back to article It's now or never for old sysadmins to learn new tricks

In most fields of human endeavour the complete invalidation of a person's formal training and skillset generally takes decades, if not generations. Within IT the tools, applications, operating systems and cloud services learned at the beginning of a bachelor's degree can already be defunct before that degree is completed. I …


This topic is closed for new posts.
  1. Anonymous Coward
    Anonymous Coward

    Good article.

    What you mentioned about certifications struck a chord with me. I'm currently doing a CCNA but i'm wondering how far to go after i've completed that. (I.E CCNP/IE etc).

    1. Ceb

      CCNP has a substantial increase in value as an opportunity to enter mid-level contracts inaccessible to those who only have Associate level Cisco certs. Most of the time when you see a CCIE headline on a job offer it'll say in the details that they prefer a CCIE but will take a CCNP with experience.

      1. Aldous

        The CCNP builds on what you did at CCNA and really gets you prepped up for networks (except multicast/mpls). The jump to CCIE is huge though and really requires in depth knowledge of the protocols and loads of lab time. However if you get it then it makes it much easier to find work

        1. Anonymous Coward
          Anonymous Coward

          Thanks for your thoughts guys.

  2. Denarius Silver badge
    Thumb Up

    Good assessment

    Your observations match my experience. Fewer sysadmin jobs around and middleware type crafting required instead. Time to exit for us oldtimers mostly, which is exactly what my older associates are doing. Many have no choice as age discrimination labelled as efficiency measures/cost cutting means jobs are vanishing whether essential or not. After all, customers are not important.

  3. tim_lovegrove

    Although I agree with the main point of this article, there are many environments, particularly in the public sector, where traditional IT methods are still carried out and are very much stil lthe norm. It *will* only be a matter of time before these get absorbed too, but the pace of change in these places is much slower. The MoD in particular is still very anti-Cloud/SaaS/anything offsite for many of its networks and facilities, and until Cloud service providers start to take the security and integrity of such systems seriously it's likely to remain so.

    But I entirely agree with the main thrust of the article - change or die.

    1. Anonymous Coward
      Anonymous Coward

      "But I entirely agree with the main thrust of the article - change or die"

      Always thought of IT as one of the most Darwinian professions :-)

      (Unlike, say, politics, where the dinosaurs still roam - "hello is Mr Dennis Skinner there? It's the 21st Century calling?")

      1. Yet Another Anonymous coward Silver badge

        >hello is Mr Dennis Skinner there? It's the 21st Century calling?")

        Yes a politician that believes in something beyond racking up expenses, lining their pockets, getting re-elected and screwing over anyone that gets in their way is so 19th century.

      2. Anonymous Coward
        Anonymous Coward

        "hello is Mr Dennis Skinner there? It's the 21st Century calling?"

        "hello is Mr Dennis Skinner there? It's the 21st Century calling?"

        Yeah, a politician on the national average wage and no more, no expenses fiddles, who votes with his constituency and his conscience not with the whips, and most of all one without a PPE degree and a network of unpaid "researchers". Very un21st century isn't it.

        Well actually quite a few people outside the Westminster/City reality distortion field might prefer to have more like Skinner and fewer like the indistinguishable Cleggeron/Milibandelsons that are currently ruining the country.

      3. John Smith 19 Gold badge


        (Unlike, say, politics, where the dinosaurs still roam - "hello is Mr Dennis Skinner there? It's the 21st Century calling?")

        But there's still no bite like a bite from an old beast.

    2. Anonymous Coward
      Anonymous Coward

      @tim_lovegrove " It *will* only be a matter of time before these get absorbed too"

      In the public sector I'm not entirely sure this is true, at least not as much as elsewhere. Particularly, education and county councils etc.

      The public sector, more than anywhere else, has a number of factors that tend to prevent these kind of changes.

      - Greater requirement to prove they are adhering to data protection standards etc. Which in a lot of cases means keeping data and services in house.

      - Greater need to be able to, in the event of failure/data loss etc, be able to investigate and explain the cause of the problem. Rather than just point the finger at a supplier and shout and stamp their feet. Sure external agencies can perform such investigations, but they are less likely to actually care about getting answers. When your cloud based solution hosted by MS, Amazon etc breaks or is unavailable I'd like to see you get a better answer from MS, Amazon etc than "there's a service outage and we're working on it".

      - Finally, and probably most crucially, public sector has a greater tendency to promote from within when it comes to management jobs. Which means that when the current CIO/IT Director, who is fond of the old ways and keeping things in house, retires or otherwise leaves he will likely be replaced by a subordinate or a lower paid peer from a similar organisation who will be of the same/similar generation and mindset.

