This discussion is archived
1 2 3 4 Previous Next 45 Replies Latest reply: Oct 9, 2012 4:32 AM by user8983130 Go to original post RSS
  • 30. Re: Are the oracle DBa's seeing an end to their career
    6363 Guru
    Currently Being Moderated
    jgarry wrote:

    I want my flying car as much as the next guy, but...
    Start saving up then.

    http://www.hammacher.com/product/11812
  • 31. Re: Are the oracle DBa's seeing an end to their career
    Mark Malakanov (user11181920) Expert
    Currently Being Moderated
    And with each one, people were predicting the new features would mean the end of DBA's ........ ;-)
    With every one of such features a fraction of DBA disappears. DBA looses some duties and have to compensate this either by taking other ones or by increasing volume of duties that left - simply, take more databases. That means for Business "we have more data and less DBA heads - less salary and other human related costs".

    I agree amount of data also growths. But automation and ergonomic of data management tools also improves at same time.
    The question is, in this fuzzy and shaky equation, what will overweight - the data growth or automation and data management improvements - this will dictate a demand on DBAs.
  • 32. Re: Are the oracle DBa's seeing an end to their career
    jgarry Guru
    Currently Being Moderated
    Thanks, I didn't know [url http://www.terrafugia.com/about_us.html]first customer delivery of a Transition® Roadable Aircraft is expected to occur in late 2012.
  • 33. Re: Are the oracle DBa's seeing an end to their career
    Mark Malakanov (user11181920) Expert
    Currently Being Moderated
    But what that article doesn't say is how long it took for the Luddite movement to get started.
    I was not talking about the Luddite movement, rather about factors that caused it.
    About similarity of these factors with our topic in regards of how automation and technological improvements can impact workforce market, specifically how these can impact a demand on DBA positions.
  • 34. Re: Are the oracle DBa's seeing an end to their career
    jgarry Guru
    Currently Being Moderated
    user11181920 wrote:
    But what that article doesn't say is how long it took for the Luddite movement to get started.
    I was not talking about the Luddite movement, rather about factors that caused it.
    About similarity of these factors with our topic in regards of how automation and technological improvements can impact workforce market, specifically how these can impact a demand on DBA positions.
    So I'm pointing out that yes, the factors impact it, but no, your time frame is off. Economists define "obsolete" differently than most of us. In a demand life cycle, greatest profit often happens when the product is economically obsolete - what we would call "mature," that is, not on a steep development curve. So not only will the demand increase for DBA's as these automation developments happen, it will increase as they mature. Some time after that is when the demand crash will happen, but that is still a long ways off. It is still possible you are right, but even so, it mostly impacts those dependent on specific skills, like those OCP's who only know the GUI (I was a DEC specialist when it fell over so spectacularly in the '80s, but it wasn't so much of a problem for me because of the transferability of the more generic programming and relational design skills, and of course, Oracle experience).
  • 35. Re: Are the oracle DBa's seeing an end to their career
    Mark Malakanov (user11181920) Expert
    Currently Being Moderated
    >
    So I'm pointing out that yes, the factors impact it, but no, your time frame is off. Economists define "obsolete" differently than most of us. In a demand life cycle, greatest profit often happens when the product is economically obsolete - what we would call "mature," that is, not on a steep development curve. So not only will the demand increase for DBA's as these automation developments happen, it will increase as they mature. Some time after that is when the demand crash will happen, but that is still a long ways off.
    >
    Interesting.
    But as you can see from previous posts DB/DBA ratio growths quite fast.
    Which means that a latency of business reaction on innovations is not too big.
    I'd say it is 3-5 years now. May be too optimistic, but not 50 years. And it seems shortens.
    In many organizations HR constantly watching for these "industry average" metrics and make efforts to comply, otherwise there is a risk to become not competitive.
    Managers jump like crazy into technological and organizational solutions that promise to drastically reduce costs. It is not that century when you have to wait a result for 50 years.
  • 36. Re: Are the oracle DBa's seeing an end to their career
    BillyVerreynne Oracle ACE
    Currently Being Moderated
    user11181920 wrote:

    But as you can see from previous posts DB/DBA ratio growths quite fast.
    That ratio means nothing. In the past, large databases required mainframes. Today - well, I have a colleague running several databases (that I look after as a favour) on several desktop PCs. Apex used as the front-end, with a user base of a 100+ users.

