Cardinality in BMM layer

Dear All ,
I have a question in the rpd in OBIEE that is what difference we will get when we have given an option in the BMM layer as 1 to many 0r many to 1 to 1 in the Joining tables .I would like to know that this cardinality will give any difference in the record count in the reports .
Please suggest me what is the usage of the cardinality in sql query out put .
Thanks

Hi,
From the guide "A key property of a logical join is cardinality. Cardinality expresses how rows in one table are related to rows in the table to which it is joined. A one-to-many cardinality means that for every row in the first logical dimension table, there are 0, 1, or many rows in the second logical table. The Administration Tool considers a table to be a logical fact table if it is at the Many end of all logical joins that connect it to other logical tables"
It really matters when you specify the driving tables if left hand side has less reocrds than right hand and makng it as driving table by menitoning the cardinality
Thanks,
Saichand

Similar Messages

Cardinality in physical and bmm layer.

Hi,
Is it not possible to define cardinality in physical layer. Do we do this only in BMM? Even in BMM why do we do it when it looks for 1:n between tables? Can anyone please expalin this cardinality in both layers?

I think your reading into it too much, think of the physical layer as merely your joins, no cardinality specified here.
in the BMM its a MUST to have a true logical starschema, you can have the same table as a source to a dimension and a fact in the BMM if you wish, as its a true star it will always be 1:N in here from your dims to facts. OLTP and snowflake sources are resolved by adding multiple tables into the Logical Table Sources (edit LTS and choose Add - you will see only tables connected to each other here).
The BI Server knows which physical tables serve logical dimensions and which serve logical facts in the physical layer based on the Logical table sources and model in the BMM layer.
can you tell us why your asking , or what your trying to achieve that your unsure about ? We can help you then.
Cheers
Alastair

Sum Aggregation Error in Physical & BMM Layer in OBIEE 11g with Essbase 11

Hi everyone,
I'm using OBIEE 11g with Essbase 11 as the data source. I'm using Sample Basic database from the Essbase as my data source. If I'm using the hierarchy for the measures (so I don't flatten the measures), and when I changed the aggregation in both physical and BMM layer from Aggregate_External to Sum, I can't create a report at all from the Answers.
Does anyone encounter the same thing? Any ideas/solution about this? Please help.
Thanks a lot!

Hi Deepak,
When I picked the "Basic - measure" alone, I got this error.
Error Codes: OPR4ONWY:U9IM8TAC:OI2DL65P
State: HY000. Code: 10058. [NQODBC] [SQL_STATE: HY000] [nQSError: 10058] A general error has occurred. [nQSError: 43113] Message returned from OBIS. [nQSError: 43119] Query Failed: [nQSError: 96002] Essbase Error: Unknown Member Basic - measure used in query (HY000)
SQL Issued: SELECT 0 s_0, "Sample Basic"."Basic"."Basic - measure" s_1 FROM "Sample Basic".
When I picked the "Gen1,Measures" alone from the measure dimension, I got this error:
Error Codes: OPR4ONWY:U9IM8TAC:OI2DL65P
State: HY000. Code: 10058. [NQODBC] [SQL_STATE: HY000] [nQSError: 10058] A general error has occurred. [nQSError: 43113] Message returned from OBIS. [nQSError: 43119] Query Failed: [nQSError: 46008] Internal error: File server\Query\Optimizer\ServiceInterfaceMgr\SIMDB\Src\SQOIMDXGeneratorGeneric.cpp, line 2610. (HY000)
SQL Issued: SELECT 0 s_0, "Sample Basic"."Measures"."Gen1,Measures" s_1, SORTKEY("Sample Basic"."Measures"."Gen1,Measures") s_2 FROM "Sample Basic"
But when I queried the dimensions one by one (only single dimension each), no error was shown.
This only happens if I use Sum in the physical and BMM layer. If I use External_Aggregation, these errors do not happen. And if I flatten the measures, these errors also do not happen.

How to determine the logical Keys of logical fact table in BMM layer?

