Sounds like possibly a permissions issue. Are the procs created with different schema's? Do you have permissions to those schema?

You don't want the second one.

If the procedure is called, does it get the parameters as you outlined or does it get the literal values? Plan guides are extremely picky....

The answer is the error message. You're trying to insert data that already exists and the primary key constraint, which ensures that the value in a column or columns is...

Either bulk insert or some method of INSERT... SELECT... You can see examples in the Books Online.

I wouldn't recommend doing this on a production system, but you could capture statement completion events from a trace. Then you'll see when the call is made from proc b.

However,...

Yeah, I saw the app. I do understand what you're describing, but I'm 100% with Gail on this. You need to gather statistics to understand exactly where the slow-downs are...

Where you see proc, substitute query. Sorry I mistyped.

And yes, you do need to determine where the slow down is occuring because it's not from compile time on such simple...

DBCC DROPCLEANBUFFERS should only get run as a test. That removes all the data from memory. Based on the queries you've shown, compiles aren't your issue. You're looking at memory...

dacto77 (2/5/2009)

---

Tsup Gente!

The output of my store has 2 tables, how can I insert these 2 tables into another 2 temp tables to manipulate the data? I use this store...

SCOPE_IDENTITY will work, but only if you ever insert a single row. If you insert two or more rows, SCOPE_IDENTITY won't adequately cover the situation. That's what the OUTPUT clause...

In addition to simply observing performance degrade, have you collected information from performance monitor to see if buffer cache hit ratio (just for one example) is maintaining over time or...

That is scary. However, I wouldn't worry too much (worry some, yeah).

I'm hesitant to suggest what to do since this is PeopleSoft, but let's assume for a moment that it's...

Possibly. But I think there might be some samples for how to do it locally.

Run the query with an explicit open transaction statement:

BEGIN TRANS

...

Don't put in a commit until you're complete with your testing.

To give you a complete answer, you'd need to post the code, deadlock graphs, database structures, etc.

To make a few suggestions, Does the table have any other clustered index? If...