Sap Interview Questions -part-2

Sunny Sep 2, 2007

22. What are the nine decision points of data warehousing?
a. Identify fact table
b. Identify dimension tables
c. Define attributes of entities
d. Define granularity of the fact table (how often
e. Pre calculated key figures
f. Slowly changing dimensions
g. Aggregates
h. How long data will be kept
i. How often data is extracted

23. How many dimensions in a cube
Total 16 out of which 3 are predefined time, unit and request; customer is left with 13 dimensions

24. What is a SID table and advantages
The SID table (Surrogate ID table) is the interface between master data and the dimension tables; advantages :-
a. uses numeric as indexes for faster access
b. master data independent of info cubes
c. language support
d. slowly changing dimension support

25. What are the other tables created for master data?
a. P table - Time independent master data attributes
b. Q table - Time dependent master data attributes
c. M view - Combines P and Q
d. X table - Interface between master data SIDs and time independent navigational attributes SIDs (P is linked to the X table)
e. Y table - Interface between master data SIDs and time dependent navigational attributes SIDs (Q is linked to the Y table)

26. What is the transfer routine of the info object?
It is like a start routine; this is independent of the data source and valid for all transfer routines; you can use this to define global data and global checks.

27. What is the DIM ID?
Dim ids link dimensions to the fact table

28. What is table partition?
SAP is using fact table partitioning to improve performance; you can partition only on OCALMONTH or OFISCPER;

29. How many extra partitions are created and why?
Usually 2 extra partitions are created to accommodate data before the begin date and after the end date

30. Can you partition a cube which has data already?
No; the cube must be empty to do this; one work around is to make a copy of the cube A to cube B; export data from A to B using export data source; empty cube A; create partition on A; re-import data from B; delete cube B

31. What is the transaction for Administrator work bench?
RSA1

32. What is a source system?
Any system that is sending data to BW like R/3, flat file, oracle database or external systems.

33. What is a data source?
The source which is sending data to a particular info source on BW; for example we have a OCUSTOMER_ATTR data source to supply attributes to OCUSTOMER from R/3

34. What is an info source?
Group of logically related objects; for example the OCUSTOMER info source will contain data related to customer and attributes like customer number, address, phone no, etc

35. What are the types of info source?
Transactional, attributes, text and hierarchy

36. What is communication structure?
Is an independent structure created from info source; it is independent of the source system/data source

37. What are transfer rules?
The transformation rules for data from source system to info source/communication structure

38. What is global transfer rule?
This is a transfer routine (ABAP) defined at the info object level; this is common for all source systems.

39. What are the options available in the transfer rule Assign info object, assign a constant, ABAP routine or a Formula (From version 3.x); example are :
a. Assign info object - direct transfer; no transformation
b. Constant - for example if you are loading data from a specified country in a
f lat file, you can make the country as constant and assign the value
c. ABAP routine - for example if you want to do some complex string manipulation; assume that you are getting a flag file from legacy data and the cost center is in a field and you have to "massage" the data to get it; in this case use ABAP code
d. For simple calculations use formula; for example you want to convert all lower case characters to upper case; use the TOUPPER formula

40. Give some important formula available
Concatenate, sub string, condense, left/right (n characters),1_trim, r_trim, replace, date routines like DATECONV, date-week, add_to_date, date_diff, logical functions like if, and;

41. When you do the ABAP code for transfer rule, what are the important variables you use?
a. RESULT - this gets the result of the ABAP code
b. RETURNCODE - you set this to 0 if everything is OK; else this record is
skipped
c. ABORT - set this to a value not 0, to abort the entire package

42. What is the process of replication?
This copies data source structures from R/3 to BW

43. What is the update rule?
Update rule defines the transformation of data from the communication structure to the data targets; this is independent of the source systems/data sources

44. What are the options in update rules?
a. one to one move of info objects
b. constant
c. lookup for master data attributes
d. formula
e. routine (ABAP)
f. initial value

45. What are the special conversions for time in update rules?
Time dimensions are automatically converted; for example if the cube contains calendar month and your transfer structure contains date, the date to calendar month is converted automatically.

46. What is the time distribution option in update rule?
This is to distribute data according to time; for example if the source contains calendar week and the target contains calendar day, the data is split for each calendar day. Here you can select either the normal calendar or the factory calendar.

47. What is the return table option in update rules for key figures?
Usually the update rule sends one record to the data target; using this option you can send multiple records; for example if we are getting total telephone examples for the cost center, you can use this to return telephone expenses for each employee (by dividing the total expenses by the number of employees in the cost center) and creating cost record for each employee in the ABAP code.

48. What is the start routine?
The first step in the update process is to call start routine; use this to fill global variables to be used in update routines;

49. How would you optimize the dimensions?
Use as many dimensions as possible for performance improvement; for example assume that you have 100 products and 200 customers; if you make one dimension for both, the size of the dimension will be 20,000; if you make individual dimensions then the total number of rows will be 300. Even if you put more than one characteristic per dimension, do the math considering worst case scenario and decide which characteristics may be combined in a dimension.

50. What is the conversion routine for units and currencies in the update rule?
Using this option you can write ABAP code for unit/currency conversion; if you enable this flag, then unit of the key figure appears in the ABAP code as an additional parameter; for example you can use this to convert quantity in pounds to kilo grams.

(7) Comments

  1. Guest Mar 8, 2008

    Both - The BW Tutorial and the BW Interview Qs - Are
    Excellent Source of Info under one place ...Thanks a Lot,
    Sandeep !! Keep up the Good Work..!
    God Bless You !!
    - Ann

  2. Guest Apr 18, 2008

    There is an important exception to the “avoid surrogate keys” principle. If the table’s primary key is large, the non-clustered indexes are also large, so non-clustered indexes become much less efficient. Not only is a non-clustered index less efficient, the value that results from the non-clustered index’s seek is large too, so navigating the primary key is slower, too.

  3. Guest Jul 24, 2008

    You have got a really good set of questions here,very useful for a lot of sap'rs ,keep up the godd work ,can you include some more questions and also try to differentiate the question from the answer,,just a suggestion,,

  4. Guest Aug 21, 2008

    sap interview

  5. Guest Oct 29, 2008

    I highly agreed with you man. You took the words out from my mouth :)

  6. Guest Jan 21, 2009

    Thank you for the information. I am having the same problem right now. Worked great.

  7. Guest Jan 21, 2009

    Very good job! I will be using this is an upcoming presentation I have to do.