Hello guys
I'd like to know more about how to know what columns of the logical fact table should be used as key in BMM layer. How OBIEE behaving differently when different keys are selected?
In my situation, I have 1 fact table that physically joins to 20 dimensions in physical layer. Therefore, in physical layer, this fact table has about 20 foreign keys but no primary keys..
In the logical layer, this fact table is logically joined to the same group of dimension tables. So in this case, how do I know what columns in logical fact table I should use as keys?
Please advice
Thanks

thanks for the reply...
The interesting thing I notice is that, when I start using fragmentation with the fact tables in its LTS, if not primary key is defined in the logical fact table, then I will be getting errors:
[nQSError: 10058] A general error has occurred. [nQSError: 14023] None of the fact sources for columns are compatible with the detail filter []. (HY000)
When I create the key by choosing random columns from the fact table as the logical key columns, this error goes way and the correct SQL is generated according to the frag..
Because of such behavior, I'd really like to know how logical key really works in BMM layer.

Unable to create logical dimension in bmm layer

Hi all,
I am not able to create a logical dimension hierarchy in the bmm layer of the rpd.
when I right click on the table I am not able to find the create logical dimension option in the popup window.
could you please help me out in this?
In the logical table we have data from sql statement also.
Thanks!

Create Dimensions is only available if the selected logical table is a dimension table (defined by 1:N logical joins) and no dimension has been associated with this table.
To create a dimension automatically, the Administration Tool examines the logical
table sources and the column mappings in those sources and uses the joins between physical tables
in the logical table sources to determine logical levels and level keys.
Hope you have defined the Logical join in BMM before looking for this option. If joins are not defined, this option won't come.
Refer http://download.oracle.com/docs/cd/E10415_01/doc/bi.1013/b31770.pdf
Page 133, Creating Dimensions Automatically
Hope its useful

How to create logical fact table in BMM layer ?

Hello,
I have 3 Dimension table - 2 are in one schema and last is another schema. Using this 3 dimension tables, I need to create a logical fact table.
So, my question is whether we can create this fact table by joining these 3 dimension table which are in 2 different schema s ?
Thanks

Fiaz,
you are correct. We can use tables from different subject area to create a report. However, my question was related to rpd design. Sorry, I was not very clear about the queries earlier.
Here is the whole scenario in the physical layer of the rpd
Table name      Databse name      Connection pool name      Schema name
AV          AV_PXRPAM     AVAILABILITY          CRMODDEV
OUTAGE          AV_PXRPAM     AVAILABILITY          CRMODDEV
COMPANY          PXRPAM          PXRPAM_POOL          CRMODDEV
AV and OUTAGE have the joins already. I want to make a join between COMPANY with OUTAGE. And then I want to include a column from each of above tables to the logical fact table in the BMM layer. then I want to do a star schema with the logical fact table to the above 3 tables in the BMM layer.
Thanks

Left outer join usage in obiee BMM layer

Hi All,
We lately started using left outer join in our bmm layer , to meet the requirement expectation .
problem : We have two tables Fact A , Dim B . There is an join int_id column to join them . Problem is... in table Dim B , we have 3 millions rows of Int_id and in fact table it is loaded for only data available(1 million rows).
Due to this we are dropping rows in the report ..as it is only getting data where fact.int_id = dim.int_id. So we used left outer join to solve to one particular report .(we know we can do ETl to load dummy int_id and make it null , but it is huge effort ..so we did left outer join)
These days , same kind of reports are in requirement . if i keep on doing same left outer join to meet the requirement in BMM layer ? is there any disadvantages due to this ? any performance degrade due to this ? please throw few lines on left outer join and its usage ?
Thanks,
Sri

Remodel your tables that might be a partition or any other way and follow in BI as per your new model.
Simply: Go for ETL kind of solution

How to resolve unresolved column error when we change column name in BMM Layer and removed alias in presentation layer

how to resolve unresolved column error when we change column name in BMM Layer and removed alias in presentation layer

Looks like the presentation column got Alias before your BMM changes, so in your case renaming logical column and deleting alias is not good to go.
Keep Alias

Drilldown in answers without creating dimensional hierarchy in BMM layer?