      Of course there will be some change in public sector. It's unavoidable, but I think it's quite likely the rate of change will be practically glacial in comparison to other areas.

  4. phuzz Silver badge

    There will always be a place for someone dealing with/explaining to end users. Which is a shame because I never liked dealing with end users :(

  5. ecofeco Silver badge

    Not eveyone can code

    Hell, a lot of coders can't code! :)

    Despite all the automaton, automation can and often does still fail and software is often badly written.

    In the end, where the "rubber meets the road" is still the providence of a live person fixing the problem. Someone still has to load, configure and troubleshoot the automation. Most coders I've met don't really understand the hardware or system quirks they are writing the software for so still need someone who is more of generalist. Those would be sysadmins.

    Nonetheless, the author is correct and the move to sysadmin automation is moving ahead.

    1. Sooty

      Re: Not eveyone can code

      in a lot of cases automation can contribute towards badly written code. It does so much for you, people who barely know what they are doing can be let loose on the systems and you'd be amazed at what sort of damage they can do.

      1. I think so I am?

        Re: Not eveyone can code

        Like turning off all your ATM's?

  6. Anonymous Coward
    Anonymous Coward

    I would agree with your comments about a CIO having technical experience is important in a small to medium company, however my experience with large companies (FTSE100 size of large) is that a CIO form the business, with appropriate advisers from an internal consultancy team is a very valuable thing. It's two different ways of cracking the same egg, in the small to medium case, you don't have enough resource in the company to have technical advisers for the CIO, so a technical CIO who understands the business is a good thing. However, in a larger company where you have resource to advise the CIO, having a CIO from the business who really understands the different requirements is also very useful.

    The key is that either CIO has to have a very good level of understanding of one of the two key skills and an ability to learn from others around him about the one that less is known about.

  7. Do Not Fold Spindle Mutilate
    Big Brother

    Very true and applies to database administrators as well.

    The ideas are very true and apply to database administrators. I chose to retire because the cuts were happening without the proper software to automate the monitoring and updates. Also Oracle was directly telling upper management that their software would do all the work of a dba. May I suggest a slight difference in attitude between us. The article seems to imply that the cause is the financial cuts from the customers but there should also be the technical person automating there own job out of existence. If similar work has been done multiple times then automate as much as possible and move on to larger problems. But what are those larger problems? Perhaps much better failover / backup / recovery or quicker security patching, or better testing / auditing to insure high quality software infrastructure?

    A real problem is the market place seems to change quicker with Apple and Microsoft causing massive changes in the OS and interface in a relatively short period of time, it becomes difficult to set out a new path of what software product to learn.

    There will always be a need for someone to install the NSA secret monitoring software.

    1. Smileyvirus

      Re: Very true and applies to database administrators as well.

      Or not :)

    2. TheOtherHobbes

      Re: Very true and applies to database administrators as well.

      "But what are those larger problems?"


      Good article.

  8. MQADM

    An interesting article. Truly successful IT staff have in-depth knowledge and experience across multiple disciplines and architectures. Pidgeon holing ones self to a single effort such as networking, databases, etc, is a sure way to limit career opportunities and employment longevitiy. No matter what the career, failure to adapt to changing needs will soon have you being shown the door. In my personal experience, those who remain hungry to learn and put in the personal effort to do so will thrive. Those who don't, won't.

    1. Trevor_Pott Gold badge

      And yet, for the past 15 years those who've made the most money in IT are the hyper-specialists. That's changing, slowly.

      1. Anonymous Coward
        Anonymous Coward

        > And yet, for the past 15 years those who've made the most money in IT are the hyper-specialists. That's changing, slowly.

        Those that make the hyper-money are usually the ones that are shown the door first when harder times arrive.

        Question is: do you want a secure future making a reasonable income by being flexible, generally knowledgeable and willing to learn, or make a short term, possibly tenuous killing while realising that you make yourself a target when your employer discovers a way of doing without your expense? High fliers' astronomical wages are tolerated because they bring a much needed skill which comes at a price. However, it's difficult to drop someone's salary when they become less useful. It's easier to just show them the door.

        In my experience, in software dev, the ones that last are the ones with a good overall and fundamental understanding of computing issues and can turn their hand effectively to a number of different disciplines.

        This usually comes from a good educational grounding which covers the eternal fundamentals in our field. When the shit hits the fan, then these guys are often the last guys standing due to their flexibility and that they know how everything works having been around a long time.

        1. Trevor_Pott Gold badge

          I'm a contractor. Crazy stupid money, please.

          1. Rukario

            I'm a contractor. Crazy stupid money, please.

            I've heard Cow-gary (Oil-gary? Harper-gary?) is the place to be for that. Certainly not Vancouver.

