![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
I was hoping I can get some help regarding subject mater located at http://msdn.microsoft.com/library/de...es_02_92k3.asp |
#3
| |||
| |||
|
|
"TZone" <TZone (AT) optus (DOT) com.au> wrote in message news:3f2e4086$0$14563$afc38c87 (AT) news (DOT) optusnet.com.au... I was hoping I can get some help regarding subject mater located at http://msdn.microsoft.com/library/de...es_02_92k3.asp snip You're correct that float and datetime are fixed length, and both are indeed 8 bytes (for your float precision, as you noted), which seems to be your main question. You just need to follow the formulas in the documentation. One point you may need to consider is how full the average variable length column will be - if you know that almost all data in an nvarchar(100) column will be around 50 Unicode characters, ie. half the length of the field, then you could count that as 100 bytes instead of 200 to get a more accurate estimate. Or if the average will be around 75, then count 150 etc. But you're the only person who knows what your data looks like, so you're the best person to calculate the size of the table. Simon |
![]() |
| Thread Tools | |
| Display Modes | |
| |