Hello All,
HELP! I have a main report like this:
title: Region EAST (prompted field- user entered: North, South.... )
heading: Departments Total Sales
rows: Meat (01) 500,000.00
Produce (02) 100,000.00
Organic Items(03) 250,000.00
Users wanted to ClLICK on either Meat, Produce or Organic Items field to get the detailed report on that particular dept. and would prefer to Click on the desired field than using department as a prompted field. :(
I did try to use the navigate function on Departments column to get to the detail report, but the results came back with ALL departments. Is it possible to drill down in answers without creating the dimensional hierarchy in BMM layer? If so, please show me how to create the link that would navigate to the detailed report when users CLICK on desired department, ex: Meat (01)? BTW, detailed report may came from different resources (data is in the detailed table...) and I still need to pass the region variable to the detailed report.
I really need a big help on this.
Thanks All in advance!

Hi Matt,
thank you for looking into this problem.
I did use the Navigate function on the 1st report, create a prompted filter on the depatment column (second report) but user can only click on the high lighted Department column header (on the first report) which navigated to the 2nd report and it displayed detail of all departments. Since users cannot click on a particular deptpartment, therefore I cannot pick up the value (01) for Meat department to pass thru the 2nd report. Is the Navigate function just works only when you click on the header column not on the data values in that column? do I need to set any other function to the department column on the 2nd report beside the filter?
thanks for your help!

Problem with BMM layer warning messages.

Hi,
I have modelled with the following tables in BMM layer.
Dimension tables: Regions,Countries,Departments, locations, Employees, jobs, job history
Fact table: HR Facts (measured columns like salary, max salary, min salary, commission)
The model is consistent with the following warnings:
Logical dimension table REGIONS has a source REGIONS that does not join to any fact source.
Logical dimension table DEPARTMENTS has a source DEPARTMENTS that does not join to any fact source.
Logical dimension table COUNTRIES has a source COUNTRIES that does not join to any fact source.
Logical dimension table LOCATIONS has a source LOCATIONS that does not join to any fact source.
and also in the answers section (Presentation services) i could not able to view data, it's giving the following error (when i am trying to display REGION NAME field values)
State: HY000. Code: 10058. [NQODBC] [SQL_STATE: HY000] [nQSError: 10058] A general error has occurred. [nQSError: 14026] Unable to navigate requested expression: REGIONS.REGION_NAME. Please fix the metadata consistency warnings. (HY000).
I could not able to trace the problem where exactly is to be fixed.
regards
chandra kanth.

Hi,
Thanks for the information. Ideally that works out without any hierarchies. But the problem with which i was facing is somewhat different.
Let me explain the problem scenario exactly:
I have an employee, region tables in physical layer. Employee table has SALARY Column. Region table dont have any
join with Employee table. In BMM layer i have created a fact table "HR Facts" and crated a logical fact column SALARY.
The logical fact table source is Employee table. I have drag dropped Region table into BMM layer and joined with HR
Facts table. While consistency check its given a warining message: Logical dimension table REGIONS has a source
REGIONS that does not join to any fact source.
What is the problem exaclty?
regards
chandra kanth.

Tables icon displaying yellow in BMM layer.

Hi all,
When i was tring to drag drop tables from physical layer to BMM layer it's showing yellow icons.....
Scenario is as follows:
tables: EMP, DEPT, BONUS, SALGRADE.
I have created complex join in Physical layer between EMP and SALGRADE (condition is "EMP$".SAL BETWEEN "SALGRADE$".LOSAL AND "SALGRADE$".HISAL). Samething i have taken into BMM layer. 4 table icons showing yellow and checked for consistency..giving errors like there is no join of logical tables.
i have created join Physical layer will it not carry to BMM layer? Do i need to create same complex join in BMM layer? After creating the complex in BMM layer will those errors come again??
Please aswer to my questions..asap.
regards
ck

Chandra kanth wrote:
Gurus,
The other two tables (DEPT, BONUS) are not required to join but i just kept it like that in BMM layer.
1. I have created complex join between EMP and SALGRADE in physical layer and i also manually created complex join in BMM layer....it did not come when i drag drop the schema from physical to BMM layer. Do we require to create manually again the complex join in BMM layer?Yes.
>
2. After creating complex join in BMM layer EMP table icon shows white others are yellow. Will complex join not suffice to make error free (Consistent) in BMM layer?
Is it require to join all tables (to maintain a star schema)?
You need to create complex joins to create a star schema in BMM layer. This is needed to create an error free BMM layer.
3. As per the requirement i dont require to join other two tables (DEPT, BONUS).
What is the purpose of these tables?
>
Please advice if i am wrong in the above cases....and also please answer to all my questions.
Thanks in advance
chandra kanth.