  9. Magister

    Have to agree

    Trevor makes some great points and I do agree with almost all of them. Many companies are going down the route of either outsourcing the function, or relying on cloud based systems. Their arguments are that it is cheaper, they don't need the technical staff etc etc.

    But the place where I am working is a classic example of how badly it can go wrong. The IT is a shambles; no-one is managing it or taking responsibility for anything. Service delivery is abysmal; basics such as getting DNS actually working seem to be beyond their capacity. Even thought they are obsessed with paperwork, there doesn't seem to be any managment control; as a result, they regularly buy hardware that's expensive, unsuitable and often sits around unused for lengthy periods of time.

    What's worse, the "strategy" is being set by people with only the most minimal of understanding of IT; and they are screwing up so badly, I cannot believe that it won't be long before the whole thing just dies and they go belly up.

    I had thought about starting my own MSP for small businesses; I'm damn sure that I could provide a significantly better service to SMEs than what I've see in a number of places. But I'm starting to think that maybe I should just call it a day; get a job stacking shelves for the next 5 / 6 years to keep me active, then claim early retirement and live out my life wandering around in a drunken stupor somewhere warm.

    1. Redsyrup

      Re: Have to agree

      Totally agree. Avoiding the cloud is the best way to maintain job security. If you're just another cog in the machine you're easily replaceable. Furthermore who can trust what they can't see or touch?

      1. PJI

        Re: Have to agree - cloud

        Re cloud, trust etc.. To anyone with half a brain who reads newspapers, listens to news and so on, it must be clear that the "cloud", being simply a marketing term for a distributed network of servers and software for remotely stored data and, possibly, applications, is a major security hole. Add to this that the most widely known and used "clouds" are in American (USA) hands, using servers on USA territory, so subject to USA laws requiring full access. So, just who would even dream of using such a service for anything other than the most trivial data or work? Any firm, or even any individual with interests in anything outside stamp collecting, would be criminally ignorant and carelss to expose their data to such a system.

        The danger of mainly automated system administration, basically in the hands of a part-time IT/most of the time business manager, is that they will be unaware of what is where and their dependance and vulnerability.

        So, we can not and should not waste time fighting it. We need to devise automation that is not synonymous with obscurity and loss of responsibility and either think again about "clouds" or, as I believe the EU is considering, make sure that there are good, secure, non-USA clouds, secure business clouds, private user clouds, encrypted clouds and so forth.

  10. Robert Carnegie Silver badge

    This may not be relevant,

    but we're still waiting for our "infrastructure team" to take the guy who quit two weeks ago out of our development team e-mail address list. When we send messages, we get an error reply back because his account has been deleted.

    So maybe they -can't- take him off the list, now. Ohdear.

    1. Anonymous Coward
      Anonymous Coward

      Re: This may not be relevant,

      "but we're still waiting for our "infrastructure team" to take the guy who quit two weeks ago out of our development team e-mail address list."

      Maybe they've got more important things to do first?

      1. Captain Underpants

        Re: This may not be relevant,


        Perhaps I'm failing to catch some sarcasm in your post, but if you think that the effective updating of internal systems to remove staff who have left the company from internal distribution lists isn't a problem, you're not thinking correctly. Most of the time, updating someone's mailing list memberships is done at the same time as administering their permissions. So if they haven't done that, there's reason to worry that said person's account is still active with the same access as when they were still employed... which is a fairly bad process failure.

        1. Anonymous Coward
          Anonymous Coward

          Re: This may not be relevant,

          @Captain Underpants

          Apparently you didn't read the part where he says they get bouncebacks because the account doesn't exist.

          But you keep imagining your imaginary problems, and I'll keep thinking his infrastructure team might have more important things to prioritise than email distribution groups. Deal?


          1. Captain Underpants

            Re: This may not be relevant,

            If the process says it's the infrastructure team who deals with this stuff, the process is wrong. Doesn't mean them failing to adhere to the process helps anything, that just means everyone's working in an ad-hoc fashion. And, you know, that'll only ever have awesome results...

            Tl,dr: there's scope for everyone to have a scoop of the blame cake here.

  11. Duncan Macdonald Silver badge


    If you are old enough and have a good pension plan then the best route may be to retire.

    System admin jobs are going the same way as most coding jobs did - either automated or outsourced to cheap countries. The only support jobs likely to continue for a while are the junior IT support - swapping keyboards and mice - swapping PCs - replacing toner and paper etc but these jobs pay peanuts. For most companies under 1000 employees, there is no business reason to have their own system administration team if they can get their administration done by an outside group for less. Using the internet, most system administration can be done just as easily from 5000 miles away as from 50 feet away.

