Log into your server sometime on sunday while the error is occuring and verify that the backup location is actually available 🙂

WITH TEMP_CTE AS(SELECT [TimeStamp], ROW_NUMBER() OVER(PARTITION BY (SELECT 1) ORDER BY [TimeStamp] ASC) AS ROW_NUM FROM #Time

)

SELECT T_ONE.[TimeStamp], T_TWO.[TimeStamp], DATEDIFF(minute, T_ONE.[TimeStamp], T_TWO.[TimeStamp]) FROM TEMP_CTE T_ONE, TEMP_CTE T_TWO

WHERE T_ONE.ROW_NUM = T_TWO.ROW_NUM...

So was this a question on why single character variable names @i and @j-2 are terrible choices?

Based on your data that update won't work as expected because the rows aren't unique.

Try this.

WITH NumCte AS (

SELECT t.CatName, t.RowNumber,

RN = ROW_NUMBER() OVER(PARTITION BY t.CatName ORDER BY t.SubCat)

FROM #TEST...

SELECT DATEADD(year, -1, DATEADD(MONTH, DATEDIFF(MONTH, '19000201', GETDATE()), '19000101')) AS StartDate1,

DATEADD(year, -1, DATEADD(MONTH, DATEDIFF(MONTH, '19000101', GETDATE()), '18991231')) AS EndDate1

dan-572483 (3/16/2015)

---

I never suggested that a weekly shrink job on a production database was a sound practice. I just thought that the commandment "NEVER shrink a database!" was a bit...

Transaction log backup.

Backing up the log does not reduce the file size, so at some point you used 500GB of log space between transaction log backups. If you need to reclaim...

If you're actually expecting data with " in it can you change the file format from say comma delimited with " qualifiers to something else like just | delimited?

Well no you haven't aliased your columns the same(you aliased your table) so SQL Server is using the column name of the first select which is why the order by...

That should...

Sean Lange (3/12/2015)

---

This thing screams of denormalized data when I see columns like IRSORRES6, IRSORRES7, IRSORRES8. If you can normalize the data you could eliminate the need to have 3...

If it's stored as a varchar you'll need to convert it to a datetime first then convert it to the formatted date you want.

The reason why shrink database should typically be avoided is that if the database grew to that size it indicates it needed that much space for some reason so shrinking...

The relatively simple way is to put your check for duplicate data into a CTE then just delete from that CTE where ROW_NUMBER != 1.