THE SQL Server Blog Spot on the Web

Welcome to SQLblog.com - The SQL Server blog spot on the web Sign in | |
in Search

Andy Leonard

Andy Leonard is an author and engineer who enjoys building and automating data integration solutions. Andy is co-host of the Data Driven podcast. Andy is no longer updating this blog. His current blog is AndyLeonard.blog.

You Need a DBA

This blog has moved! You can find this content at the following new location:

http://andyleonard.blog/2014/06/24/you-need-a-dba/

Published Tuesday, June 24, 2014 8:00 AM by andyleonard

Comment Notification

If you would like to receive an email when updates are made to this post, please register here

Subscribe to this post's comments using RSS

Comments

 

Dave said:

Andy, I was one of the primary commenters on Grant's original article.  I think I didn't make my point clearly.  I absolutely agree with you that DBAs are needed.  In fact, you qualified it by saying "to properly administer a production relational database instance".  I would take out "relational".  I've worked on NoSQL solutions that were touted as not needing a DBA and I found that a DBA was even more of a requirement if you wanted DR and performance.  The DBA skill just becomes HIGHLY specialized with NoSQL.  

At the big companies where I consult (that should know better) there is a push to move relational out and NoSQL in.  When I ask why it is invariably, "we don't need DBAs then".  When I mention specific use cases where a DBA is still needed even for NoSQL then I hear (paraphrasing of course), "well, our relational guys hold us up too much by wanting every change to be reviewed in design sessions.  This means we can't put together a working POC of this new thing we are building without a DBA making us go back to the drawing board to get everything in proper 3NF.  Our NoSQL tool is schema-less so we can evolve our designs as the requirements evolve."  There you go.  Of course you know this is a lie.  There are plenty of NoSQL use cases where modifying a schema requires an entire export-and-reload.  

I don't believe any of the above comments about relational folks, but it appears to be the perception, at least sometimes.  If it wasn't, then why do NoSQL vendors state on the first few pages of their marketing tools that a)you don't need a DBA b)you don't need a rigid, structured schema?  They know how to press the right buttons with the business folks.  

I believe we, as relational practitioners, can change some of these perceptions while still maintaining our relational fidelity.  One way is by being less rigid about changing our schemas.  At some places it is common for a schema change to require 2 weeks to get approval because of those "nasty" DBAs and their process.  And the perception is this adds no business value (not my belief, but a repeated perception).  Using tooling, like Grant mentioned, we could get those changes turned around in a day or two.  

Anyway, I agree with you that DBAs are needed and important.  But sometimes the perception is that (development) DBAs get in the way.  I hope I'm wrong and the places where I consult that are pushing to move to NoSQL are the edge cases.  I hope the perceptions that I hear about regarding relational folks (not just DBAs) is wrong.  

June 24, 2014 8:56 AM
 

Robert said:

Hi,

I agree in full, but your final statement is contradictable. You say "You need a DBA", that is singular it should have been "You need at least TWO DBAs"

June 26, 2014 10:07 AM
 

Scott said:

There is a saying about the importance of having a back-up in case one fails:

two = one, and one = none

This applies in so many ways.  It could mean having an extra key to your house or car.  How about having a second somke detector somewhere in your house?  It also applies to DBAs.

July 2, 2014 2:45 PM

Leave a Comment

(required) 
(required) 
Submit

This Blog

Syndication

News

My Latest Book:

Community Awards



Friend of Red Gate

Contact Me

Archives

Privacy Statement