    1. Redsyrup

      Re: Retire

      Are you kidding? 1000 employees and nobody on site? Or one person from the read of your post? Impossible my friend.

    2. Anonymous Coward
      Anonymous Coward

      Re: Retire

      An "automated" system admin? I'd like to see that... Me thinks you don't know what a real Sysadmin is.

      1. Suricou Raven

        Re: Retire

        More accurately put, automation means that a system admin can administer a much larger, more complicate network with many more servers.

        The company still needs an admin.

        But they no longer need *two* admins. Someone is now surplus to requirements.

      2. Anonymous Coward
        Anonymous Coward

        Re: Retire

        Did you notice this link?

    3. Captain Underpants

      Re: Retire


      I've seen a fair bit of how well that works from a friend working in a school that decided to go down that route. Now, instead of their "expensive" on-site sysadmins (2 guys, from memory) they have a helpdesk number for a helpdesk based in Norway (I think), with something like a four-hour response time. As opposed to the "how long it takes to get to the classroom from the IT office" response time when they had proper on-site support.

      Don't get me wrong, there's scope to do outsourcing of that sort properly, but it's not at all cheap to do properly.

  12. Alistair

    The problem with specialization

    Part of the issue is that with specialization, ones focus can become too narrow.

    If you happen to be the guy/gal who is specialized too narrowly, its possible to have the specialization override your view of business process. This immediately makes you less valuable.

    25+ years into the business I realize that my utility to my customer(s) comes from a very broad range of knowledge and the ability to understand the business processes that these customers (use/need/rely on). I'm far from afraid of adding to my pool of knowledge, indeed, quite happy with learning more.

    Automation, be it puppet, cfengine, chef, or some proprietary tool(s) from a large commercial vendor are absolute requirements. And being willing to dig deep and hard into *small* issues to resolve them before they become large ones (and recognizing which small issues will blow up) is a skill that one needs to hone as well.

    As with all tools, however, your automation can and will break. If you don't have the patience or skills to tear it apart and find out WHY it broke, you are in very very deep trouble. Usually your automation has allowed your task pool to expand far beyond your ability to fix things *manually* by the time it breaks.

    A broad general skill set and a very healthy understanding of the business processes your customers use and need, and strong automation skills. This is what will keep you employed in my view.

    But then I could just be a cranky old fart with no patience left.

    I am currently on vacation ---- I think I need another drink.

    1. h3

      Re: The problem with specialization

      Don't see why people need these automation tools. (If you understand the OS properly you should be able to code what you need). Using someone else's tool makes it harder than just writing something from scratch in a suitable language. (Lisp or ocaml even perl). Powershell if it is Windows. (If you are a big enough company have someone who can make everything work with Powershell).

      Automation is ok if you understand everything that the tools are doing. (Pretty easy with stuff like Solaris Jumpstart very difficult if it is some Windows tool that does everything using COM objects).

      I have not done that much with it recently I did Jumpstart stuff for thousands of servers. (And worked under a really decent contractor who used Smartstart and could do anything anyone could want with it in the time before anything was easy).

    2. This post has been deleted by its author

    3. Solmyr ibn Wali Barad

      Re: The problem with specialization

      Weber's Definition: An expert is one who knows more and more about less and less, until he knows absolutely everything about nothing.

  13. Ambivalous Crowboard
    Thumb Up

    What he said

    I've taken a slightly different approach. Being of the tender* age of 30^H^H31, I'm in a position where my house is being sold, I've become a single man, so I've taken a look at the national occupation shortage register, and decided on a completely different path.

    Sure, I could become a developer, or a website bloke - and to be honest I already run a small MSP and have a few tech support customers. But I'm just not feeling good about IT any more. Perhaps a career in social work, where I get to deal with actual people, may be more engaging...

    Four years at college/uni doing a people-related qualification (for free, cause there's a skills shortage in this field) and I can then go into teaching, assistant/deputy/headteacher, or do social work for £30 an hour (as a friend has just done).

    Interesting times ahead!

    1. Dr Who

      Re: What he said

      Aha! So you've just become single and you're looking to spend four years at university. I can certainly see where you're coming from there.

      1. PCPuss

        Re: What he said

        Ha! Got the T-shirt...

    2. Suricou Raven

      Re: What he said

      "can then go into teaching"

      Be warned - teachers are severely overworked. Not only can you expect to work the normal eight hour day, but another few hours a day on top of that at some, plus a sizeable chunk of your weekends, just to keep up with the work piled on you.

    3. Anonymous Coward
      Anonymous Coward

      Teacher shortage

      I'm nearly twice your age. A few years ago, when I had 10 years left to "classic" retirement age, I responded to one of the ads offering a route into teaching for those with IT experience. The ad specified a first degree with >50% IT content.

