If you start your procedure name with "sp_", put it in the master db, and mark it as a system object, it can execute in the context of the current...

If you just want to run the proc for a single db at a time, you don't need dynamic SQL at all, so I suggest avoiding it. The code's...

I, too, don't have a problem with a cursor per se here; it could be done w/o one, but it's not a big deal for such a limited number of...

Btw, it's always frustrated me that SQL Server can't seem to read only the non-leaf pages of a clustered index. If SQL needs only key column(s), why does it...

Don't automatically assume the PK makes the best, or even a good, clustered index, in particular if the PK is an identity column.

You need to review the existing index usage,...

GSquared (1/7/2013)

---

paul.j.kemna (1/7/2013)

---

Wow, that makes a lot of sense. Such a simple concept that i may have overlooked in some places.

But in this case, we are assuming that I...

Your...

If all the columns are from a single table, you definitely should created a computed column that has the computation: that way, the definition is in only one place. ...

🙂

GilaMonster (1/7/2013)

What is the basis for you claim that if only 1% of rows in the table match that SQL will almost certainly do a table scan? If the...

GSquared (1/7/2013)

---

ScottPletcher (1/7/2013)

---

GSquared (1/7/2013)

---

Assuming "correct results" means dates between the two variable values, including the end-date value, then yes.

As mentioned, using <= @EndDate can give odd results.

For example:

DECLARE @T TABLE

...

GSquared (1/7/2013)

---

Assuming "correct results" means dates between the two variable values, including the end-date value, then yes.

As mentioned, using <= @EndDate can give odd results.

For example:

DECLARE @T TABLE

...

GilaMonster (1/7/2013)

---

The most selective one (the one that, based...

I suggest a hybrid method.

Do something like the first method except that you begin new transactions only at certain record counts, for example, every 2000 values processed:

record #1; record #2001;...

With a clustered index on ( A, B, C ) and a

WHERE A = 6 AND C = 7

I would expect SQL to a clustered index seek.

That is, it will...

paul.j.kemna (1/7/2013)

---

The actual syntax would be :

DECLARE @StartDate as date

SET @StartDate = some calculated date

DECLARE @EndDate as date

SET @EndDate = another calculated date

CONVERT([varchar](10), MyDatetimeColumn, 101) AS MyDate

...

WHERE MyDate >=...