I suggest creating the proc in the master db and marking it as a "system" proc. It will then work within the context of the current db, whatever that...

Maybe?:

SELECT CONVERT(varchar(8), DATEADD(SECOND, DATEDIFF(SECOND, ISNULL(QueueDate, StartDate), EndDate), 0), 8)

If you decide to run it locally, clustering the tables to match the join logic will help:

IF OBJECT_ID('tempdb.dbop.#RFI') IS NOT NULL

DROP TABLE #RFI

CREATE TABLE #RFI

(

COMMIT_NO CHAR(4),

BODY_NO...

Does the clustering key start with Monthkey? And the partitioning key?

kenambrose (3/10/2016)

---

why do the same flawed arguments keep popping up?

[[If you deal with a report and want all columns via left joins, then there will be nulls. ]]

We don't know...

For efficiency, the WHERE clause should include just the specific date ranges you need, rather than the entire year. Particularly if m1_KF.dbo.ShipmentLines is clustered first on smlCreatedDate (which it...

Try using a CROSS APPLY to assign an alias to the random number:

SELECT

RowSequence AS CaseID

,DateSequenceas [CaseDate]

,random_number

,CASE random_number

WHEN1THEN'Black'

WHEN2THEN'White'

WHEN3THEN'Red'

WHEN4THEN'Blue'

WHEN5THEN'Silver'

WHEN6THEN'Grey'

WHEN7THEN'Green'

WHEN8THEN'Yellow'

ENDas [VehicleColour]

FROM #RowGen

CROSS...

Zidar (3/10/2016)

Somewhere at the begginig it was said that rows in a table (tuples in a realtion) represent only TRUE propositions (declarations). If 'a piece of data is unknown' it...

Look at the missing index stats, index usage stats and index operational stats. SQL itself will help you determine the best clustered index, which should be your goal.

You'll get the most performance benefit by determining and implementing the best clustered index for that table, in this case very likely the date.

But you might also gain some performance...

What is the data type for:

sys_updated_on

?

If it's smalldatetime/datetime/date/datetime2, then you should compare it to format YYYYMMDD, because that format is 100% universal across any/all SQL instances. The convert code...

What is really critical is how the information table is clustered. Insure the table has the best clustering index to reduce rows to be read as much as possible...

In the real world, there can be unknown values at a given point in time in a data store. For example, birth date. If you need to create...

Zidar (3/2/2016)

---

I am not done 🙂

In T-SQL it takes two CHECK constraints to enforce the rule:

C1: CHECK (NOT [Job] = 'DBA" OR [CertificateNum] IS NOT NULL)

C2: CHECK ([CertificateNum] ...

kenambrose (3/2/2016)

---

[represent having no value]

it was in earlier post. put nullable column into it's own table, enforce 1 to 1 relationship to parent table.

no data for a row in...