      I have an Elec Eng degree from Imperial, '73 - '76. I focussed on the computing and "light" electrical (as it was known then) components of the course. I got an "A" in the 2nd year optional computing course. My final year project (25% of the degree) was programming a PDP11 vector graphics device. I also got an "A" in 3rd-year digital circuit design. Apart from the first 11 months post-grad, my entire career has been in IT hardware and software. Including hardware and software in the high-energy physics group at Imperial.

      However, this wasn't good enough for the bureaucratic numpty I spoke to. 40-odd percent 1st-degree being IT-related ddn't reach the bar. No discussion. No allowance for the nature of "IT" in degrees being very different back then.

      Oh well, not my loss......

      I wouldn't recommend social care, not children and families anyway. When it all goes mammaries-vertical, your chances of being selected as the scapegoat will be very high. Foster Care is a worthwhile option, but carries equally high risks - though they are more manageable if you have your head screwed on right.

  14. Ceb

    I've been involved in contracts since I left the military and recently added technical instruction to my repertoire. I've found that being in the business of instruction keeps me from getting pegged into a specialist's role and eventually eliminated from the market. When new technologies come down the pipe I have word they're coming and time to prepare. As well, being an instructor relies on charisma as much as expertise, so I found myself nodding along to what I saw in most of this article.

    As stated in the article though staying in isn't for everyone. In this industry you're constantly learning new information. While it's certainly true that older technologies drop off the plate eventually, you certainly pile it on a lot faster than it falls off.

  15. Anonymous Coward
    Anonymous Coward

    What a load of nonsense, another trolling article on a friday to keep the reg site ticking over over the weekend.

    If only it were as simple that a dev could write an app and plonk it down on a cloud service and scale to thousands of users but back in reality this takes people with skills to work out the kinks and keep running. They are slightly different skills than before but in IT which job can you do that doesn't require constant learning?

    1. N2 Silver badge


      There are still plenty running NT4, 2000, 2003 & other classic servers in their domains who dont give a shit about the modern crap because its expensive, all fucked up & weve got to change this because it wont run with... (substitute expensive program of choice). They just want something that works and people who know how to run & fix it.

  16. Anonymous Coward
    Anonymous Coward

    Say it ain't so

    But if you've tried dealing with the companies that are "automating" everything then you'll eventually realize that the entire edifice is a pack of cards, nobody understands the entire picture. This works well when everything works well but as soon as something steps out of bounds the results can be catastrophic.

    For example, Snowden doing a runner with data - that was very predictable but by all accounts NSA was caught completely flat footed and plans to solve the problem by further automation and elimination of sysadmins. Hands up anyone who thinks that this will actually solve the problems?

  17. Anonymous Coward
    Anonymous Coward

    Good times

    I actually am quite optimistic.

    I think it's an exciting time - the whole DevOps movement has given the industry a kick up the proverbial. The move to the cloud has opened up the opportunity to think about new ways of providing support. The APM space is moving on quickly and giving opportunities for support teams to demonstrate what they can do in real financial terms. Big data techniques are offering the potential to find (and solve) failure patterns in a new way.

  18. Henry Wertz 1 Gold badge

    Do they think they'll do it themselves?

    So, I've seen plenty of CIO positions where the person is (in my opinion) quite overpaid for what they are doing, and have subordinates reporting to them who seem to do all the real work; in these cases that CIO is really rather redundant.

    But, by the same token, one could say "Why does your company have a CFO? Finances are even more mundane and standardized than IT, so a CFO is not needed." "Why does your company have a CEO? After all, you have a board and department heads that keep things running, so CEO is unneeded." It sounds silly when put that way.

    But, the thrust of this article, and the ones linked, seems to be that there doesn't need to be anyone around at all to make IT decisions, that bosses are now plenty technically savvy to decide for themselves.


    For on-site services, I don't expect some random manager to REALLY know a about capacity planning, SANs, the higher performance LAN components, and so on; I don't expect them to know that some kind of RAID setup is NOT a backup; I don't expect them to have kept up on reports of some particular product having problems before they choose what product to purchase. For managed services (i.e. "cloud"), I don't expect a random manager to have the time to look into whether a particular cloud provider is REALLY following good practices for data redundancy, backups, security, and so on. Obviously, something like GMail there's no worry about this... but for other services, they may end up picking the lowest cost solution, only to end up using a "cloud" that is actually a single rack-mount computer. It's not unheard of! I think this is a potential recipe for disaster, with vendors taking these not too tech savvy executives for a ride (this does happen even with a CIO but is easier without one.)

  19. chris lively

    I hate to break this to you, but the problem you are pointing out has been happening for a long long time ( decades ).

