InfoCube Design

dear gurus, I am familiar with creating standard Infocube's in BI7. I have 2 dimensions in this cube and 3 infoobjects that I have inserted in these dimensional. This is a flat file loading. Now how will I know which infoobject will be acting a primary key/foreign key in the fact table?

Hi.....
Look in the Dimension table...........key field will be DIM id.......
And in Fact table key field ........Combination of DIMids......
Actually Dimension table will contain........DIM id and SID...with this SID............DIM table will be connected to the SID table of the Master data.........It is a concept of Extended star schema.........
So u don't need ti think about the key field...........Since u hav three Infoobjects only.............use Line item Dimension...........This means the dimension contains precisely one characteristic. This means that the system does not create a dimension table. Instead, the SID table of the characteristic takes on the role of dimension table.................ie create three Dimension and include one infoobject in each Dimension......
Removing the dimension table has the following advantages:
a) When loading transaction data, no IDs are generated for the entries in the dimension table. This number range operation can compromise performance precisely in the case where a degenerated dimension is involved.
b) A table- having a very large cardinality- is removed from the star schema. As a result, the SQL-based queries are simpler. In many cases, the database optimizer can choose better execution plans.
Nevertheless, it also has a disadvantage: A dimension marked as a line item cannot subsequently include additional characteristics. This is only possible with normal dimensions........So remember this point...........but it will improve performance.........
Check this..........
http://sapbwneelam.blogspot.com/2007/10/extended-star-schema.html
http://syedtayyabali.blogspot.com/2007/11/sap-bwbi-extended-star-schema.html
Regards,
Debjani.....

Similar Messages

InfoCube design - high granularity by inclusion of document level detail

When considering the inclusion of sales document level detail in an infocube (0DOC_NUMBER/0S_ORD_ITEM), two possible high-level designs come to mind:
1. Make sales document number/sales order item master data carrying infoobjects and include in infocube as line item dimensions. Activate all necessary characteristic navigational attributes for reporting.
2. Model infocube dimensions to contain all necessary charactistics for reporting with inclusion of characteristics sales document number/sales order item therefore, fact table will have highest level of granularity.
Knowing the 'textbook' does not suggest the above designs, has anyone attempted maximum granularity in infocube design with success or otherwise? If successful, has this design outperformed DSO or InfoSet reporting performance?
Thank you - Brendan.

Hi Brendan,
we have implemented both solutions; it all depends on the number of documents. if it is much more than 1 mio; I recommend storing your data in the target; if it is 1 or 2 mio you can have master data (more flexible).
We have some 4 mio records in 0MAT_PLANT and are starting having performance issues.
by the way; if you are using ORACLE, NEVER mark a dimension with high cardinality; that doesn't work.
hope that helps
Olivier.
Message was edited by:
Olivier Cora

Infocube Design/Development

Hi,
What are the factors we consider for designing Info cube and also what are factors we should consider in the requirments gathering for infocube design
and
How to choose an infoobject to declare it as line item dimension
Thanks

Hi BW Developer,
As for requirements gathering you may want to go through the blog: http://www.sdn.sap.com/irj/scn/weblogs?blog=/pub/wlg/15737. [original link is broken] [original link is broken] [original link is broken]
When desinging an InfoCube, U need to consider factors such as:
- What purpose the report will be used for reporting?
- What are the characteristics & KeyFigures required in the user report as per requirements
- What is the amount of data that will be used for reporting?
- What is the frequency for user reporting?
- How many users are expected to use the report at a given time?
- How do u seperate (design) dimensions using 1:n relationships.
- Perfomance measures such as Aggregates, Line Item dimensions, etc.
These are some of the pointers that can be considered.
Hope the above helps.
Regards,
Kunal G

InfoCube Design and Partitioning

Hi All, when we design an InfoCube, like when we decide which craracteristics to go in which dimension, does partitioning decided at this time as well? If os, please provide some details?
Can somebody guide me on this? thanks in advance.

