The DBA of the Future

  • Great comments, thank you all for reading and participating!

    Blandry, I was struck in particular by your comment and I'll think on it more. I do wish DBA were more well defined, and I think about Andy L's comment on data professional, at this point in my career I'm more of the latter, but the title tends to be DBA because businesses don't understand anything else (even if DBA could mean most anything). I'll definitely chew on it some more.

  • LSCIV - I don't think IOPS will disappear, just that we'll get a budget of them and/or an SLA, and someone else will handle all the ugly details of disks and raid and such. The tools, as much as they have advanced, still have a ways to go, but with the DMV type info evolving, we should get to the point in the next gen or two where the tools will tell us whether we're using or abusing resources. At that point the business pays us to try to reduce consumption, or pays to increase resources. True today of course, but often a swag, I look forward to it being something we can easily measure which way has the better ROI.

    I guess it's also that my visions matches the flow I already see. I rarely deal with disks any more, and it's taken a bit to get comfortable with that, less control you know! But specialization is the trend and I think it's healthy, just makes sense to have someone that focuses on storage.

  • here is my thoughts:

    1.Performance front:

    We can expect 256 GB RAM and 100 TB better version of SSD, 4 times faster processors and may be consistant 100 MBPS wireless Broad band speed

    Impact: there wont be performance problem in small/medium scale databases as most would be in memory databases. We can expect cheeper in memory solutions in another 10 years.

    2. some DBA tasks:

    We can expect better in build options like

    1. options for auto archival

    2. options for intelligent auto indexing and intelligent defragmentations options

    3. better options to handle too much file growth (like tempdb/log file growth)

    Impact: day to day DBA tasks would be inbuild options in another 5 or 10 year.

    3. High availability:

    We can expect same Clustering/Mirroring/Backups features with some improvements.

    Impact: Even a junior DBA/developer will know on how to recover data with 0 data loss.

    4. We can expect inbulid team based development suite (like vss)

    5. DBAs will be forced to learn .Net, Cloud and SSIS & SSAS for survival.

  • Ramkumar (LivingForSQLServer) (4/17/2011)


    here is my thoughts:

    1.Performance front:

    We can expect 256 GB RAM and 100 TB better version of SSD, 4 times faster processors and may be consistant 100 MBPS wireless Broad band speed

    Impact: there wont be performance problem in small/medium scale databases as most would be in memory databases. We can expect cheeper in memory solutions in another 10 years.

    2. some DBA tasks:

    We can expect better in build options like

    1. options for auto archival

    2. options for intelligent auto indexing and intelligent defragmentations options

    3. better options to handle too much file growth (like tempdb/log file growth)

    Impact: day to day DBA tasks would be inbuild options in another 5 or 10 year.

    3. High availability:

    We can expect same Clustering/Mirroring/Backups features with some improvements.

    Impact: Even a junior DBA/developer will know on how to recover data with 0 data loss.

    4. We can expect inbulid team based development suite (like vss)

    5. DBAs will be forced to learn .Net, Cloud and SSIS & SSAS for survival.

    Hmm...allow me counter:

    1) Small & medium scale DB's will not live on this hardware. They will live inside virtual clusters on this hardware. Performance problems will live on.

    2) Fully functional DBA capability has been vaporware for 15 years. If the product stopped evolving, then maybe. Possibly some limited auto-indexing.

    3) A junior DBA/developer should know how to recover data but that doesn't mean they will know. Technology changes, people do not.

    4) VSS is dead. Long live TFS.

    5) Jack-of-all trades a good DBA does not make. I do agree with knowledge of the Cloud, though.


    James Stover, McDBA

  • Nice to see your comments James Stover.

  • James Stover (4/17/2011)


    5) Jack-of-all trades a good DBA does not make. I do agree with knowledge of the Cloud, though.

    James - I love this response to that statement. Too many times I have interviewed for a DBA position that was actualy a BI development position. Once I was hired as a DBA, put into a Database developer role as well, and then the BI person was let go and I was asked to produce new product filling all three positions. What a nightmare.

    I can't think of an environment where it would be good for the DBA and BI developer to be the same person.

    For Example: I would never think to ask a good DBA (that knows how to create a DB schema backup with TSQL or write a SPROC and table set that automates nightly Index maintenance and logs it) to create a new floating dynamic date range deminsion in a data cube on old and existing data.

    I must be missing something, becuase I do not understand the need for a DBA if your database is in a cloud.

  • Nice to see comments from James and SanDroid.

    I would like to share my comment on James reply.

    "Jack-of-all trades a good DBA does not make. I do agree with knowledge of the Cloud, though"

    yes. I do agree with you. but my thougt here is, nobody can expect what improvement/new feature will available in another 10 year.

    one simple example

    We had to rely on Profiler to understand pain areas and painful transactions. but now we have n number of DMVs and Activity monitor and inbuild performance/activity related reports. Guess what we can expect in next 4,5 releases.

    so, i think, If all current DBA stuff are enhanced in future versions, believe after 10 years a junior level DBA can do all in new version. I am sure that you and me will be forced to learn new stuff (love to learn new stuff is a different story)

    We can't be master of all. but We have to be Proficient in new skills related to SQL for survival.

    else person who have more skills than Us may replace Us.

  • Ramkumar (LivingForSQLServer) (4/18/2011)


    Nice to see comments from James and SanDroid.

    I would like to share my comment on James reply.

    "Jack-of-all trades a good DBA does not make. I do agree with knowledge of the Cloud, though"

    yes. I do agree with you. but my thougt here is, nobody can expect what improvement/new feature will available in another 10 year.

    Very true. I can compare 10 or 15 years ago to today. In SQL 2000 (and SQL 7) a good BI developer and a good DBA had very differant skills as they do today.

    We can't be master of all. but We have to be Proficient in new skills related to SQL for survival.

    else person who have more skills than Us may replace Us.

    If you are not proficient or a master at something should you be doing it as a professional?

    I know there are many thousands that are more skilled than I at many things.

    The only one that could take over my position without training would be the one that trained me on "how it is here".

    There are reasons other than skill that allow persons to be hired and stay in positions they have no skill or expertise.

    I would hope these chagne in Ten years, but I doubt updates to SQL server will have an affect.

    I would suspect that would rely more on social changes.

    As for the BI and DBA skillsets. I know that there are more new skills important to both areas today than Ten years ago. New features, require new skill. Both of these feilds are full of both. I do not see that changing in Ten years. Mores Law does not either.

  • The technology will improve but the appetite to store, retrieve and report on information will always stay a step ahead as the appetite cannot be fulfilled, it is limitless. Also there will still be a requirement for human intervention to prioritise processing and maintain what it is the business want but which the business won't define. Unless we get into the realm of true artificial intelligence with empathy for human desire and processing as reflected in 'business' then this will not be done by software.

    Nigel Moore
    ======================

  • I think DBAs are going to have to learn BI. I recently went to a MS BI conference and learned a lot.

    It was a little wake up call though when Michael Tejedor (Senior Product Manager on the Microsoft Business Intelligence team) showed some of the new features in Denali.

    One thing that stood out was project Cresent. He basically showed that more business users are going to be able to play with data much more easier than before. I am sure he made it seem simpler than it actually was but it looks like thing kind of thing could really take off. I can see CFO's, CEO's, etc drool over this.

    I feel this is going to make the DBA seem less important. BI analysts will do all the manipulation, massaging, and structuring of the data. Less technical people will be able to play with data easier.

    Time to brush up on my cubes!

  • Yin - Great example...

    This is essentially the first step down the primrose path to n00b-Tube City (what I was eluding to in my last post).

    We are seeing this happen at our company (think KableTown :-P) right now. Essentially all data storage/aggregation is being consolidated into a giant Analysis Services Cube, which we were already starting to do internally with the DBA/Dev teams collaborating to create our own Data Warehouse(s). However, the company has been consistently moving away from internal DBA/App Dev and moving towards BI Analysts/Project Managers and other less technical people leveraging GUI interfaces to Third Party Analysis Services Cubes, essentially automating the technical aspects of the job through the GUI, and outsourcing large parts of the IT/data center staff to the Third Party provider of the SSAS solution.

    I would take it one step further (especially at some of the larger companies)... DBA's... Don't just prepare to learn BI skills... prepare to become a BI expert and/or go find a job working for the Third Party Vendor(s) who will now be providing the services and support that you once did 😉

Viewing 11 posts - 16 through 25 (of 25 total)

You must be logged in to reply to this topic. Login to reply