Viewing 15 posts - 1 through 15 (of 136 total)
Based on your description, Merge replication doesn't sound like the correct option. Merge replication is used when you want changes made to the Subscriber database(s) to be "merged" back to...
January 24, 2012 at 5:10 am
How wedded are you to the schema of your base fact table? Is this a work in progress that you can make some changes to, or is it already in...
January 21, 2012 at 9:10 am
How many rows are in the base fact table? What will the growth rate be? What is the maximum number of periods that will be stored in the fact table?
January 19, 2012 at 5:12 pm
ckoster (12/29/2011)
I have been doing the calculation for DUEDATE in...
December 30, 2011 at 1:19 am
You can't reference a calculated column in another calculated column. However, you could encapsulate the logic of the first computed column in the definition for the second.
ALTER TABLE TableA ADD...
December 29, 2011 at 6:36 pm
I'm puzzled about the inclusion of the HAVING ( COUNT(Column_Name) >= 1 ) bit. Every value that exists in Column_Name will have one or more occurrences, so this doesn't do...
December 21, 2011 at 6:45 am
Ok, I think I was pretty close then with my assumptions. The code I posted should work for whenever you want to find out how much vacation time an employee...
December 19, 2011 at 5:17 am
Ok, taking into the account the issues I mentioned above, here's a possible solution that attempts to be as robust as possible (based on what I know/guess about your underlying...
December 18, 2011 at 11:55 am
This can definitely be done in a more elegant manner, with much better performance to boot. I'll post one possible solution in a moment; first though, I have some questions...
December 18, 2011 at 10:55 am
I think this thread from StackOverflow may explain the behavior you're seeing:
http://stackoverflow.com/questions/126401/sql-server-2005-numeric-precision-loss
From that thread:
Since you multiplied NUMERIC(24,8) and NUMERIC(24,8), and SQL Server will only check the type not the content,...
December 15, 2011 at 6:30 am
GSquared (12/13/2011)
or (simpler):
join CalendarPeriod cp on cp.CalendarDate = OrderDate and (CalendarPeriod = @Period or @Period = ‘All’)
Except, the way the original CalendarPeriod table is structured, I'm pretty sure that...
December 13, 2011 at 8:00 am
I'm not sure about this bit:
join CalendarPeriod cp on (cp.CalendarDate = OrderDate and CalendarPeriod = @Period) or (@Period = ‘All’)
That works fine if @Period != 'All', but as soon...
December 13, 2011 at 7:30 am
drew.allen (12/9/2011)
WHERE H.SALESSTTUS = 1 AND
H.SALESORIGINID IN ('DSM','DSM-Trf') AND
L.STARTDATE < @IndvoiceEnd AND
@InvoiceStart > L.ENDDATE
This also emphasizes the fact that this relationship is commutative....
December 9, 2011 at 8:55 am
No problem. Did the last method I posted not solve the performance problems, or have you just not had a chance to test it yet?
December 9, 2011 at 6:24 am
sqlfriends (12/8/2011)
I hear sort really affect performance, for this number of records, does it matter?
Honestly, I haven't worked with SSIS enough at this point to be able to say...
December 9, 2011 at 4:55 am
Viewing 15 posts - 1 through 15 (of 136 total)