Arranging Hierarchy tables in BMM layer

Hi
I have a different structure of dimensions in the database.
Year->Season->Quarter->Month->Week
For every level in Hierarchy I have table i.e, Dim_Year, Dim_Season, Dim_Quarter, Dim_Month, DIm_Week
They are joined in physical layer in as per the hierarchy thats is Dim_Year is joined to DIm_Season by Year_id, Dim_Season is joined to Dim_Qaurter by Season id and so on.
In BMM layer I have to craete an Hierrarchy.
How should i keep these tables
Option1: Pull the all the tables separately
Create comples join between all the dim tables year->Season->Quarter->Month->Week
Create hierarchy from these tables
Option2:
Create a logical tables as Time
Pull all the tables into Time table with different Logical table sources.
Then create a hierarchy from these tables.
This will avoid the comples join in BMM layer
Please suggest which option will lead to less performance issue and which is ideal approach in such scenarios.
Thanks in Advance

Hi,
I am new to OBIEE but still I am suggesting one option....
Drag and drop any one of the Time table (for ex Dim_Year) from physical layer to bmm layer. create a complex join of this table with the other tables in the bmm layer. then simply drag and drop the columns from other tables (Dim_Season, Dim_Quarter, Dim_Month, DIm_Week) from the physical layer one by one on the LTS of Dim_Year.
Hope I haven't given you a funny answer.
Edited by: user9149257 on Mar 8, 2010 9:20 PM

Filtering values in BMM layer

Hi All,
can we filter values or add filter condition for logical column in the BMM layer, i could see we can add where condition only in LTS of the logical table, i could see we could use FILTER display function but it can be used only with measure, i want to filter store values which is a dimension, is there any way round we could do this without creating any extra aliases and extra logical table.
Thanks,
Sreekanth
Edited by: JUPS on Sep 6, 2010 6:21 PM

hi,
1)Create one session variable by writing a query in initialization block like
select 'order' ,order_name from order_dim where order_name like '%Abc%'
2) Variable to be row-wise-initialization
3) In report pull order_name and click on Filter-->Advanced-->convert to sql
give condition like
SRC_ORDER_LINES.ORDER_ID = (VALUEOF(NQ_SESSION.order))
Let me know if this doesn't work
Thanks,
Saichand.v

Changing the LTS mapping in BMM layer to new Physical Alias table

Experts
I need to change the source table mapping in my BMM layer to new physical tables aliases.
I have finished BMM layer design (column name changes, business friendly names etc.) before actually creating the alias tables in the Physical layer and I know I hv to run for hiding now. I expected changing the logical source mapping to the new aliases will work but it's not the case though it works partially. Since the Column names are changed in the bmm, I have to manually change the Expression, Physical Table in the LTS column mapping tab for each non-matching name logical columns.
Column mapping is intact where both the BMM columns and the new Physical alias names match.
eg: CALENDER_YEAR(BMM) ---> CALENDAR_YEAR(Physical alias) mapping is intact but
Week Day(BMM) ---> WEEK_DAY(Physical alias) need to be changed manually.
Any tips/suggestions will be appreciated.
Regards,
Tom

Hi Tom,
I too do not see any other options here other than manually modifying the individual logical columns. When there is a new logical column, we have to map it manually to the physical column explicitly because the Admin tool will not know which physical column to pick up.
Hope this helps.
Thank you,
Dhar

What is the use of Multiple source in Logical Table under BMM Layer?

what is the use of Multiple source in Logical Table under BMM Layer?
thanks in advance

Or more generally speaking:
To create a "logical table" consisting of multiple "physical tables"...hence the names. It's the backbone of the whole BI server concept and the way the rpd handles the requests.
I.e.: how would you otherwise proceed for making an analyzable table from SAP, Teradata, Oracle and a flat file with no full change-access to all the data sources?

Cardinality in BMM layer

Similar Messages

Maybe you are looking for