I suggest creating the following indexes to support the query.  If you'd like to adjust other indexes on the tables also, just let me know.

CREATE UNIQUE...

            
July 21, 2021 at 2:14 pm
            
#3912474

            
            
                                                
        

Results attached.

I can't offer index recommendations without seeing the DDL for the tables, including all current index definitions.  The most critical index is the clustered index, so that should be...

• This reply was modified 3 years, 2 months ago by  Jackie Lowery.

Thanks guys.  The coalesce solution is what I needed.  Phils solution would work, except I want to return an empty string if there's less than 2 characters.

I guess i should say the goal here is to do a :

left(column1,len(column1)-2)

and I'm trying to account for Null or Empty fields so the left function doesn't...

I didn't know of the new persisting sample percent feature.  Thanks for that info.  I found a query that shows the date statistics were updated and days since last update. ...

• This reply was modified 3 years, 6 months ago by  Jackie Lowery.

I may not have had to rebuild them, but doing a full sample on the statistics definitely fixed the slowness.

Should I have to do a full sample on...

I may not have had to rebuild them, but doing a full sample on the statistics definitely fixed the slowness.

Should I have to do a full sample on the statistics...

• This reply was modified 3 years, 10 months ago by  Jackie Lowery.

I found that i can add:

	@StatisticsSample = 100 ,

to the script.  I run this script on Fridays at 10PM.  Is there any reason I shouldn't set the...

I finally found what i was doing that had sped up the query.  I rebuilt the index manually.  Normally, my indexing is done by Ola Hallengren's maintenance scripts, but it...

Now I'm really confused. The problem seems to have come back.  It's taking 4 to 5 seconds again.  If I remove the TOP 1 part it returns 32 rows.   When...

• This reply was modified 3 years, 10 months ago by  Jackie Lowery.

Oddly enough, rebuilding the indexes on the wspkg and wspkglin tables fixed the issue.

If the value passed matches the histogram for 66,000 rows, wouldn't the estimate  be high not low.  Again, I'm stupid on this, so sorry if I'm not asking the wrong...

Again.  Thanks for the help.  I don't understand why the estimate is so low.  There's a big difference between 1200 rows and 66,000 rows.  It causes a problem when displaying...