Is the following correct? I think you want a property value other than "IsPrimaryKey" based on the description and the output (in the article).

Alternatively, you can use the system views...

And fairly straightforward to accomplish. All I was after is that it is a common failing that usually manifests itself at the worst possible time. I was also reminded of...

What I haven't seen in this conversation is any control for uniqueness. Certainly one can argue that a monotonically increasing record (not row) counter is unique. It does, however, allow...

And now I do...

I don't see an error message in the text, just whitespace.

Based on your conclusions about unit testing I'll ask if you've worked with tSQLt as a formal test framework? Without getting into the details of your UDF or stored procedure...

Granted, restoring a backup to a live database after a deploy with issues is risky at best and virtually impossible in most cases. Typically we just move forward...

Same here. Our developers have a template script to use as a starting point so some of the details are provided for them. We also have a scratch database on...

The ability for developers to delete data simply is a feature is mismanaged security. It has nothing to do with Dev-Ops.

To @jeff Moden's point, there are absolutely times...

• This reply was modified 5 years, 1 month ago by  Bryant McClellan.

I guess I've yet to encounter a tuning problem that required NOLOCK to handle.

Agreed. That is also why having some demonstration of the potential consequences is so valuable. For some applications it is less risky.

The biggest problem is having it peppered throughout a...

I've worked with SQL Server since v4.2.1. I have never put any code containing NOLOCK in production. At my present employer we actively remove WITH (NOLOCK) from code as it...

I am always surprised that the default isolation level for SSIS package objects is Serializable. Granted, it could be useful or even necessary in some cases. Isolation Level settings other...

Whether you intended to or not you again illustrated the fallacy in "...and the execution plan showed all index seeks and no scans...". Seek good/scan bad is not a valid...

Thanks for the straight-forward example, Steve. Things like this show that tSQLt testing isn't difficult and helps you avoid deploying bugs in the first place.

We adopted tSQLt for database unit...