Here is a sample CROSS APPLY for your situation.

SELECT Part#, CustomerID, ca1.Wk_Num, ca1.Qty

FROM (

SELECT 'A' AS Part#, 1 AS CustomerID, 10 AS Wk1, 20 AS Wk2,...

Yeah, sorry, in too big a hurry, left something out of the code:

SELECT

/* @IdCompany = */ e.IdTCompany,

/* @IdClient = */...

Did you look at the code I wrote??

SELECT

/* @IdCompany = */ e.IdTCompany,

/* @IdClient = */ e.IdClient,

/* @address = */ COALESCE(a1.address, a2.address)

FROM emp e

LEFT...

My guess would be that somehow task schedule is using a batch size of 1 row (or some very small number of rows). See if you can change the...

I'm not sure what you mean about procs and CDC.

CDC uses the SQL log to get changes, so it shouldn't require use of procs, or even know how any change...

Jeff Moden (1/6/2014)

---

ScottPletcher (1/6/2014)

---

Actually, it is a terrible idea to add an identity to this table.

Depending on your specific data, future inserts could be sequential anyway. Even if not,...

Actually, it is a terrible idea to add an identity to this table.

Depending on your specific data, future inserts could be sequential anyway. Even if not, rows are inserted...

Identity is a terrible idea on that table. Use the existing columns, in whichever order best matches your use of the table: either ( Item, ItemNumber ) or (...

Jeff Moden (1/1/2014)

---

fahey.jonathan (12/12/2013)

---

If you don't want to create a Tally table, you can use sys.columns instead and create a tally number in memory. According to my research, this...

Rando (1/2/2014)

---

Still searching for a solution. Scott's solution yield's syntax errors and Chris' solution yields the entire date field...

Again, I'm just looking for the results to be the following;

Date: 12/30/2013

Time:...

SELECT

CONVERT(varchar(10), CAST(CounterDateTime AS datetime), 101) AS Date,

RIGHT(CONVERT(varchar(20), CAST(CounterDateTime AS datetime), 100), 7) AS Time

FROM table

Jeff Moden (12/24/2013)

---

Awesome! I don't believe anything will...

If you don't have any data longer than 24 bytes (or so), it might be easier just to "brute STUFF" it:

WITH SampleData (MyStr) AS

(

SELECT '12345678'

...

Jeff Moden (12/20/2013)

---

ScottPletcher (12/20/2013)

---

Evil Kraig F (12/20/2013)

---

ScottPletcher (12/20/2013)

---

Admittedly the gain is much clearer when the clustering key is the ISNULLed column. So, let's just flip the clustering keys on...