightly. Larger companies are building / have built their own "cloud tech" infrastructures - so the wranglers still have a place to go work. can't speak for small co's never worked in that end of the market.

      probably the best way to learn this stuff nowadays, given that almost everything is virtualisable, is to get hold of a small pile of 2nd hand old-spec tin and build your own cloud. the traditional way of learning on the job is probably no longer valid if the small-co where you start out in a dogsbody IT support job is purely a consumer of cloud Xaas services (i.e. not far up the stack from dumb end user).

  2. InITForTheMoney

    Demise and rise of the IT generalist

    As a contractor in IT Infrastructure who moves between clients fairly frequently, I am often pretty shocked by the comparitiviely low level of expertise of the Data Centre operations staff that I meet on my clients sites, I'm not complaining, it keeps me in a job, but I wonder about the sustainability of it. The problem I think stems from the way that IT departments are structured now. I am not conviced that there is a pathway that really enables traning in IT Infrastructure, because there aren't really any basic jobs that are very useful at educating staff in the more complex stuff anymore.

    I'm only 33, but when I took on my first IT support job in the year 2000, it was doing break/fix desktop and server support, we had a mix of 10BaseT and 10base2 networking across the 3 floors of the office building, we also had a Citrix MetaFrame farm of 2 servers that supported about 60 thin clients that were old 386 and 486 PC's running a DOS based Citrix ICA client, there were then about 40 high end CAD workstations that used a mix of WinNT and Win2000 and accessed a few published apps via Citrix.

    In order to support this environment it was necessary for me to understand the basics of the various Windows operating systems, of DOS, of Citrix and a little of Linux. I then also had to undertsand plenty about the various different network types and protocols that we operated in the environment (primarily TCP/IP and IPX), about bridges, routers and switches. There was a team of 3 of us and we did everything from supporting the servers, configuring the firewalls, building / imaging the PC's and 'thin clients' and running through the office moving terminators in order to locate faults in and fix our decrepit 10Base2 network. In this environment, I learned the basics and root of almost all of the IT knowledge that I use today, everything I know now is an extension of one of those concepts, however, I think I am one of the last people that had this type of generalist 'apprenticeship' in IT.

    The fact is that people dont gain this type of experience now, if you have a fault with a PC in the office environment, you re-image it or replace it, you dont diagnose a fault with the hardware or software, you just reset things to a known good configuration and move on to the next problem. The office IT department has been relieved of most of it's responsibility and major applications and services have moved to the Data Centre. First line support calls, instead of being attended to by the local office IT person, are now dealt with over the phone from call centres, these call centres are miles from the data centres that would provide a career progression path for those staff, meaning that there are few junior positions in IT support and a massive skills gap to cross in order to work in the complex parts of our environments.

    In the Data Centre itself, the industry has in the past encouraged staff to specialise in specific areas, there are few IT generalists now, people have instead become compartmentalised in to teams according to skills. What this means is that people who are young blood in IT now, do not get as broad a spectrum of experience as their predecessors, what they instead learn is how to implement the specific products that their employers use, often only in the employers specific usage scenario.

    This is a problem, not only because those products themselves become obsolete (and so the skills associated with them have a finite life and value), but because things have already been changing for some years in a way that means that this specialising is already hurting both staff and businesses... things are becoming more integrated and people once again need to have knowledge of many areas of IT as things like Virtualisation begin to require knowledge of many disceplins, including networking, storage, compute hardware, operating systems, services, hypervisor technology and management suites, the problem is that usually most support staff have only knowledge of 2 or 3 of these and if they have been in IT less than 10 years it is harder to explain to them the concepts required to understand the pieces that they are missing, they often lack the experience that provided this base knowledge, by starting their careers in a specialism.

    The question is... how do we fix this problem? Vendor training is not the right approach, because vendor training centres around products, but how do we make sure the next generation of IT staff is able to understand and adopt new technology quickly?