Hi,
Both Infocube design and Partitioning will help in improving the performance.
Partioning will be decided based on the data to be stored in the Infocube. If for example, lets assume that the cube will hold 15 yrs of data. As you know, partiontioning can be done only based on Fiscal Year or Calmonth
If partioning is done on Fiscal Year, we might decide to go for 15 partitions as we have 15 yrs of data. If we decide to go for 7 partitions, then 2 yrs of data will be stored in each partition. Like wise for Calmonths, we might go for 15*12 partitions.
If your system is 3.5, data should not be there to do partioning, if your system is BI7, partioning can be done even if data is present in the cube.
Regards,
Sekhar

Infocube Design question

Hi, experts
We have two infocube and we need to know which is best option for customer. First infocube was designed with eleven (11) characteristics and three (3) key figures, the other one with four (4) characteristics and forty-three (43) key figures. Both infocube upload data from the same source correctly, but we need to know wich design is best for performing?
Regards
Tokio
P.D. Points rewards

Tokio,
Without knowing the characteristics and key figures is a bit complicated to tell.
But, i guess that if you use first option, you'll need to create calculated and restricted key figures in the queries.
In the second case, i guess you make those calculations and restrictions in the data loading (transformation or rules). Then, in the queries you just include the KF.
so, based on this, you have:
For loading process cube 1 is faster than cube 2.
For reporting, cube 2 is faster than cube 1.
Also, cube 1 will be more flexible to KF definition changes, since the data is stored as it's generated in the source system, you only need to change the query. This can also be aplied to older data. In cube 2, you need to modify the transformations (or rules) and then reload if you want to update older data.
The best option depends on data volume and the query performance that you need.
Hope this helps.
Regards,
Diego

InfoCube Design for Variable data - Use of Line Item Dimensions

I have an infoprovider based on billing conditions which we have extended the extractor structure for 2LIS13_VDKON and we now have a requirement to add Customer fields such as Customer Purchase Order Number and Contract Number. These fields are obviously highly variable. I have added to them to the reporting DSO and now need advice on what is the best way to add these types of fields as reportable dimensions to the infocube so as to not impact performance? I currently have 9 dimensions with multiple charachteristics and a time dimension. Should I just create a line item dimension for Purchase Order? Problem is I have 8 other line item dimensions to add which are customer specific reporting fields that we capture on the sales order and wish to report on. I know there is a limit of 16 dimensions and I am also concerned about performance.
Any advice is greatly appreciated
Lee Lewis

Hi,
To make sure that the infocube you have created should not have any performance issue: Please do the following
Go to RSRV > All elementary tests-> Database---> database information about infoprovider tables
Upon clicking on database information about infoprovider tables, on the right hand side, in the parameter enter your infocube name and execute and see the log: ( log will automatically popping up once you are done with execution )
There see the database infomation about infoprovider:
This log wil make you to understand how well you have designed your infocube and make sure that each (f ) table corresponding to each dimension will not exceed 20 % of the infocube size.
You create dimensions of infocubes in a such a way ( whether line item or normal dimension ) so that any of the dimensional F table will not exceed 20 % of the infocube size.
Actually this will give us the information of the size of the data of particular dimension and there by if any particular dimension is exceeding the 20 % of the infocube size, then you need to create line item dimension for the characteristics existing in that dimension .
After creating again, test it and see whether any of the dimension table exceeding 20% infocube size .
Repeat this process until you see all dim F tables less than 20 % of the info cube table size
This will negate any performance issues arise in reporting
Edited by: S Simran on Nov 6, 2009 10:11 AM

InfoCube design- partition of chararcters over IC - compression rate etc

Hello,
prerequiste for aggregation in an IC is that technical key is identical. technical key is formed of all DIM IDs and
those in turn are formed by characters in appropriate dimension. That means all characters of all dimensions form up the
technical key. So I have to organize an appropriate set of characters in an IC to get max. compression rate. That
partitioingn may result in to many InfoCubes if I consider semantic or project aspects. On the other hand aggregates are also
a means for realizing compression. Could you support me with guidelines for that or support me with links for this
subject.
Thanks in advance
Oskar

Arun,
thanks for your answer but I think I must change my misleading question to get the answer I want to have
Assumption :
1) InfoCube A has dimension d1 with characters ch1 and ch2 and dimension d2 has character ch3 and
ch4.
2) If ch1 and ch2 and ch3 and ch4 are identical standard aggregation is done ex. sum
Lets say thats what I want to have as aggregation level and which I get If I compress table F to table E
Til now I used an IC with 4 chararcters in 2 dimensions. Now I assume that I have additional 20 characters which
I didn´t organize in Cubes but they are semantical belonging to characters used in InfoCube A. If I pack them in Cube A
may be I destroy that aggregatio level I want to have ( because now all 24 characters must be identical for case of
aggregation ) . But if I organize them in seperate ICs may be I get to much ICs.
How would you solve that modelling ? Would you for example pack all characters into one IC and solve necessary aggregation level with aggregates , because they allow to use a subset of characters in the underlaying cube ?
Thanks in advance
Regards,
Oskar

Infocube design - Navigational attribute

Hi Gurus,
0ASSET is my characteristics and ZLOCCODE is a navigational attribute of 0ASSET.
In infocube , the ZLOCCODE is defined in dimension and as well in navigational attribute. So , gurus could you please explain why we need to define the ZLOCCODE in dimension and in navigational attribute.
What is the use of it?
How the data will differ?
Thanks in advance.

