Kerberos Authentication and SQL Server (Part 1)

It’s been a while since my last post (i.e. almost 1 & 1/2 years, ouch!), but I hope to get back into a more regular pace with my updates going forward.  My next few posts will be a short series related to Kerberos Authentication, particularly in relation to the SQL Server product family.  Kerberos is a critical component needed for some advanced security features and configurations and Microsoft recommends it be enabled in every SQL Server deployment, pending outstanding circumstances.  Some people will find it difficult to setup, but those challenges tend to relate to misunderstanding what’s going on and what permissions are required to get everything working.  In this post, I’ll discuss what Kerberos is and what is needed within your Windows environment for SQL Server to utilize this authentication protocol.

Soup to Nuts: Oracle 11g Client on Windows (Part 2)

The Nuts: (Un)Deinstalling the Oracle Client

In part 1 of this two-part post, I described how to install the Oracle 11g Client and how to fully configure it to work for what I believe is a majority of use-cases on Windows.  In this follow-up post, I will discuss how to uninstall (Oracle likes to say deinstall) the client both automatically and manually so you can start over,  upgrade to the latest and greatest client, or walkaway from Oracle being on your machine once and for all.

Soup to Nuts: Oracle 11g Client on Windows (Part 1)

The Soup: Installing The Oracle 11g Client

Chances are higher than 90% that if you work in a corporate environment you are logging onto a Windows OS every morning.  If you happen to dabble in the Oracle domain, chances are also highly likely that you have (or will need) the Oracle client installed on your machine.  The Oracle client installer, like many things Oracle-branded, is based on Java, which can best be described as “tolerated” by Windows (though some corporations are becoming less so).  Frankly, I dislike Java-based installers as they tend not to tie in well with the Windows operating system, and the Oracle client installer is no exception.  Regardless of the underlying technology or my opinion, my need to have the client installed is still very real as I work with both SQL Server and Oracle database solutions.  In my experience, it’s quite easy to end up with either a bad, broken, or partially working Oracle 11g Client installation without even knowing how or why.  This 2-part post will help avoid these situations before they begin or help rectify them if you already find yourself in a bad spot.  Part 1 will outline how and what to install using the installer, and Part 2 will outline how to uninstall the client, both automatically (if possible) and manually (if necessary).

Making NetApp SnapManager “Play Nice”

Implementing NetApp SnapManager for SQL Server

I’m a firm believer in using any software at your disposal to it’s fullest.  There are times though that I curse my own name for adopting this mantra, because let’s face it, some software just refuses to “Play Nice”.  I don’t want to sound too critical here, but NetApp’s SnapManager for SQL Server product is one of those, how shall I say, onerous applications.  If you find yourself running into issues with the tool or are exploring the possibilities of adopting it in your environment, read on as I go over some of its idiosyncrasies.

Truncating Tables Without Elevated Rights

A common request from users is to receive “Truncate Table Rights” within a database.  Because a truncate statement is a DDL command, this isn’t nearly as easy to fulfill as most people would expect.  Faced with this challenge, a DBA might even find themselves choosing between elevating the user’s rights (e.g. including them into the db_ddladmin or db_owner role) or being a bad guy and flat-out denying the request.  Neither option is a win for both parties, so thankfully (in SQL 2005 and later) there is another solution.

Eliminating Static Output Logs

Using Tokens to Create Dynamically Named Job Step Output Logs

There are situations that make me think I need to find and enroll in a “Making Better Google Searches” class.  After some brief digging into Ola Halengren’s Maintenance Solution I inadvertently came across the way of creating log file names dynamically for job step output.  While this may not sound all that wonderful, I’ve asked a lot of people how to do this over the years (including MS support during various incident resolutions) and no one has been able to tell me what it took 2 minutes of digging into Ola’s solution to find.  I think the problem isn’t the folks I asked, but how I asked the question.  Anyway, read on and I will show you the way to create dynamic output log names within a Job Step’s Advanced Tab.

Bucking a “Best Practice”

Why avoiding the “Change port 1433 Best Practice” might be the “Better Practice” to follow

A lot of quality resources in the DBA community (e.g. MS, Jonathan Kehayias, Green SQL, etc.) state that changing a default instance’s port to something other than 1433 is a security “Best Practice.”  While I may agree that you want to avoid using port 1433, I contend there are better ways to go about it.

Hooray! You finished installing SQL Server, now what?

John’s post-install configuration recommendations

Let me start by saying this is the procedure I perform when configuring a new SQL Server instance, immediately after installation.  “Best Practice” is a term a lot of people love to throw around, but I will not definitively say performing the following is a “Best Practice” as your environment and situation could be wildly different from mine.  I find though, that these steps work best for me in most situations.  With that said a lot of the steps outlined below do have “Best Practice” associated with them from one source or another, but they could range anywhere from highly useful to absolutely worthless for your environment so keep that in mind as you read on.