Regarding SQL Versions and restore to older versions: the "assuming no new keywords are used" would be a bad sign for a new version.

I don't care about all the fancy...

I think triggers can be agood. On the database of my previous employer we had tons of triggers to ensure business logic. Of course you could argue, that it should...

why should I do this or how would this work? If I take a single file from a filegroup offline, the data in the file will no longer accessible. And...

I'd assume that the NoSQL data model fits better for very specific approaches, where you can encapsulate a bigger part of data and assign it to a specific user etc..

Imaging...

• This reply was modified 4 years, 4 months ago by  Thomas Franz.

I had a coworker once (luckily sitting in other room) who was a ~40 years old single and lived in a guesthouse/motel. I don't know, what he did wrong, but...

oh, I did already - about 20 questions 🙂

I hope, this will be not the next Question of the Day, but sometimes you have the value not in a binary column but in a VARCHAR (e.g. because you...

I used this behavior

• while setting a column to the MIN/MAX() result of a subquery, where I needed this value for further use in a variable too
• setting a column...

• This reply was modified 4 years, 8 months ago by  Thomas Franz.

We can't use it either. We have an generic file loader, that searchs in the directories for new files and calls the fitting package to load the file. The (sub)package...

Counterquestion: why should it be forbitten? It does not write anything to the log, changes no data ...

I don't know how it is realized internal, but theoretical an USE [db]...

For the sake of completeness, CHOOSE() ignores errors too, until you access this value:

-- works well
DECLARE @i INT = 4;
SELECT CHOOSE(@i, 10,20,30,3.14,50,'Test', 1/0)
GO
-- returns division by zero...

When disk I/O is no longer a problem, it comes to CPU and network I/O... To reduce the CPU load (and disk I/O) you could e.g. tune your queries /...

Even after several minutes of thinking, I can't really anything I know about indexing flag as only nice-to-have. Ok, maybe you could just skip the whole in-memory-table stuff with its...

theoretical you could have added one more file to the filegroup and used

DBCC SHRINKFILE (N'<original file>' , EMPTYFILE)

to move all data into the new files (will be evenly distributed in...

When your 500 GB database would use multiple filegroups (and the bigger tables are ideally partitioned into different FGs too), then you would be able to do a partial restore...