Hi Vaish,
Here we have two cases.
--> Use ZLOCCODE in cube Dimension and populate its value by 0ASSET master data
In this case ZLOCCODE value is populated from 0ASSET master data at the time of data load.
--> Use navigational attribute ZLOCCODE in Cube
In this Case ZLOCCODE value is the currect 0ASSET master information
Yes, both value can be different based on business logic.
EX.
You have loaded data in Cube on 01.01.2011
at that time for Asset 'X', ZLOCCODE value is 'A'.
So cube will store ZLOCCODE = 'A'
but on 01.08.2011
ZLOCCODE value of Asset 'X' is changed to 'B'.
in this case if you see Value in Query today then you can see Cube dimension ZLOCCODE value ='A'
while nav attribue ZLOCCODE value = 'B'.
Kindly note that not all the master data can be change, some of the master data value is always fix.
For Ex. Employee Gender never change so if you see Nav Attribute or Dimension Gender then value must be same.
Regards,
Ashish

Infocube design alternatives

Hi,
wats IC design alternatives,plz gv me any design alternative
thanks,
Akhilesh

akhilesh,
IC design alternatives ?? could you give the context for the same ?
I would presume that this would refer to possible design options for an IC considering the dimensions and their characteristics ... this would of course be based on your intial design and hence we would be able to provide a design alternative only after seeing the initial design..
Arun

Analyze InfoCube Design

Whats the importance of using it? and the suggestions that you can give...

hi michael,
the important factor of anlyzing cube is tom improve performance of loading .
1.so at anlyzing time we should decide which objects to put in which dimension(if relation is 1:1 or 1:M put in same dimension or if relation is M:M put into different dimension.)
2. if fact table size is morethan 20% of dimension table then we should go for line item dimension.
u can check in DB02
OR below link give u step by step procedure,
http://www.sdn.sap.com/irj/scn/go/portal/prtroot/docs/library/uuid/d0427b8a-7fe5-2c10-f1be-a8be71fa2c06?QuickLink=index&overridelayout=true
or u can run a program in se38-SAP_INFOCUBE_DESIGN.
HOPE IT HELPS. if it is really helpfull asign points.
regards,
naidu123

Methods used in design infocube

Hi Folks,
Is Star Schema only method used to design (represent) infocube or is there any other methods. if you have any links or notes please provide information.
Thanks in advance
Regards
RK

hi RK,
there is other method snowflake, but bw use star schema and extended star schema, you may find the detail in bw data modeling doc here
http://help.sap.com/bp_biv235/bi_en/documentation/Multi-dimensional_modeling_EN.doc
there are doc about infocubes design may useful for you...
'.. Efficient Usage of BW InfoProviders... strategy and commonly known multi dimensional modelling concepts when designing single SAP BW InfoCubes....
https://www.sdn.sap.com/irj/servlet/prt/portal/prtroot/docs/library/uuid/79f6d190-0201-0010-ec8b-810a969028ec
hope this helps.

What is best method for designing InfoCube?

Hi BW Guru`s,
I have some doubts on InfoCube designing.
I have a scenario like, I have 25 Characteristics ( Customer, Sales Org, Brand, Material, Material Group etc..) and 10 keyfigures (Sales, Revenue, Amount,etc..). By using these objects how to design the Infocube. But I know the normal design process. Does any body tell me best design method of Infocube and it should not effect any kind of performance issues ( load performance and Query performance).
good points will assiagn.
Thanks in advance,
venkat

hi,
I have a scenario like, I have 25 Characteristics ( Customer, Sales Org, Brand, Material, Material Group etc..) and 10 keyfigures (Sales, Revenue, Amount,etc..). By using these objects how to design the Infocube. But I know the normal design process. Does any body tell me best design method of Infocube and it should not effect any kind of performance issues ( load performance and Query performance).
As you dont have much charecteristics and key figures to design your cube, create Dimensions for CUSTOMER, SALES DATA, MATERIAL, BRAND etc. and put the logically related charecteristics into the dimensions.
For example: Customer Dimension( Customer ID, Customer name, customer Address etc.)
try to create the smaller Dimensions for performance reasons.
assigning points is the way to say thanks in sdn
regards
vadlamudi

Which on is better for reporting ODS/INFOCUBES

Hello expert's
1.I want to know which one ODS or INFOCUBES are better (technically/functionally) for reporting purpose in BW.if some one is having any comparisons on ODS/INFOCUBE then plz pass on to me.
2.Are Z tables reliable in case of data updation and fetching when we create ODS/INFOCUBE.
3.HOW RELIABLE THE "Z"(*TABLES/REPORTS) are in SAP-BW/R/3-ABAP.

Gaurav,
1. Cubes are definitely better for repiorting if cubes are used as cubes which would mean :
a. Cubes should have aggregated data and nto line item data
b. Cubes should not be used to store master data
c. Cubes with large amounts of data are regularly compressed and statistics and relevant indexes maintained
d. You can check the infocube design by using RS_INFOCUBE_DESIGNS ( I do not remember the program name properly )
ODs on the other hand :
1. Should be used for storing detail ( line item ) level data
They need not always be used for reporting
2.As for your other questions , please define reliabilty , there are many definitions for reliability and if you can elaborate on reliability it would make it easier for a directed reply.
Arun
Assign points if helpful

Different table used in infocube dataloading?

hii
   we load data in to the cube by using star schema and extended star schema
here differnt data we have to load
here different ids like dimid and sid also generated
now my question is
1. which are the different table where the record and ids are stored?
2. how and when the dimid and sid are generated?
3. can we see the dim id and sid?if we can then where?
4. this ids are stored in which tables?
thanxxx

Hi Deepa,
Whenever you want to analyze the star schema of a particular cube, execute LISTSCHEMA transaction in your system. Once you execute this transaction you will see the list view of star schema and you will get to know all the underlying tables in which cube stores the data and reads data.
which are the different table where the record and ids are stored?
In case of infocube mainly Fact table and Dimension tables are used to store the transaction specific data.
2. how and when the dimid and sid are generated?
In the infocube design you can see the various dimensions and underlying characteristics in each dimension. During data loading one DIMID is created for each unique combination of all the characteristic values. Dimension table stores the SID's of each characteristic. For e.g. Your dimension consists of Material and customer then it will have different DIMID's as follows,
DIMID          MATERIAL          CUSTOMER
1                  MAT1                  CUST1
2                  MAT1                  CUST2
3                   MAT2                 CUST1                 and so on.... In reality you will see SID's in DIM table.
3. can we see the dim id and sid?if we can then where?
You will have to check data in SID and DIM tables, use LISTSCHEMA transaction to get the table names.
4. this ids are stored in which tables?
SID and DIM tables.
Regards,
Durgesh.

Creating Infocube

Hi,
I am trying to create infocube in SAP BI 7.0 but facing a lot of problem.As I donu2019t have screen shot to
refer. I am referring SAP BW composed by fufu but it is mainly focused on sap bw old version.
Can any one send me all the steps along with screen shot. I am looking for all steps
Business Scenario and SAP BW
Creating an InfoCube
Loading Data into the InfoCube
Checking Data Quality
Creating Queries and Workbooks
Managing User Authorization
InfoCube Design
Aggregates and Multi-Cubes
Operational Data Store (ODS
Business Content
Generic R/3 Data Extraction
Data Maintenance
Performance Tuning
Object Transport
BW Implementation Methodology
Hoping prompt response from members of this group
Regards,
Vivek

Hi,
Creating infocube in 7.0 is almost similar the only difference in 3.5 u have to create update rule & transfer rule in 7.0 DTP is used for transformation.
if u mention the errors u faced while creating infocube then it is easy to find out the solution.
Pls go thru this links,
http://help.sap.com/saphelp_nw04/helpdata/en/cd/cae63b60c24c6ee10000000a11402f/frameset.htm
http://help.sap.com/saphelp_nw04/helpdata/en/cd/cae63b60c24c6ee10000000a11402f/frameset.htm
Hope it helps u,
Thanks & Regards,
SD

InfoCube Design

Similar Messages

Maybe you are looking for