    Here's a clue: guys who configure stuff are always going to be a "dying breed" as that stuff becomes more robust and easier to deal with. When a new tech comes out that needs twiddling to work right, then people will be needed to twiddle the dials. The number of those people will be commiserate with the success of the vendors sales group. When that tech starts twiddling its own dials, those "engineers" will need a new thing to fuck with.

    If you want a solid job: move to the side that actually builds new stuff. You'll have a job for a long long time.

    1. Anonymous Coward
      Anonymous Coward


      "If you want a solid job move to the side that builds things" - yeah thats never been outsourced to india, china, poland, Ukraine, <insert your country here with smart people who work for less money/>

  20. gerdesj Silver badge

    I do that

    I do a similar thing as the author - have done for quite a few years: 200 odd customers and you will have heard of several of them. 20 odd staff in the firm. t/o ~£1M. I am the MD.

    My advice:

    * Read the manual and/or the source code if it is available - whatever it is (yes I'm a bloke but it helps)

    * Don't get hung up on certification: I *will/do* bin a CV that mentions MCSE without evidence of intelligence. It's evidence of lack of originality. CCxx is nearly OK amongst other qualis for various reasons.

    * If you are wondering what to do at Uni - don't do an IT quali. Do engineering or maths or anything apart from IT. I can teach you how to program or whatever I need but I'd rather you had some useful extra knowledge. Besides if it turns out IT doesn't work out for you at least you have another path.

    Well, that's my own personal point of view - oh and BTW I don't see things as quite as bad as the author of the article.

    That's the thing about personal experience - IT'S NOT FUCKING DATA - it's just personal experience.



    1. Getriebe

      Re: I do that

      @gerdesj I agree with all you have said - and I wrote a similar post a few months back. I run what might be called a managed services / hosting company for a big software house - so similar to you but I don't own it.

      I recruit on the person and not qualifications, after they have gone through some initial rounds of weeding I will ask them some questions myself and its not the absolute answers its the approach and methodology and also the willingness to engage with the customer, I look for.

      I do see things as bad as Tevor for the sysadmin who has no business skills or who is not involved with business process.. Knowb twiddling as the person above said is going to constanty die - where are ostlers these days? - but being involved in the money making process will always be key.

      Also to refer to a fllowing post - we do not get involved in hardwrae at all. No real money in it - its knob twiddling again - its the setting up of the business processes, managing them, improving them and helping the customer plan for their future that creates the cash.

  21. GreyWolf
    Black Helicopters

    Cloud is so over

    or at least, cloud provided by a third party (see recent news about NSA, GCHQ, ISPs and telcos spreading their legs). The Data Protection Act 1998 (and EU legisation) makes it illegal for you put personal data into the care of a cloud provider that is not wholly compliant within the correct jurisdiction.

    OTOH, having your own cloud implementation makes sense if (1) you have the problems that cloud solves (2) you have the expertise in-house to design and build it. Few places have (1) and (2).

  22. Bob500000

    There's a few inherent issues to cloud technology.

    Firstly, The advantage of a custom in-house secret-sauce solution is that it gives your product or service unrepeatable value; the disadvantage of going with PAAS or SAAS is anyone can get the same thing for the same price. That won't help you differentiate yourself in the market. So while competition is in fact driving adoption, once the market has changed you need differentiate yourself or someone else will. And that means you need to adapt new products and innovate.

    Secondly, External SAAS\PAAS are minimally configurable for the client; cloud providers aren't going to add your secret sauce to the mix for you because that increases cost of automation. Additionally, if their staff will sell your secret sauce to your competitors, they will.

    Thirdly, there needs to be a well-documented change-control process for the given automation. A Decentralized control system sounds like a great idea, but when it comes time to modify that automation it becomes a Political nightmare. A key sign of that break down is the users manually managing credentials (handing each other their passwords) to work around the technology.

    Fourthly, cloud providers are risky; what business continuity plans does the provider have for economic, political, environmental, or institutional disasters? If your provider is gone, tomorrow, and it's all over but the cryin', can you survive? What about corporate espionage e.g. being targeted by a spammer? As competition increases dangerously underfunded companies become more difficult to spot.

    Fifthly, ALL Automation is inherently customization, and to support that properly, someone needs to be there who understands the system. For each measured unit of automation added, complexity doubles. This excludes adoption in some areas, and accelerates it in others. The only time this isn't an issue is when the automation is standardized or templatized.

    Finally; outsourcing has it's limits, as the fast food and retail workers unionization shows, people only allow themselves to be outsourced and downsized so far and then they get into the habit of fighting and fighting back hard. I'm of the opinion we are in for a gargantuan market crash where massive amounts of debt will be discharged in bankruptcy and social unrest will reign supreme for awhile because the political class has been too cozy in doing a poor job. Believe me when I say the first people to go are the outsourced firms because they become too risky and are unable to adapt to either rapidly changing architecture or rapidly growing companies.

