Dimension relationship

Hi,
Is there a way to define relationships in 2 BPC dimensions or any other work arround to achive below requirement?
Let say I have 2 dimensions (Facility and Products)
One product can exisit in multiple facilities, how can I link products to facilities?
User is ok to combine both dimensions to make a hierarchy products by facilities but since we cannot have duplicate members so we cannot define a product at multiple nodes.
This requirement is for input screens and reports, when user select the Facility it should show only the products that are associated to that facility.
Thanks in advance for yout expert advice.
Thanks.

Hi,
You can crate two dimension Facility and Products. In product dimension add Facility as property .
As you said one product can exist in multiple facilities, you can add multiple facility value by using comma "," in facility property in product dimension.
Hope this help...
Regards,
Shrikant

Similar Messages

Dimension relationships

I have 1 fact table (granularity of a day) and 3 dimensions (above). Each measure in the fact table has a Site, a Well and a Date.
However, a Site has MANY Wells and a Well has ONE Site.
With the following structure I can run the following queries:
- Get measures for Site by Date (Year - Month - Week - Day)
- Get measures for Well by Date (Year - Month - Week - Day)
- Get measures for Date (Year - Month - Week - Day)
However, do I require a relationship between the Site and Well dimensions in order to create a:
- drill down by Date - Site - Wells for that specific Site
- roll up Wells - Site - Date

Nope it has to be via a intermediate dimension
Yes it can handle many to one and many to many scenarios
Nope. Referenced relationship means lookup so its snowflake schema
Yes. It can still remain
Please Mark This As Answer if it solved your issue
Please Mark This As Helpful if it helps to solve your issue
Visakh
My MSDN Page
My Personal Blog
My Facebook Page

Ignoring role playing relationship

As is the case in most cubes, I have a role playing Date dimension. The dimensions are [Date Created] and [Date Snapshot]. [Date Created] is snowflaked off the Ticket dimension using the DateCreatedKey. [Date Snapshot] is related to my
fact table using the DateSnapshotKey (the day the snapshot of the fact was taken).
With this schema, I can find the number of tickets that were created in January 2014.
SELECT
[Number of Tickets] ON COLUMNS,
[Date Created].[Month Year].&[201401] ON ROWS
FROM
[MyCube]
I can also find the number of tickets that were in an Open status in January 2014
SELECT
[Number of Tickets] ON COLUMNS,
[Date Snapshot].[Month Year].&[201401] ON ROWS
FROM
(SELECT [Ticket Status].[Ticket Status].&[1]
ON COLUMNS FROM [MyCube])
Now what I need to do is the get the number of open tickets in January 2014 (regardless of when they were created) and divide that by the number of new tickets in January 2014. Here's what I tried
WITH
MEMBER OpenTickets AS
([Number of Tickets]
, [Date Snapshot].[Month Year].&[201401])
SELECT
{[Number of Tickets], OpenTickets]}
ON COLUMNS,
[Date Created].[Month Year].&[201401] ON ROWS
FROM
[MyCube]
However, what I get is the number of tickets created in January 2014 and the number of open tickets that were also created in January 2014. All other open tickets are ignored. How do I break that connection between dates so that the open tickets
doesn't look at the date created?
Stephanie

In ssas, you can use the Dimension Usage tab in Cube Designer to view and edit the dimension relationships between cube dimensions and measure groups in the cube. If you break that connection between dates, you can to to the Dimension Usage
tab to set it.
http://msdn.microsoft.com/en-us/library/ms183558.aspx

SSAS relationship help

Hi all,
I have created an SSIS package which takes forecasting info from one system, and has lookups that lookup a bunch of stuff from another system. I then dump all this info into a table as below. Users, as they are, have now said ooooo wouldn't it be great
if we could analyse all this by customer, time, parts, pricing etc. which is all stored in this table. That immediately made me think SSAS cube. Now i'm pretty comfortable in SQL and SSIS. When it comes to SSAS i'm a bit of a noob.
My attempts to turn this into a cube have been a bit in vain. Therefore considering the following relationships how could i implement this in SSAS.
Customers have their own parts and share parts with other customers. (some customers have identical products)
Even if a customer has the same product as another they will have different PRICING. All other data i.e. weights, qty's etc. can just be aggregated.
For a customer and product the price could also CHANGE. So therefore a customer's product can have multiple prices. (but 1 active price)
I was thinking of creating a TIME dimension and linking that to W/C date. Basically I just need a bit of guidance into how to break this up and create in SSAS. A fast response would be much appreciated as they need this info for financial reporting
etc. Thanks for your help!
Could someone help me with this please it's getting quite urgent...

Hi jhowe1,
According to your description, you want to create a cube for the data in your above table. Right?
In this scenario, these data should be from different tables. In database, I think you have dim tables like Customer, Product, Date. We can create
dimensions based on these tables. And you can define attribute relationship in dimension, like different customer has identical product with different price. Please refer to a link below:
Dimension Relationships
Creating a New Dimension Using the Dimension Wizard
For data like Quantities, Sales, which need to be aggregated based on dimension. We should have some fact tables in database, and we can create
measures based on these fact tables.
For creating a entire cube, please see an article below:
Create First OLAP Cube in SQL Server Analysis Services
Also you can download an
AdventureWorksDW sample database for practice based on the tutorial below:
Multidimensional Modeling (Adventure Works Tutorial)
Best Regards,
Simon Hou
TechNet Community Support

Objects are not referenced in cube data source view. Large model.