    1. Erik4872

      Re: Demise and rise of the IT generalist

      "The question is... how do we fix this problem?"

      That's a hard one. I've been doing the generalist thing for close to 20 years. And at every place I've worked, there's a huge pressure to specialize. If I actually have a specialty, it could broadly be described as "systems management and integration," and even that's a generalist job.

      Like you've noticed, the big money in IT is the contract mercenary who has an incredibly narrow but extremely deep skill set. In the "systems management" specialty, I've bounced between Altiris, Microsoft's System Center products, CA's (crappy) products, LANDesk, etc. I've had to learn how each of these do things differently, but at the end of the day they're all systems management tools that do the same basic stuff. I work for a professional services company, so I get the chance to bounce around a lot more than I would in a traditional "big corporate" IT role. And yet, I'm not the mercenary type - I have a family and can't travel 40+ weeks out of the year. Those who can, and become absolute geniuses on one or two of these products, make way more than I do. I've worked with more than a few consultants who work many different contracts per year, don't have a permanent residence, and basically live in hotels making multiples of my salary. The downside is that these guys don't really have a handle on the overall picture, from end to end, like generalists do.

      Here's what I think _might_ happen to fix things. Corporate IT jobs are increasingly going to be more about herding the vendor cats to make your virtualized cloud based whatever-aaS work. Getting all of this reliably functioning requires a generalist's training...even if it's not your problem to fix, you need to know enough to tell who's to blame.

    2. Lusty

      Re: Demise and rise of the IT generalist

      I agree that young people are not getting the opportunity to learn, but I'm not so convinced it's a problem. We're moving very rapidly to a SaaS model in all aspects of IT, and that SaaS will be running on PaaS or IaaS in any one of a handful of cloud providers. In the short term, folk like you and I will run things and we won't need new blood. In the longer term there will only be a need for a few good people to run the cloud environments. Thanks to scale, the pay for these people can be very high and so it will be a worthwhile thing to train up for. Most of the industry though will become either programmers where there's a low enough entry barrier that it's not a problem, and DevOps people to get the code from the developer to the cloud platform to provide SaaS. No doubt PC world will by that time have an army of very low skill PC re-imagers as I can't see that function remaining on-prem long term either.

  3. Anonymous Coward
    Anonymous Coward

    "[...] the industry has in the past encouraged staff to specialise in specific areas, [...]"

    Somewhere in the 1990s it became necessary to choose a narrow specialist area if you wanted the quick career progression and pay. Youngsters with the aptitude to be IT generalists soon realised which side their bread was buttered - and went into specialisations or management.

    To show their specialist competence they needed certificates and accreditations. Furthermore they had to re-validate them every few years. The really dedicated ones also had to pursue the necessary learning in their own time. That left little opportunity or incentive to look outside their own specialisation - that was an S.E.P (possibly coloured pink).

    In the end the generalists were usually the oldies who had managed to avoid specialisation or becoming managers. They had cut their teeth when it was still permissible to make a mistakeinnovation that took a whole system down until the wrinkles were ironed out. They had often been involved in pioneering both hardware and software simultaneously.

    That "do it all" generation is now either retired or getting close to it.

  4. Gannon (J.) Dick

    When Dinosaurs Roamed ...

    ... and I was in School, the half-life of an Engineer was said to be 5 years. Of course, there the Mathematicians programmed the computers. Dirty, dead end job, but somebody had to do it. Mathematicians had some sort of magic academic longevity - how many times has that tired 2+2=4 thing been trotted out - in any case they were captured and tasked.

    So, here we are, just where we knew all along we would be.

    It's true. History does not repeat itself, historians repeat each other.

  5. Bluto Nash

    Trevor - I've noticed a couple of your comments and writings of late where you've referenced the "square root of bugger all." A lovely turn of phrase, I must say. I'd like to propose a new acronym for common use - SQROBA - as a quick substitute for it, as it gives the general "gist" while still having a nicely negative sound (ref. "screwed," "scrotum," etc.).

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon