June 2, 2017 at 6:41 am
Hello,
I passed my exam 70-467 and I had this question, please do you know the reply ?
Thanks.
You are designing a SQL Server Analysis Services cube un SQL Server for a company that sells insurance plans.
The insurance plan data is stored in a SQL Server data warehouse. Billing data for the insurance plan is extracted from an Oracle database. You need to recommend a solution to merge the insurance plan data and the billing data into a single cube. What should you include in the recommendation ?
Answer area :
Number of data source views : ?
Minimum number of measure groups : ?
Minimum number of partitions in the cube : ?
June 6, 2017 at 9:56 am
mokojo51340 - Friday, June 2, 2017 6:41 AMHello,
I passed my exam 70-467 and I had this question, please do you know the reply ?
Thanks.You are designing a SQL Server Analysis Services cube un SQL Server for a company that sells insurance plans.
The insurance plan data is stored in a SQL Server data warehouse. Billing data for the insurance plan is extracted from an Oracle database. You need to recommend a solution to merge the insurance plan data and the billing data into a single cube. What should you include in the recommendation ?
Answer area :
Number of data source views : ?
Minimum number of measure groups : ?
Minimum number of partitions in the cube : ?
This sounds like a bad test question, but I can answer it with utmost confidence in saying, "It depends".
If this were a real work question, you most certainly wouldn't be focusing on details like how many data source views. I would imagine that there would be many higher level questions to answer before you got down into the weeds. E.g, do the Oracle and SQL Server systems contain the same business keys to identify plans?
Rob
June 6, 2017 at 12:20 pm
Actually as the questions asks for minimum number of.. and also asks for a single cube then the answer isn't "it depends".
The answer is as follows (assuming I can remember my SSAS training)
Number of data source views : 1
Minimum number of measure groups : 2 - 1 for the Insurance Plan Data fact table, 1 for the Billing Fact table
Minimum number of partitions in the cube : 2 - 1 for each measure group.
I agree never something you would be asked in the real world but I think a reasonable question to ensure you know DSVs are related to cubes and how measure groups are related to fact tables etc.
June 6, 2017 at 12:44 pm
FridayNightGiant - Tuesday, June 6, 2017 12:20 PMActually as the questions asks for minimum number of.. and also asks for a single cube then the answer isn't "it depends".
The answer is as follows (assuming I can remember my SSAS training)Number of data source views : 1
Minimum number of measure groups : 2 - 1 for the Insurance Plan Data fact table, 1 for the Billing Fact table
Minimum number of partitions in the cube : 2 - 1 for each measure group.
I agree never something you would be asked in the real world but I think a reasonable question to ensure you know DSVs are related to cubes and how measure groups are related to fact tables etc.
Thank you very much for your help. !
June 7, 2017 at 5:44 am
FridayNightGiant - Tuesday, June 6, 2017 12:20 PMActually as the questions asks for minimum number of.. and also asks for a single cube then the answer isn't "it depends".
The answer is as follows (assuming I can remember my SSAS training)Number of data source views : 1
Minimum number of measure groups : 2 - 1 for the Insurance Plan Data fact table, 1 for the Billing Fact table
Minimum number of partitions in the cube : 2 - 1 for each measure group.
I agree never something you would be asked in the real world but I think a reasonable question to ensure you know DSVs are related to cubes and how measure groups are related to fact tables etc.
>>>The insurance plan data is stored in a SQL Server data warehouse. Billing data for the insurance plan is extracted from an Oracle database. You need to recommend a solution to merge the insurance plan data and the billing data into a single cube. What should you include in the recommendation ?
I stand by my "it depends". What if the fact table has 100 records, or 100 billion records (over multiple years). Would you recommend 1 partition for that measure?
Rob
June 7, 2017 at 3:26 pm
robert.gerald.taylor - Wednesday, June 7, 2017 5:44 AMFridayNightGiant - Tuesday, June 6, 2017 12:20 PMActually as the questions asks for minimum number of.. and also asks for a single cube then the answer isn't "it depends".
The answer is as follows (assuming I can remember my SSAS training)Number of data source views : 1
Minimum number of measure groups : 2 - 1 for the Insurance Plan Data fact table, 1 for the Billing Fact table
Minimum number of partitions in the cube : 2 - 1 for each measure group.
I agree never something you would be asked in the real world but I think a reasonable question to ensure you know DSVs are related to cubes and how measure groups are related to fact tables etc.
>>>The insurance plan data is stored in a SQL Server data warehouse. Billing data for the insurance plan is extracted from an Oracle database. You need to recommend a solution to merge the insurance plan data and the billing data into a single cube. What should you include in the recommendation ?
I stand by my "it depends". What if the fact table has 100 records, or 100 billion records (over multiple years). Would you recommend 1 partition for that measure?
Rob
It is an exam question though. It doesn't give you any information on data sizes or even where the data resides. You have to forgot real world and figure out what the question is trying to establish that you know.
If you were setting the exam and wanted to establish that the exam taker is aware that measures are grouped by fact table into Measure groups and that by default a measure group will be a single partition (we will ignore the DSV part) how would you word the question?
Totally agree in real world the answer is it depends but on an exam question the answer can never be it depends.
Viewing 6 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply