Thanks for the article Adam. It is well thought out, and well presented and written. I have used a exists then update or add methodology. I've only ever added extended...

I would use sys.sql_modules or INFORMATION_SCHEMA.routines. However, more importantly than that, I would convert the @@IDENTITY to the output parameter method suggested by Microsoft since there is a bug related...

I agree, there is a lot of confusion about logging with truncate table. Let's discuss your scenario related to the issue at hand.

In your scenario about rolling back transactions that...

Truncate Table is logged, it's just that the pages are deallocated instead of each row being logged as with the delete. For proof of this, try rolling back a truncate...

I agree with theSQLGuru - ERStudio can be very helpful in larger models. I am curious why you focus on the big table(s) and the one(s) that are most frequently...

Great article! I am looking forward to the future articles in the series. I wonder if power-pivot deals adequately with rendering axis beyond the second in excel. I worked with...

Oh, there is one issue I wanted to discuss. It's not a problem with your article, but a problem I have struggled with in the past.

I often set the file...

Excellent article. I also wanted to let you know about your additional predicate, but you've already identified it. I have written server side traces like this for years, but I...

Excellent article Gail. You are pro, and pro's lean on other pros. Nice to see such a practical treatment of log reuse while still giving significant background information in such...

Thanks for the article. I try not to use distinct, especially in multi joined select statements, but your article delivered on what you set out to deliver as far as...

Great artical Gail! Your commentary about intersections and the factors around whether to place the most selective column in the leading edge of the index are very practical and helpful....

Lowell, you are exactly correct, Bhuvnesh, I am only interested in whether it can be done, and if so what the procedure(s) is/are.

The way I see it is like getoffmyfoot put it where in your outer proc you submit an insert/exec statement.

--outer proc code

CREATE TABLE #temp (columns from your inner proc)

INSERT...

True, the heap/table/index will be scanned with the leading wildcard, but I am wondering if the optimizer plans for a full row count or a rowcount based on an average...

it sounds like a poor execution plan is being created. Usually if statistics are up to date the optimizer won't choose the wrong physical join operators because the estimated row...