Viewing 15 posts - 961 through 975 (of 1,353 total)
It is possible to set dimensions so that only certain items show depending on the user. This is too complicated to go into in this environment. You need...
May 17, 2011 at 9:54 am
fact_, dim_ and stg_... can we make this an unofficial standard and pass it along as accepted practice when teaching others? I'm in.
I use ftblName, dtblName, and my...
May 17, 2011 at 9:50 am
Without data to analyze, these can only be suggestions, but there is some chance that supplier and country are related. Commodity type is probably independent, so I think you...
May 13, 2011 at 4:56 am
Almost all your comments seem geared to SSIS. Here are some from SSAS that I wish I'd known at the start.
1. Beginning in SSAS 2005, the same dimension can...
May 12, 2011 at 2:39 am
My experience has been that when many developers and DBAs say denormalized, they actually mean unnormalized. You can't reverse what you never did in the first place...
Completely agree
Atomicity is...
May 4, 2011 at 12:49 pm
As no one has posted, I'll just say I hope you solved this. I have never had an issue with using Refresh All to bring any updates into preexisting...
April 29, 2011 at 3:26 am
Oh how I love (fast and easy) denormalized data; just write your code well enough to keep the tables in sync.
That is asking a lot. In a large...
April 29, 2011 at 2:52 am
That's why I feel that normalization rules are really "mostly guidelines"
"Normalization" is a matter of differering orthodoxies, really. I've seen arguments about whether NULLs should be allowed in "normalized data"...
April 28, 2011 at 4:37 am
Nice question and an interesting example. Would renaming the columns "Telephone 1" and "Telephone 2" to "Home Phone" and "Mobile Phone" mean that the table was then in 1NF?
I...
April 27, 2011 at 3:53 am
The key issue in deciding if something can be in one cube or broken out is the granularity of the item measured. If the measures are on the same...
April 27, 2011 at 3:48 am
I do like your approach to implement separate markers for "A" and "I" Null Values but I'm a little concerned about the use of zero - as one of them...
April 27, 2011 at 3:17 am
Sorry, but in my view, it's not a valid option. I appreciate this isn't universally accepted, but there are many who would say the same thing.
April 27, 2011 at 3:14 am
I don't know what you mean by restricting in this case. I find that the system works very well. One of the fields I track BTW is status....
April 26, 2011 at 3:30 am
First create a package level variable. Insert a row count transformation between the start and end points of the data flow. Assign the output to the variable that...
April 26, 2011 at 3:26 am
I would add a record counter to the error stream. Once done, check the value. If it's more than zero, you have errors.
April 25, 2011 at 6:47 am
Viewing 15 posts - 961 through 975 (of 1,353 total)