SQL Server Central

Over the past couple of years, I’ve been using the bejeezus out of AI. I don’t really feel like an expert by any means – it’s so hard to keep up with the advancements – but I’ve been enjoying reading Grant Fritchey’s series on building stuff with AI, and I thought you might enjoy some of my lessons learned as well.

Introduction A finance manager suddenly discovers that an employee, who moved to a branch office six months ago, can still browse confidential sales reports in an SSAS cube. Nobody remembers why. Nobody knows which role grants the permission. And checking hundreds of SSAS roles manually could take hours. This is a far more common problem than many DBAs realize. OLAP cubes are often considered "internal" systems, which means they rarely receive the same security attention as relational databases.

Group Managed Service Accounts (gMSAs) solve the classic service-account password problem - just look at the glowing comments on this Reddit thread. It's a wonderful solution to a genuinely difficult and essential security problem. When an experienced DBA first learns about gMSA, they search the internet and invariably land on a handful of excellent how-to posts that are clear, step-by-step and easy to follow. That's the old approach.

Indexes are supposed to make SQL Server faster – so why do so many databases end up slower, bloated, and harder to maintain when they have more of them? It usually comes down to misapplied indexes rather than missing ones. There may be too many that are too wide, tuned with settings that don’t fit the workload, or built on assumptions that stopped being true years ago.

In a Power BI project I built for a company, I started with the classic map showing revenue per shop. To make it more interesting, I decided to add the locations of competitors as dots on the map. Not a bad idea, but harder than I expected. Getting the competitors’ database with location wasn’t straightforward, and I eventually had to rely on OpenStreetMap to obtain that information.

In the modern data landscape, most of the effort doesn’t go into building models or dashboards. It goes into getting the data into a usable state. Generally, a large portion of time is spent extracting, cleaning, and reshaping data before any analysis can begin. While Python and SQL together form a powerful stack, the boundary between them is often where things start to break down.

Introduction When we began learning SQL, almost everyone of us started by using the wildcard operator: "SELECT * ..." It is incredibly convenient. It tells the database engine to fetch every single column from the particular table, and we don't have to type out every column name individually.

Disclaimer: The following is built with Claude Code. Just need to justify my $100/mo subscription cost. This was meant for a private demo so I can learn more about parameter sniffing and memory grant, but I decided to open it to public. And you may say “Marlon, this isn’t you, this is too good.” Again, I will say it again, originally I was to keep this a private learning doc for me to understand Parameter sniffing. I asked Claude Code to build me a test for param sniffing and memory grant feedback.

T-SQL Tuesday is a monthly blog party hosted by a different community member each month. This month, Jeff Taylor (blog) asks us for our take on temp tables: So, do you use temp tables, or do you pull them out wherever you find them? Make your case. I’ll be reading every post. His whole post is worth your time, so please pop over there once you’re done here and check it out.

I developed these rules a while back, before COVID. I had a period of self-reflection when I realized I was spending time wastefully arguing with folks on the various platforms with no real expectation of changing my mind or theirs. I review these rules periodically to reduce the likelihood of following into those same patterns. I’m not always successful, but I’m far better than I was. I also re-post these rules from time-to-time in part to share them but also in part for accountability.