I suggest making index changes scripted below, following these steps:

1) Capture and save the existing index missing/usage stats for that table immediately, before any hanges are made.

2) Run the code...

1) Maybe you can combine the three separate UPDATEs into one, as shown below, avoiding repeated joins of the same tables.

2) Depending on the row counts, you might consider creating...

You could also consider something like below, particularly since presumably you are on Enterprise Edition and thus can recreate all the non-clustered indexes ONLINE:

Before the shrink:

1) verify you have current...

Try changing:

!= ''

to

IS NOT NULL.

The variables should contain NULL, not be empty, if no row is found.

For performance reasons, it would almost certainly be better to have a trigger on the Calculations table that automatically identified and saved all variables used in a normalized table whenever...

Maybe something like this?:

declare @replace1 varchar(max)

declare @replace2 varchar(max)

select @replace1 = (

select 'REPLACE('

from #CalculationVariables

for xml path('')

...

FYI: the wheel's been re-invented countless times, or we'd all be riding around on wooden wheels with no rims!

SELECT RTRIM(LEFT(CustomerName,CHARINDEX('*',CustomerName + '*')-1))

As long as you have fixed leading values, i.e. 'value1%' and not '%value1%',

an index on myColumn should handle the query fine. If it's a large % of the...

If you're using Enterprise Edition, just use Change Data Capture and let SQL do all the hard work for you.

If not, the overhead of what you want to do is...

Sorry, extremely busy, but here's the re-write. I couldn't test it, of course, so you'll have to do that ;-):

Set NOCOUNT ON

If Object_ID('tempdb..#auditlogidsToArchive') > 0

Drop...

Need to clarify whether the other part of my initial impressions was true or not:

Is logtimestamp from the original insert, and so will also always ascend in conjunction with the...

Stamey (12/4/2013)

I have figured it out, thanks to Google and an example from guys from 'rola. ... I changed it around so that the commit is in the Try and...

The quick and very-dirty-in-this-case "solution" is to put a COMMIT TRANSACTION immediately before the END TRY.

But the whole process overall could be a lot cleaner and faster, particularly if:

auditlogid is...

keyser soze-308506 (11/27/2013)

I have a query like that

select c.client_group, s.product_id, count(*) as q

from Sales s

inner join Client c on c.client_id= s.client_id

group by c.client_id, s.product_id

the Sales.pk (clustered index) consist of...