May 3, 2007 at 9:32 am
i know each suituation is different, but are there any rules/steps to go through or things to look out for when converting a cursor into set based logic?
May 3, 2007 at 10:39 am
Obviously, When the situation deals with a lot of records, it is always best to use the set based logic rather persisting with CURSORS. Using Set based logic would give a better edge over the cursor option.
May 3, 2007 at 10:41 am
the critical obvious one is that you still get the same end result - I know this sounds rather silly but knowing what you should get and making sure you do when you change code is important.
I'd also make sure that there are savings made - I'm not a great lover of cursors, but there's no point spending hours or weeks to change a piece of code just for the sake of it - and of course remember "if it ain't broke don't fix it!"
[font="Comic Sans MS"]The GrumpyOldDBA[/font]
www.grumpyolddba.co.uk
http://sqlblogcasts.com/blogs/grumpyolddba/
May 25, 2007 at 12:34 pm
Not exactly sure what u mean by set based however i like using home made cursors as oppose to implimenting real cusors.
load into a mem table with what would normally be loaded into the cursor w/ an added field at the end labeled processed and this gets set to zero for every record
while exists ( select top 1 * from mem_table where processed - 0)
begin
Store all fields from the record your working w/ into local variables
do whatever u would normally do w/ these values
update memtable
set processed = 1
where id =id
end
these are really useful
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply