This blog started in 2017 as somewhere to write down what I was learning as a SQL Server DBA. Over time it collected two quite different things: SQL Server material, and a lot of Windows, PowerShell and WSL notes from the same job. Useful, but not a library anyone could navigate.
So the SQL Server side moved. It now lives at sqldba.blog, rebuilt properly rather than copied across. This post is a short account of what is over there and how it is organised, for anyone who has been reading here for a while.
What Is Over There
Four things, each built as a library rather than a pile of posts.
DBA Scripts
150+ production scripts, each with its own page explaining what it does, when you would run it, and what the output means. They are grouped by the job they belong to: backups and recovery, performance, security, storage, monitoring, HA/DR, migration, and general server configuration.
Every script carries a safety classification in its header, because a script that reads DMVs and a script that changes configuration should never look the same at a glance. They also live in a public GitHub repository if you would rather clone the lot.
The Error Library
Paste the error message, find the write-up. Each one covers what the error actually means, what causes it, how to fix it, and the checks to run afterwards. It is the part I use most myself, because the useful version of an error page is the one written by someone who has hit it at 3am rather than one that restates the message back at you.
The Wait Types Library
A page for practically every wait type, more than 230 of them, plus the guides that explain what to do when one dominates. Wait stats are the single best starting point for “why is this server slow”, and most of the material online stops at defining the acronym.
The SSMS Complete Guide
Installing, configuring and fixing SQL Server Management Studio, kept current with the SSMS 22 line. There is also a builds and lifecycle reference that lists every build number, its CU, and exactly when each version falls out of support, which is more useful than it sounds the first time an auditor asks.
The Part I Did Not Expect to Build
There is now an MCP server that exposes the same reference data to AI assistants: error numbers, wait types, build and patch levels. The reason it exists is simple enough. Ask a general-purpose model for a SQL Server build number or a support end date and you will often get a confident, wrong answer, because that is exactly the kind of detail models misremember. The server answers from checked data and cites its source, or says it does not know.
That is the honest summary of my position on AI in this job: useful for drafting and explaining, not to be trusted on specifics unless the specifics come from somewhere verifiable.
What Happens to This Blog
It stays where it is. The Windows, PowerShell and WSL material was never going to fit on a SQL Server site, and it is still the part of this blog people actually read, so it is being tidied up rather than retired. Two new guides have just gone up that pull the older posts together:
Older SQL Server posts here now point at their current equivalent on sqldba.blog, where one exists. If you arrived from a search result and landed somewhere that felt out of date, that is why.
Thanks for reading, whichever era of this blog brought you here.
Leave a Reply