    With today's technology, anyone with a little bit of tech savvy can deploy a database. And many do. It is easy. It is simple. I recall that creating a database on a mainframe back in the 80's easily took a week - it was a slow process (down to coding each and every column's 3 byte unique identifier). Today - point and click and some minutes later you have a database.

    So obviously, there has been a tremendous growth in databases. Most anyone can go point-and-click and create a database (and many do just that in business).

    But yesterday's 10 database cannot be compared with today's 100 databases.. how do you compare a mainframe database of several million US$ with a bunch of Open Source mySQL databases that have been created mostly on the fly to address some or other issue that involves storing a couple of rows? And then make the jump to say we had a single DBA back then and a single DBA now, and then try to draw conclusions from that?

    I fail to see the logic in that.
  • 37. Re: Are the oracle DBa's seeing an end to their career
    Mark Malakanov (user11181920) Expert
    Currently Being Moderated
    >
    But yesterday's 10 database cannot be compared with today's 100 databases.. how do you compare a mainframe database of several million US$ with a bunch of Open Source mySQL databases that have been created mostly on the fly to address some or other issue that involves storing a couple of rows?
    >
    If you are questioning a methodology of the comparison this question should be addressed to Forrester Research.

    >
    And then make the jump to say we had a single DBA back then and a single DBA now, and then try to draw conclusions from that?
    I fail to see the logic in that.
    >
    I fail to see what do you fail to see.
    Yourself just admitted
    Most anyone can go point-and-click and create a database (and many do just that in business).
    Excellent! It is exactly inline to what I said.
    And I be dare to just continue your thought - Most anyone can go point-and-click and manage a database.
    Why? Because they are easy to be created and managed now. Many features are automated, other so ergonomically advanced, foolproof and user-friendly so one does not need to be super-duper DBA to manage such DBs.

    Then why Business would need so many DBAs?
  • 38. Re: Are the oracle DBa's seeing an end to their career
    jgarry Guru
    Currently Being Moderated
    user11181920 wrote:
    >
    But yesterday's 10 database cannot be compared with today's 100 databases.. how do you compare a mainframe database of several million US$ with a bunch of Open Source mySQL databases that have been created mostly on the fly to address some or other issue that involves storing a couple of rows?
    >
    If you are questioning a methodology of the comparison this question should be addressed to Forrester Research.
    Why not to anyone pointing to Forrester Research?

    >
    >
    And then make the jump to say we had a single DBA back then and a single DBA now, and then try to draw conclusions from that?
    I fail to see the logic in that.
    >
    I fail to see what do you fail to see.
    Yourself just admitted
    Most anyone can go point-and-click and create a database (and many do just that in business).
    Excellent! It is exactly inline to what I said.
    And I be dare to just continue your thought - Most anyone can go point-and-click and manage a database.
    Why? Because they are easy to be created and managed now. Many features are automated, other so ergonomically advanced, foolproof and user-friendly so one does not need to be super-duper DBA to manage such DBs.

    Then why Business would need so many DBAs?
    True story: Overheard, pretty HR lady in the next cubical on the phone to tech support for some PC HR app: "Restore the database?"

    The problem with most people managing a database is they simply have no clue how to manage a database. You can't really automate around that. You have to account for the bigger idiot that is created by supposedly idiot-proof systems.

    That HR database was toast. And not the warm, tasty kind. I dearly wanted to say, "Maybe you should have a dba managing all these SS databases that are popping up," but I knew that would be a political disaster. Later, after I became an employee, I did say that, but of course, SS databases don't need management, they're much better than Oracle, you know. And the poor network admins on the help desk get more piled on them, because the databases actually do need administration. And Oracle is still there, because it scales better, even though the apps run on SS.

    Perhaps a more reasonable argument would be DBA jobs being replaced by Data Management jobs, but that's a specialty too. One that is also improperly managed, leaving developers to deal with reconciling disparate data definitions between and within systems.
  • 39. Re: Are the oracle DBa's seeing an end to their career
    EdStevens Guru
    Currently Being Moderated
    user11181920 wrote:
    >
    But yesterday's 10 database cannot be compared with today's 100 databases.. how do you compare a mainframe database of several million US$ with a bunch of Open Source mySQL databases that have been created mostly on the fly to address some or other issue that involves storing a couple of rows?
    >
    If you are questioning a methodology of the comparison this question should be addressed to Forrester Research.

    >
    And then make the jump to say we had a single DBA back then and a single DBA now, and then try to draw conclusions from that?
    I fail to see the logic in that.
    >
    I fail to see what do you fail to see.
    Yourself just admitted
    Most anyone can go point-and-click and create a database (and many do just that in business).
    Excellent! It is exactly inline to what I said.
    And I be dare to just continue your thought - Most anyone can go point-and-click and manage a database.
    Why? Because they are easy to be created and managed now. Many features are automated, other so ergonomically advanced, foolproof and user-friendly so one does not need to be super-duper DBA to manage such DBs.

    Then why Business would need so many DBAs?
    Because a fool with a tool is still a fool.

    If your thesis were correct, the traffic on this very board would be reduced by 95%.
  • 40. Re: Are the oracle DBa's seeing an end to their career
    Osama_Mustafa Oracle ACE
    Currently Being Moderated
    What i Want To say

    In Any company database is not getting smaller but its getting bigger which is need more monitoring and administration that's not all , but maybe your company will need other solution to make this database is available all the time , Now we are talking about DBA Tasks , Companies usually don't throw their applications and say we will not use them they trying upgrade this application and databases , which indicate that you need to monitor more applications more databases .


    I See lot of customers that hire DBA but they also take Support and consultation from other company . if the futures for developers then DBA will be with them also who will tune bad sql , improve performance , maintenance database .

    in my opinion We are talking about hard future contain long of working hours .
  • 41. Re: Are the oracle DBa's seeing an end to their career
    Mark Malakanov (user11181920) Expert
    Currently Being Moderated
    Because a fool with a tool is still a fool.
    I agree with it. That is why it is very important to keep them separated - a tool from a fool.
    There is several ways to do it. (all can be implemented to some degree and used simultaneously)
    1) to hire a Smart that can use tool.
    2) to convert Fool to Smart - training.
    3) to make a Tool to be easy to use and fool-proof, so any Fool can use it.
    4) to automate - to materialize a smartness of Smart into an automated entity that uses Tool like Smart, and clone its instance together with every Tool.

    There is a price tag for each solution - costs and other implications.
    for 1) - training, salary and other employee related costs.
    for 2) - training, and additional duties for Fool can impact his main duties.
    for 3) - Vendors development costs will be reflected on a Tool price.
    for 4) - Vendors development costs will be reflected on a Tool price.

    As I already wrote that RDBMS vendors gradually but steadily introduce more and more features that can be placed into 3 and 4 categories.
    Which effectively means that some cost burden can be offloaded from 1 and 2 categories. Which in case 1 means less and less DBA (per data unit) is required.
  • 42. Re: Are the oracle DBa's seeing an end to their career
    BillyVerreynne Oracle ACE
    Currently Being Moderated
    user11181920 wrote:
    Most anyone can go point-and-click and create a database (and many do just that in business).
    Excellent! It is exactly inline to what I said.
    And I be dare to just continue your thought - Most anyone can go point-and-click and manage a database.
    And that is where you are wrong. A year or two ago I spend a month with Oracle Support attempting to recover a production database that was build (and "managed") by network specialists (experts in their field) that have little to no DBA (and not much technical database) experience. After the crash, I rebuild a new database on a spare server - but some critical data was lost. And thanks to some undocumented parameters and data block editing and patching, Oracle Support managed to assist in getting the broken database opened for exporting data.

