Cody Konior

@codykonior.bsky.social

Expert at scaling SQL Server and PowerShell. Happily married with 4 cat children 😻

@owen.nz Just wanted to drop a line to agree and say yes Elastic Jobs are super chatty for auditing. I think mine is way more than 24qps as well. I also had to exclude them from auditing in a similar way to you. Lol. Microslop strikes again.

If a logging process goes crazy and starts spinning up sessions dumping into a table, you could disable the user temporarily waiting for the other side to die, or add an instead of insert trigger for a little while that just does a return 😛

TIL if you have an empty folder you want to keep in git (so you don’t need to test and recreate it in every script that touches it) the convention is to use an empty .gitkeep file in there.

I had resorted to doing all users in a post-script in SSDT but it turns out you can add a publish flag that lets you create the users but not deploy the fake logins you define for them. That stuff is important when you want to drop all users not defined in your project.

If you open SQL Prompt and the dropdown menu only has 3 items on it (about, check version, something else), it means it has lost its license information. Don’t bother messing with the folders or reinstalling. Click the about and it’ll log you back in to authenticate 🤦‍♂️ Shame it doesn’t tell you!

I think I found a weird optimiser error in Azure SQL Database. You can do a SUBSTRING CHARINDEX CHARINDEX with valid values and it’ll barf with a length error even though it’s 100% valid. If you add a WHERE clause with that CHARINDEX > 0 to act as a filter, it works even though it’s not NULL/0/-.

Be careful trusting Set-AzSqlServer to enable or disable identity types of SystemAssigned/None. Shit doesn’t work. Only the portal works. What a fucking piece of shit. No error or anything either.

I’m putting a 100% tariff on SQL Elastic Jobs. You know how you can tie a job to databases in an elastic pool? Well if one of those steps fails but is between retries, you can move the database to another pool and that step will still retry and complete. But tomorrow it won’t because wrong pool.

It bothers me when I look at an online store and get an automated email, “Hey, we noticed you checking out the store, did you want XYZ item?” Yeah? Did you notice me looking at your products and they’re all out of stock in my size, which is Medium? The lack of data analysis everywhere sucks.