back to article I used to be an Oracle DBA ... but now I'm a Big Data guru

As the demand for Oracle skills fades along with VB and as even Java loses its shine, the smart developer is looking at what will pay the bills for the next decade. As an ITpro you have to bet your career every few years and Big Data is too obvious an opportunity to pass up. The problem being that it’s not a single product …


This topic is closed for new posts.
  1. John Smith 19 Gold badge
    Thumb Up

    Have BS, will travel.

    Highly entertaining but also informative.

    Sadly the dumbos who feel they need this (PHB has been on a conference) will probably hire an agency that will run a keyword search on their CV bank and find the people with the best engineered CV in this area.

    Who will probably be some of the least equipped players to actually do it, prompting the first (next?) round of "XXXX's Big Data project goes titsup" headlines.

    Let the games begin.

    1. Dominic Connor, Quant Headhunter

      Re: Have BS, will travel.

      It's entirely possible that I've been at the same conference as your PHB (he may mention a loud old headhunter who kept asking questions of the speakers :)

      I wish you were more wrong, I asked a good number of people if there were yet any qualifications they'd respect in this field and the response was entirely negative, so it's the standard model of buzzwords and random interview questions.

      1. John Smith 19 Gold badge

        Re: Have BS, will travel.

        "I wish you were more wrong, I asked a good number of people if there were yet any qualifications they'd respect in this field and the response was entirely negative, so it's the standard model of buzzwords and random interview questions."

        Oh that was just my experience of recruitment agency SOP. One of the best recruitment experiences I ever had was on an employer 2nd interview when they put me in a room with a login, a work description, the language manuals and told me they'd be back in 2 hours. It runs. I'm in.

        Of course for that to work management has to understand what they want in the first place.

        I'd agree with other posters it's not the size of any single database table, it's the cross-referencing (across multiple tables/databases/sites/clouds). ETL on the grand scale.

        Not to mention the contents of Damian-in-Marketing's spreadsheet which hold the (very secret) weightings for how to spot a sure fire prospect from a tire kicking time waster.

        And then there's all that "unstructured" data from fapbook (and possibly myspunk) you want to snaffle off the kiddies before they wise up....

  2. Joe Harrison

    Can anyone point me at a good newbies guide?

    I just happen to have some large traditional-style databases within which I suspect there is valuable information I am not finding. I would like to experiment with "big data" (whatever it's called) alternative approaches, hopefully bolstering my CV in the process. Now what?

    1. Ting

      Re: Can anyone point me at a good newbies guide?

      You could take a look at Teradata Aster as the article mentions. This website has some useful getting started guides and you can download VMs to play wth.

    2. Ian Michael Gumby

      Re: Can anyone point me at a good newbies guide?

      Depends on how you define 'big data'...

      Hadoop, just head over to Apache's website or look for Tom White's book (3rd Ed)

  3. Steve Davies 3 Silver badge

    Cue Job postings asking for...

    10 years experience with 'Big Data'

    then anyone who has even smelt Oracle or DB2 out in the wild will apply.

    Only those assimilated by the Borg will mention SQLServer.

  4. Titanosaur

    Mann from heaven for Empty Suits

    "Big Data" is a buzz phrase much loved by Empty Suits. I work in pharma and our IT departments have suffered major Empty Suit infestation. "Big Data" has now replaced "Cloud Computing" as the latest bandwagon for such empty suits to pile onto in the hope of career advancement. One could also add another current buzz phrase: "Data Scientist", many of those who like to bandy this one around know little about data and even less about science. Having said all the above, "Big Data" is important, much too important to be left to gasbags!

    1. Charlie Clark Silver badge

      Re: Mann from heaven for Empty Suits

      Amen to that, brother. One of the keynotes from PyCon a few years back was from Bitly's Chief Data Scientist and I've viewed the term with disdain ever since. I saw Mr Matt "Buzzword" Asay was banging on about data scientists the other day* which I guess is further evidence of the meaningless of the term.

      * I've long given up on reading the badly thought-out, poorly written pieces I just them to prime my internal spam filter.

  5. James 47

    From the job ads I've seen (I'm a C++ dev) the most money these days is in Java for HFT (figure that out) and python/JavaScript for web.

    1. Charlie Clark Silver badge

      I, for one, can't wait for the transaction tax to put an end to this particular branch of "finance".

    2. Ian Michael Gumby

      @ James

      If you can't figure it out then you really don't know much about the languages or how you can do HFT and market making using Java.

  6. Anonymous Coward

    I've been very happy with Oracle as of late...

    Since Oracle's latest (mis)fortunes really helped my (small!) firm on making some rather drastic decisions. We're a small IT company doing a lot of things; main tasks lie in systems administration, website hosting, development (usually customer specific software) and information management.

    Development has always been a rather (very) small portion of it all and because I've been quite experienced with Java that has always been the language which was used. For several reasons, amongst which my own experience, whenever I needed help (hired temporary co-worker) there was hardly a problem finding a Java guy, safety / security (Java actually has a Security manager (link to API docs) build right in which enhances this feel for security) and of course the cross-platform aspect (though it has mostly been Windows).

    Alas; the latest Java mishappens also eventually found their way towards our totally non-technical customers who started asking questions. And that is something you really don't want to happen; esp. since security has always been a major aspect for us. Of course the main issue sits at browser level right now, but for a customer "Java" is "Java"; and right now "Java is dangerous / bad".

    And that's where you get a little forced into a situation...

    And so we're currently migrating. From Java to .NET; the web part is being rewritten in ASP.NET, the desktop parts are likely going to be a combination of C# & VB (depending on external help) and Centos / Glassfish / Apache is being replaced with Win2k8 / IIS.

    Whether this is for good or bad; I dunno. So far its massively more expensive and a huge investment for my company (VS 2012 vs. NetBeans, CentOS / Glassfish vs. Win2k8 (both VPS so the extra costs isn't /that/ high) and not to forget all the work which goes into all this). But a good thing (IMO) is that we're also dropping MySQL in favour of PSQL for this environment. Since we're changing anyway we might as well do it right (and with ODBC / .NET integration the underlying engine doesn't differ all that much).

    Even so; we have Oracle to thank for it. And although I don't quite feel "at home" with C# as much as I do with Java I have to say that the transition so far going smoother than I had anticipated. At least I can continue working with VP UML since that easily embeds itself into Visual Studio just like it does in NetBeans. Without extra costs! You don't have to get VS Enterprise for UML support you know; a mere $100,- is all you need.

    Well, that's one down, how many more to go? :-)

  7. Charlie Clark Silver badge
    Thumb Up

    Good piece

    I guess it's largely prejudice but I am struggling with DBA and VB in the same context.

    Otherwise bang on - if you all ready know what you are looking for then you're likely to best-served by a relational system with the right collection of analytical and statistical tools and skills (eg. knowing what normalisation of both data and distributions are). Otherwise, say you're looking at the background radiation of space, the big data tools have some interesting approaches to finding out what might be interesting.

    1. Dominic Connor, Quant Headhunter

      Re: Good piece

      DBAing and VB are different skills, but quite a lot of people do both, especially in smaller dev teams.

    2. Ian Michael Gumby

      Re: Good piece

      Unfortunately you don't grok 'Big Data'.

      It's not something to be ashamed of. Really. There are a lot of 'Big Data' pundits who don't really understand it either.

      If you want to be useful, you should know both and understand how to put both systems in to play working together.

      BTW, the price of one Oracle license of a large 'enterprise' server would cover the costs of an entire cluster of Hadoop machines along with a licensed support contract. Then add in your hardware to boot.

      That s why people are looking at Hadoop.

  8. AceRimmer
    1. Anonymous Coward
      Anonymous Coward

      Re: You could always

      Yeah, but *Dundee* ffs.

      1. Anonymous Coward
        Anonymous Coward

        Re: You could always

        Hey, Dundee's not that bad!

        I spent a bit of time there a few years ago, and I think it's under-rated. Plus you can easily hop on the train to Edinburgh, one of my favourite cities.

    2. Dominic Connor, Quant Headhunter

      Re: You could always

      Entertaining Chicken/Egg issue, as it's a new course how does one evaluate it ?

    3. Michael Wojcik Silver badge

      Re: You could always

      There are a whole bunch of these programs. A post on the Analytic Bridge site lists several in the US, based on a report from KDNuggets.

  9. Anonymous Coward
    Anonymous Coward


    Quite witty, but still the hoarse ramblings of a worn out pimp.

  10. Michael Wojcik Silver badge

    Stumbling around in the dark

    Pieces like this are always fun - an author does a surface treatment of a complicated and quickly-changing subject, in an area filled with misconception and distortion, and a bunch of commentators who have even less understanding of it chime in. (Though props to a few folks, such as Ian Michael Gumby, for trying to inject a bit of sense.)

    Sure, "big data" is a buzzword that's routinely abused. Ditto "data science" and "data scientist" (though those have been around a bit longer, and before they were picked up by the pundits they did have a degree of specific meaning). That doesn't mean there isn't real work being done under the "big data" and "data science" labels. The presence of charlatans does not mean the absence of genuine practitioners.

    And there's a hell of a lot more to "big data" than wrangling disparate data sources. Classifiers and predictive models, serious statistics and probability, visualization - if all you're doing is glorified ETL and naive-Bayesian, you're not doing "big data" (much less data science) in any significant way.

    People who are interested in actually learning a bit about the real field, as opposed to the hype, could do worse than spend some time on sites like Analytic Bridge. (I have no connection to Analytic Bridge other than reading material posted there.) I'd recommend some of Vincent Granville's posts in particular:

    Fake data science

    The curse of big data

    The 8 worst predictive modeling techniques

    Those help cut through some of the cruft.

    1. John Smith 19 Gold badge

      Re: Stumbling around in the dark

      "The 8 worst predictive modeling techniques"

      Now how many of these are GCHQ going to deploy in their new find-a-terrorist system?

This topic is closed for new posts.

Biting the hand that feeds IT © 1998–2021