    Most anyone may THINK they are able to manage a database via a point-and-click admin interface. But does this equal an experienced DBA? Of course not.
    Why? Because they are easy to be created and managed now. Many features are automated, other so ergonomically advanced, foolproof and user-friendly so one does not need to be super-duper DBA to manage such DBs.
    Foolproof? User-friendly? It is dangerous allowing non-technical and non-experienced people to take on the roll of computer/database/network administrators. Especially if what is being administered falls into the category of business production.

    Real story. Back in the 80's I worked on-site at a customer doing GIS stuff. They purchased a mainframe. As they had not a single IT staff position, they trained one of their technical surveying employees as mainframe admin. Bright guy.. but with no real IT experience and no real understanding of operating systems. One day he started the initial process loader (IPL) with the F option as he heard of a "fast start". A standard IPL load and boot took around 30+ minutes, depending on subsystems configurations. The F option was a factory start - this is what we do in the factory as the mainframe rolls off the production line. It does a full initialisation (which includes resetting/reformatting all attached storage). Took us a couple of days (booting from magnetic tape and restoring) to get the mainframe back to a usable state.

    I have seen situations like this (though less spectacular than resetting a mainframe, but as costly) more times through the years than what I care for.

    Administration may seem simply via a point-and-click admin interface. But the fact is that technology is getting MORE complex. Thinking that people without technical knowledge and experience of the technologies being used, are able to effectively manage these technologies?

    It is a laughable concept...
  • 43. Re: Are the oracle DBa's seeing an end to their career
    jgarry Guru
    Currently Being Moderated
    Training and documentation are two major weak spots in what you propose. It is difficult to write, coordinate with the actual usage, and propagate the docs, I would think that is obvious.

    Training is worse. You have to make arbitrary assumptions about previous knowledge and experience, and that is multiplied by the various imperfect ways people learn, retain and react to information. Given a general user population, you wind up with people in over their heads when they apply the training, even if they are well qualified on the functional application level. What are you going to do, fire them? (I've seen the answer be yes. This week.)

    Once problems reach a certain level of complexity, fixes should not be automated. Predicting that level is impossible, because the variables involved are not fully known. Even the things you can know are sometimes [url http://www.utsandiego.com/news/2012/sep/09/tp-blackout-one-year-later/?print&page=all]ignored.

    Well, back to figuring out why the statistics I locked down months ago to keep some plans stable let the optimizer suddenly decide to switch plans anyways.
  • 44. Re: Are the oracle DBa's seeing an end to their career
    Mark Malakanov (user11181920) Expert
    Currently Being Moderated
    I can tell cases and stories from my experience too. Both for advocating absolute necessity of hiring of experienced DBA and totally opposite ones.
    For example, just yesterday one of my friends told me. He works in a development shop company, that also provides their AppAsAService to customers. They have multiple production instances of MS SQL and numeruos test and dev ones. And they do not have DBA. At all.
    Sysadmins use enterprise backup solution for backup, developers write SQLs...
    They work in this mode for years and nothing bad ever happen...

    So what?

    My point was not in this. I do not disagree that for many environments DBA is a necessity.
    I am just saying that with evolution of DB Management tools and automation need for DBAs decreases. The ratio DBA/DataUnit decreases. So demand on DBA may also decrease on job market.

    Of course I would love to be not right. But my humble analysis of how Oracle is progressing with automation and ergonomic improvements shows that they (Oracle) gradually pushing out an experienced DBA by introducing convenient and fool-proof tools, and they completely replacing DBA in some DB management and tuning areas by automation.

    To see it, just answer on a simple question:
    How many 50GB databases with 200 users (identical OS, apps and workload) you could administer for :
    Oracle 5
    Oracle 6
    Oracle 7
    Oracle 8
    Oracle 8i
    Oracle 9i
    Oracle 10g
    Oracle 11g
    Oracle 12c

    You can use all features you need that are available for the version - management packs, Grid Control etc...

Legend

  • Correct Answers - 10 points
  • Helpful Answers - 5 points