Viewing 15 posts - 20,521 through 20,535 (of 22,202 total)
Are the updates and the reads against the same data? Before you do something drastic (the very next post is probably going to suggest using a query hint on the...
April 30, 2008 at 6:24 am
Lots better.
If nothing else, instead of dynamic queries, use a table valued function to pivot your comma delimited list into a table and then join against it. It's not always...
April 30, 2008 at 6:18 am
Use Modulo 4 and Modulo 6 to limit the values to quarterly & every six months. Of course, you can only use June & December for the six month breaks.
April 30, 2008 at 6:16 am
Parameter sniffing?
Or, if we've got a procedure that consists of views, calling views, calling views... bad execution plans?
Look at the execution plan when it performs badly and when it...
April 30, 2008 at 6:15 am
svhanda (4/30/2008)
I had problem in updating a table ...
My query is suppose there is table like
EMP_ID Comp_ID
100 ...
April 30, 2008 at 6:06 am
There is no way to "link" databases in SQL Server so that you can have cross-database referential integrity. Period. End of Story.
However, you could write a trigger that will validate...
April 30, 2008 at 6:02 am
Just run "File", "Export", "Script Trace Definition", "SQL Server 2005" and you can define the file from there. Attach it to a message (cause posting it would be ugly).
April 29, 2008 at 1:03 pm
Wild.
That's sure a surprise that you shouldn't run trace in production... Since I've got it running off & on on most production servers and I have yet to see...
April 29, 2008 at 12:51 pm
We carried straight jackets... That's all I'm going to say about that.
April 29, 2008 at 11:23 am
Short answer: it depends.
Slight longer answer: When looking at a system like this, you have to decide, is it a single integrated system or a series of apps that have...
April 29, 2008 at 11:22 am
OK.
Then without details... I think the insert & update is better.
April 29, 2008 at 10:32 am
I agree. You should minimize what is maintained there to only what is needed to run the processes that move the data in & out. Assuming it's not also used...
April 29, 2008 at 9:28 am
He's right. With very, very few exceptions, every table should have a clustered index. Further, every table should have a primary key and that constraint can be clustered (since it...
April 29, 2008 at 9:25 am
Kevin Durham (4/29/2008)
April 29, 2008 at 9:23 am
Viewing 15 posts - 20,521 through 20,535 (of 22,202 total)