Viewing 15 posts - 3,301 through 3,315 (of 6,036 total)
But if capacity is more important for you than precision to the last digit then you probably will go with FLOAT result.
For N = 115 it returns 2.9250936934930141E+188 and still...
November 6, 2007 at 8:59 pm
There is a solution without loops - single "select" (according to what "SQL" stands for).
It involves very famous table Tally (regards to you know who ;))
[Code]
DECLARE @I BIGINT
select @I...
November 6, 2007 at 6:56 pm
David McFarland (11/6/2007)
Sergiy (11/5/2007)
That you were wrong. No big deal, as I've done it before, and I'll do it again (based on your "maths" skills, probably...
November 6, 2007 at 12:52 pm
After TAble B is populated:
CREATE VIEW dbo.mytable AS SELECT * FROM dbo.tableB
Then do data update in tableA
After it's finished:
DROP VIEW dbo.mytable
CREATE VIEW dbo.mytable AS SELECT * FROM dbo.tableA
November 6, 2007 at 12:11 pm
Matt Miller (11/5/2007)
Sergiy -On the INSERT part... you SURE you want to NOT include the keycol?
I'm assuming that's a cut/paste error?
How did you know?
😀
November 5, 2007 at 9:04 pm
wackoyacky (11/5/2007)
But wouldn't it be an additional hit on the transactional tables, especially if the traffic on that table is very high? But I will consider it as...
November 5, 2007 at 8:56 pm
Let's assume that:
TheTable is your table you update;
StTable is Staging Table you just inserted;
KeyCol - one or more columns uniquely identifying rows in TheTable.
[Code]
DELETE FROM TheTable
WHERE NOT EXISTS (select 1...
November 5, 2007 at 7:57 pm
John N (11/5/2007)the data is not really that much but the problem is it's coming from a vendor api with a standard data for all their users. Unfortunately they don't...
November 5, 2007 at 6:34 pm
David McFarland (11/5/2007)
You don't believe that covering indexes are a representation of what a real production system would have.
I said no such thing. I said they're not a representation of...
November 5, 2007 at 3:38 pm
David, what did you prove?
My statement was exactly this:
Probably 2k5 is more comfortable for procedural language programmers, but for those who understand relational model and use to operate with datasets...
November 5, 2007 at 3:27 pm
wackoyacky (11/5/2007)
Anyone there who have done report tables in SQL 2000? How did you denormalized your tables?I'll be doing some test now. Thanks for the input guys.
Why tables?
Views...
November 5, 2007 at 1:58 pm
John, how much of you data is actually changed during those updates?
From my experience it's typically 5-10% of the total amount of data.
Then you need to upload into some staging...
November 5, 2007 at 1:52 pm
Matt, there is no any risk.
Just do it proper way.
To make sure the way is really proper copy proven solution.
Open this table in "design table" window in EM.
Change datatype for...
November 5, 2007 at 2:54 am
Steve Jones - Editor (11/4/2007)
I like Sergiy's copy/drop/rename solution if it works.
I did not invent it.
😉
It's proven working by years of thousands developers experience working with EM:
this approach...
November 4, 2007 at 9:04 pm
Good story about "copy-drop-rename" approach is that it does not require any stoppage during copy process.
After it's finished you may revisit the data and find what's changed since you've started...
November 4, 2007 at 3:57 pm
Viewing 15 posts - 3,301 through 3,315 (of 6,036 total)