Viewing 12 posts - 16 through 27 (of 27 total)
It's a bit early in the morning for me to be thinking this hard but here goes...
Use a left outer join
then use a case to decide what data you want...
April 21, 2004 at 8:55 am
I dont think you want to use fetch prev, I think you want to create a temporary variable that holds the previous batch start time. then you can use this...
April 9, 2004 at 6:30 pm
April 9, 2004 at 6:25 pm
quote:
Mikey:Say you have like in David's example a big amout of records containing dates... Say that the joins on that table still...
July 9, 2003 at 6:30 pm
Something like that. I like to have the joined element clustered, it definitely helps join performance. Also, when you are using a where clause it is helpful to...
July 1, 2003 at 12:03 pm
If you are 80% of the time joining on the identity then you will probably want that to be your primary key
If you are joining on identity and using the...
June 30, 2003 at 12:58 pm
David, quick question. When you select from this table are you joining to it or from it? When you join is it based on the identity column?
If you...
June 27, 2003 at 11:23 am
I thought it was common knowledge/best practice to always put text columns last. No one is saying we should rearrange the table design at all, just explicitly select like...
June 27, 2003 at 10:56 am
someone said
I tend to save my clustered index for something that would benefit from it. Normally this would be the primary key but I don't feel there is much to...
June 25, 2003 at 11:09 am
It is going to be dependent on what you are trying to accomplish. The best bet is to use the execution plan and see which you beleive is better...
April 28, 2003 at 11:26 am
Some motherboards send interrupt requests to one CPU only. This can cause this issue.
April 24, 2003 at 10:24 am
First: With a Clustered PK the data is *always* returned in the natual table order, so order by is not necessary. Second: why do we need to get the...
March 5, 2003 at 12:24 pm
Viewing 12 posts - 16 through 27 (of 27 total)