September 14, 2008 at 11:06 pm
First:
Define a function that maps employeeID to col1's value.. If you can create such a function all you have to do is:
update table1
set col1=select convertEmployeeIdToCol1(id)
Secondly: Updating data for 200,000 rows is inviting trouble. Instead, go for limited no of updates at a time. like update few 10,50,100...number of rows at time....
You can also thread the operation with .NET
Hope this helps
September 14, 2008 at 11:18 pm
Hi Guys,
Instead of doing it with query I did it through the coding
Thanks for your help.
It's really valuable help for me thank again 🙂
Thank,
Rohit.
September 15, 2008 at 6:43 am
First:
Define a function that maps employeeID to col1's value.. If you can create such a function all you have to do is:
update table1
set col1=select convertEmployeeIdToCol1(id)
My concern about this from my own experience, is that functions, while they can make a problem appear much simpler, can result in some major performance problems. If the function contains a single SELECT from some table, you have just converted a single implied SELECT (where we are getting our data from for the UPDATE) into 200,000 individual SELECTS.
Scott
September 15, 2008 at 7:25 am
Perry Whittle (9/8/2008)
where existswould be more efficient than
where in
Are you sure about that one? Like I tested it and in this case exists is faster than in?
September 15, 2008 at 10:33 am
Ninja's_RGR'us (9/15/2008)
Like I tested it and in this case exists is faster than in?
thats what i'm asking
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
Viewing 5 posts - 16 through 19 (of 19 total)
You must be logged in to reply to this topic. Login to reply