![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
I am trying to find the most optimal approach to load Fact table which has FK(es) to Dimensions(8 dimension) without using Cursors. MSIS or t-sql or C# code or all of above? Highly appreciate any direction/architecture approach/white papers or any source of info to optimize the peformance for upload. ( I have 10 mln at least records for single fact table to be uploaded...cursors...will be done by next week...) thank you, Alex Deiden |
#3
| |||
| |||
|
|
Hello Alex, If you can do it in tsql then do it in TSQL. You would not necessarilty need cursors. Essentially you will take you source data, run this past the dimension tables and extract surrogate keys from the dimension tables where the Natural/Business key maches the input data. You can of course use all the methods you describe but if you can get away with using TSQL then do it. After that my preference is for SSIS. Allan Mitchell www.SQLDTS.com www.SQLIS.com www.Konesans.com I am trying to find the most optimal approach to load Fact table which has FK(es) to Dimensions(8 dimension) without using Cursors. MSIS or t-sql or C# code or all of above? Highly appreciate any direction/architecture approach/white papers or any source of info to optimize the peformance for upload. ( I have 10 mln at least records for single fact table to be uploaded...cursors...will be done by next week...) thank you, Alex Deiden |
![]() |
| Thread Tools | |
| Display Modes | |
| |