Just to refer to the orginal problem:

The root problem of the big log file could be fixed / prevented, if you would partition your tables. Ola's script would in this...

when you are working with dynamic SQL and particularly, when it uses varchar-parameters and adds it to the dynamic code, you have always to use QUOTENAME(), otherwise someone could call...

@hmbacon: you could either do 6 Updates each joining the base table or - if the select of the base table is slow / difficult - write the ids, that...

The DELETE WHERE id IN (SELECT TOP x) is still a very bad solution. You wrote that it makes a table scan in your first example, so I assume that...

@dave_rogers: I disagree with your "rule". It is often much easier to query a log table (even if it will be never joined to whatever) than to read / search...

• This reply was modified 4 years, 10 months ago by Thomas Franz.
• This reply was modified 4 years, 10 months ago by Thomas Franz.

ok, theoretical you could for the cursor "problem" declare an additional variable, increase it inside the loop and write a WHILE @i < 999999999 instead of 1=1

But what, if someone...

Did you just click onto the link and tried it out? It is only a simple database / table on a webserver and a web frontend where you could enter...

Nice, but it would be nicer, if I could also search for

• functions as CONCAT_WS(), STRING_AGG(), APPROX_COUNT_DISTINCT(), TRIM(), TRANSLATE()
• OPTIMIZE_FOR_SEQUENTIAL_KEY (Index option for faster inserting MANY rows with increasing key)

...

• This reply was modified 5 years ago by Thomas Franz.

@jeff - but both it is linked together 🙂

@Jeffrey: of course I'd cluster your table by PatientID, DocID too. But when you are not a very new hospital 99.x% of...

• This reply was modified 5 years, 2 months ago by Thomas Franz.

Stupid question, but why are some people use a low fill factor as 70 or 80%, I'm usually fine with something as 95-98 (or even 100 if it is on...

Nice article, but I think it could need a second follow up article

• In your example you created the certificate by yourself (you = SA).
• The reason for Always Encrypted...

I manage a big DWH database (about 12 TB). Of course (for this size) the database is partitioned. The biggest tables have their own set of annual filegroups (e.g. FG_ORDERS_2019,...

we are saying, that #1 and #3 are correct, not #3 and #4.

LEAD with descending order is the same as LAG with ascending order (except that the returned rows are...

@scott: The question, if they should cluster for it or not depends, how else this table is used. Yes, 2,400 deletes per day may be a lot, but when they...

regarding the clustered index recommentation:

you do not necessary have to cluster your table by MyTime. When MyTime is ascending too (e.g. the creation timestamp) you may e.g. simply selecting the...

• This reply was modified 5 years, 6 months ago by Thomas Franz.