  23. 404

    Great time for a morality(?) question, Mr Pott.

    Like you, I've taken the managed services provider route, started my business in 2005. I saw the writing on the wall when I, a senior systems administrator at an ISP in Arizona, was laid off. Surprised the hell out of me, life had been great up to that point. Then there were no jobs. Pure survival led to the creation of my company.

    I have always had a personal issue with ripping off clients so I put a policy in place from the beginning that my company would not provide equipment because I consider it a direct conflict of interest. We would determine the requirements, shop for the best price, set up the purchase for the client, and never make any money on it. This way the client has direct control of warranty issues if they occur - this is a black hole with 3rd party purchasing, ever tried to track down a company that somehow fell into the abyss for warranties? I make my money installing and managing networks, not upselling a hardware store with a server cluster. Lost a bid with an Amazon fulfillment center by a $1000.99 because the winner used copper-clad aluminum CAT5 instead of solid copper, I know because I subcontracted the install from the SOB lol.

    Anyway I recently lost an initial contract (not fully out, I provide the best service in the area) with a company that insisted I profit/steal from them - wanted me to purchase equipment and slap ten percent on each piece. I actually felt uncomfortable about it and made me kinda sad, but there it is, couldn't do it and explained the why and how my company could help them by keeping costs down. No effect. Looked at me like I was crazy.

    My question to you is: Am I crazy?


    1. Demosthenese

      Re: Great time for a morality(?) question, Mr Pott.

      Yes, you are crazy.

      The advantage to the client of buying both hardware and software from a single source is that there is no opportunity for the hardware and software suppliers to blame each other for a problem and neither be willing to fix the mess.

      1. Anonymous Coward
        Anonymous Coward

        Re: Great time for a morality(?) question, Mr Pott.

        "The advantage to the client of buying both hardware and software from a single source is that there is no opportunity for the hardware and software suppliers to blame each other for a problem and neither be willing to fix the mess."

        I presume you've never done business with HP, in a situation where the HP software isn't wrong and the HP hardware isn't wrong but the results are?

    2. Trevor_Pott Gold badge

      Re: Great time for a morality(?) question, Mr Pott.

      No, 404, you aren't crazy. I see the same things here. There are really only two paths forward on this: get engaged with some professional marketing wonks and start building a story around your company that says you are quality providers, not low-ballers. This doesn't mean you aren't SME friendly, but it does mean that the SMEs who have had bad experiences with their current guys and who are willing to spend a few quid more for decent service choose you.

      The other path forward is to pull the eject handle. Leave the MSP racket behind, and with it, systems administration. I have to admit that this is what I'm in the process of doing. I maintain a few shops on my roster, but I"ll be honest when I say that's largely so that I have legitimacy when I write about this stuff. Writing and marketing are way better money than systems administration or development.

      Ideally, I'll get another good marketing client and that will give me the buffer room to hire a sysadmin to handle the scut work. Then I think I can take on a few more sysadmin clients than I have. Honestly though, I think the days of "just being a sysadmin" - even as an MSP - are coming to an end in the SME space. It's hard to compete with the scum-sucking low-ballers on the local scene let alone the encroachment of over-the-net types from abroad.

      If you have a sense of ethics and want to do well by your clients, but you play in the SME space then it may just be that in many cities you simply can't make a decent living any more. Each region will be different, so YMMV.

    3. dajames Silver badge

      Re: Great time for a morality(?) question, Mr Pott.

      My question to you is: Am I crazy?

      Maybe, maybe not. I applaud your atttude and I can understand your desire not to get stuck in the middle of warranty disputes ... but as you already take the responsibility for specifying the hardware you will still pick up some of the flak if things don't work correctly.

      I think I would go back to the client and insist on a higher percentage. Say that it's not worth your time getting involved in the paperwork of ordering kit for your customers and dealing with warranties, etc., for less than (say) a 25% markup when your time can be better spent carrying out tasks for which you are uniquely qualified.

      You wouldn't do this expecting to get 25%, of course, you'd do it get the client to see sense and take on the purchasing responsibility himself.

  24. jake Silver badge

    Old sysadmins don't need to learn new tricks.

    We're the folks inventing them.

    HTH, HAND.

    Hint: Learn COBOL and Fortran ... it'll see you through your retirement years ;-)

    1. Destroy All Monsters Silver badge

      Re: Old sysadmins don't need to learn new tricks.

      Useless. You want the guy who KNOWS ABOUT THE DOMAIN, especially for Fortran. Of course, being bored enough to put a fork into his ballslearn about Fortran (which frankly has no redeeming feature left and is only kept because of the undead codebase) helps.

      Yeah, ship in someone who learnt Fortran to check out that Finite Element Analyis code, I dare you..

    2. Getriebe

      Re: Old sysadmins don't need to learn new tricks.

      The COBOL comment - oh so true. We have some legacy stuf based on Cobol - goos specialist software - the two people who maintin the core COBOL are creaming it - they are in their 60s but we an't get replacements. We will need COBOL programers for 10 years minimum.

  25. Michael Sanders

    Watson Admin

    Dude. Just wait until IBM rolls out Watson 2.0 as your sysadmin. Add a small service contract to wipe up any possible mistake. And that's all she wrote. Novell had an awesome transaction tracking system. Deploy something like that and the phone support guy just hits rewind. But realistically Watson isn't going to make a mistake. He will have the entire technet library coded into his brain. You didn't think they just wanted to win at Jeopardy did you?

  26. FrancisT

    Move to (network) security

    Security issues aren't going to go away. In fact in a cloudy, BYOD sort of world they are even more important. I think many sas admins (and particulalry network admins) could usefully move sideways into network security. Exactly how you position yourself (as a MSSP? consultant? in house expert? other...) will vary but there's a crying need for people with clue about security issues and many of the security problems are ones that a regular admin has been handling for years in an intranet/local server environment

    1. Trevor_Pott Gold badge

      Re: Move to (network) security

      The need for security people is great. The number of those willing to pay for security people is few.

  27. Trevor 3

    I have recently found that System Administration is incredibly boring, and I was worried about the risks to that role as I seemed to be writing the scripts and the programs to automate my job. (If I can do it, someone with more business acumen can do it too and sell it for load of money to all sorts of firms)

    I decided that instead of "doing IT" I would do "Applied IT". There are computers in everything nowadays, so it can't possibly be (and isn't) beyond my wits to learn new software/applications/services etc... so that's what I did.

    I have learned a great deal where I am now, and applying my IT skills. I am working a lot harder though ;-)

    Now I can say that I don't work in IT, I work in broadcasting - which is a damn site more interesting and people talk to you...even if they do try and weasel free Satellite TV out of you ;-)

  28. Patrick R

    I could lose my job simply...

    Because El Reg keeps putting nudies on the main page. How old are you ?

  29. loosebruce

    The minute you start using buzzwords like Big-data, cloud computing you are done for!

  30. OzBob

    In the words of Arnold,...

    "Judgement Day is inevitable". So, innovate slower, just enough to show progress, and suck it up when you have to throw it out every 2 - 3 years and essentially redo it again, only slightly better. If we innovate too fast, we will do ourselves out of jobs quicker.

  31. jcuk

    Interesting article and I think most in the private sector would agree with these comments.

    Someone mentioned something about the MoD not wanting to put their data on to cloudy based service providers - quite understandable. This is actually the case for a lot of private sector businesses where their clients data and reputation is their lifeline.

    I reckon the best route for sysadmins to go down is setup shop and target a specific type of client, focusing all your work on what their needs are. This way you create your own niche, building up a trusting relationship with those specific client types and monopolising that individual market with your solution. Defense, for example - understand why they don't want to evolve their technology and help them understand how *you* can eliminate those negative feelings.

    From what I can see, sysadmins need to jump ship immediately. Those in a more secure position should at least widen their skill set and climb on to the next big marketing balloon.

    A sysadmin today needs to be able to sell, people don't look for things now, they expect them just to come through their inbox with a fat BUY-ME-NOW sticker on the front.

  32. nrunge

    Confusing Jobs and Roles

    The natural evolution of technology makes this type of article incredibly easy to write at any point in time. Technology "x" eliminating job "y" articles predate the internet. The sad thing is that we continue to bite.

    Frankly there is too much noise in the data and not enough signal.

    For instance these types of predictions don't take into account the increasing global need for technology as a whole that creates new roles as old ones are no longer necessary.

    Which is where I make my main point. You say jobs but you really mean roles.

    I can say this for a fact because IT employment data is not indicative of any epidemic. I would challenge the author to provide quantifiable data that shows entire jobs are being eliminated rather than the natural re-cycling of roles.

    So don't sweat it IT Pro's. The change is usually so gradual you will evolve your skill set and not even realize it.

    The author of this article makes it sound as if changes like virtualization appeared overnight and now you we are behind the curve.

    That is simply not the case. As long as you have embraced change your skills should be just fine. If you have not then I would argue you never belonged in IT to begin with.

This topic is closed for new posts.

Biting the hand that feeds IT © 1998–2021