Looking at large model. 100+ measure groups, 200+ dimensions, one cube. Deployed to multiple customers in production as a part of custom developed software.
in data tools 2010 I'm looking at the cube on sql 2008, sql 2008 R2 (EE, DE).
In the cube data source diagram we observe 50+ objects that do not have any links to them or from them.
1)
Some of these I was unable to find in DMV's of the cube at all, it turns out they are present in the Data source that is in Data Source Views section.
How did these end/could end up in the cube data source model? they are not available in dimension usage window at all.
2) Others do not have any links from them or to them in cube’s data source diagram, however the dimension usage clearly shows these objects are part of relationships
both regular and many to many depending on the object in question.
We are trying to identify what objects we can ‘trim’ from the model. I’m not sure what am I looking at in this case: data tools have a trouble interpreting diagram for 200+ dimensions and
100+ measure groups (it is a large existing model) and this artifact should not be used to decide what in dimensions is obsolete and can be decommissioned.
We are looking at cube that has over 40 NxM just to one fact not counting regular relationships.
Am I looking at a feature or maybe an issue with the model itself that somehow has developed in the particular cube.
It is not something I came across before.
thanks

Hi，
According to your description, you are designing a large SQL Server Analysis Services cube, the problem is that you are unable view the relationship in the data source
view diagram, right?
In your description, you said that you can see the relationships in Dimension Usage window. In your scenario, can you see the relationship under the Tables pane when expanding
a table's Relationships folder?
General, the relationships will be defined automatic if the relationships were defined in your data source of the cube. And the dimension relationships between cube dimensions
and measure groups in the cube are defined manually in Dimension Usage. In your scenario, you can define relationship manually in the data source view, please refer to the link below to see the details.
http://technet.microsoft.com/en-us/library/ms178448.aspx
Regards,
Charlie Liao
TechNet Community Support

Text Appearing as 1-Pixel-Sized Boxes in CS2

Greetings,
I'm having a problem with text inside Photoshop CS2. I've never had this problem before so I'm not sure what's causing it or how to fix it-- obviously.
I first noticed this problem when I was doing some work in Photoshop earlier today. I was working on a 1024 x 1024 image and noticed that when I tried to add text that I couldn't see it at all. It wasn't hidden behind any layers as the text layer was at the top, the opacity was set to 100%, the layer wasn't hidden, so it should have been visible... but it wasn't. I tried increasing the font size up to the max but again I couldn't see it.
Wondering what the heck was going on I created a new image file at 190 x 190 pixels and tried to add in text. This time something appeared and I realized why I couldn't see the text on the previous piece... because the text, for some reason, is appearing as tiny 1 pixel by 1 pixel sized boxes (and the font size is set to 72). If I lower the font size to anything below 72, than the boxes simply disappear. It doesn't matter if it's custom font or the standard font, it all does the same thing.
I was doing some work yesterday in Photoshop so I'm not sure if I accidentally hit some option that's causing this and I didn't realize it; though the text was working yesterday...
I've been trying to trace back through the option panels to see if there's anything that might cause this to happen but I haven't found anything yet.
Does anyone happen to know what's going on here and how I can fix this?
Any help would be much appreciated. Thank you for your time.

> Well, I went into Image > Image Size and it does look like the resolution is set to 1 (I don't remember ever changing it, but who knows). But if I try to set the resolution to 72, the width and height jump up to 73,728 pixels... which, when I tried to have it increase to that size, made Photoshop and my computer wig out and took me an hour just to get the program to close.
For sure you had "Resample Image" checked. You should have had that unchecked. You don't want to re-sample (increase or decrease the number of pixels), you just want to change the resolution-to-dimension relationship. You want to increase the resolution without re-sampling, which will cause your dimensions (in inches or centimeters) to decrease (probably by quite a bit if your resolution was set at 1 pixel (per inch?).
All this works better if you sort of think things through beforehand. If you know you are producing an 8.5"x11" sheet and you know that you will be printing it out at about 300 pixels per inch, then you create your canvas with those dimensions and resolution or, if you are using an image instead of a blank canvas you crop and Image Size the image to that dimension and resolution. Then, when you add your text, the font size in points will be roughly what you would expect from a word processing application. 10-12 pt text will come out similar to a normal "typewritten" text document, for example.

Errors in OLAP storage engine when processing application

Hi
After changing a logic in the application, we processed the application but it did not complete successfully due to the following error:
Error message:: CreateOLAPCubeForApplication:CreateCube:Errors in the OLAP storage engine: The attribute key cannot be found when processing: Table: 'dbo_tblFactQuality', Column: 'QACCOUNT', Value: '%ACCOUNTS%'. The attribute is 'QAccount_ID1'.
When trying other applications they showed errors with reference to the measure group, e.g.:
Error message:: Errors in the metadata manager. No dimension relationships exist within the 'Ownership' measure group.
Thanks for your help.
Melanie

Hi,
This kind of error will come when you have some invalid member defined in your fact table. you can use the following SQL query to check it.
select * from tblfact"yourapplication" where "dimension" not in (select id from mbr"dimension" where calc = 'N')
The same query needs to be run in wbtable and fac2 table.
These selection should return 0 records.
If it is returning something then you have to delete these records (replace "select * "with delete).
You can run this for all the applications and all the dimensions. However, I believe, you can check in your Quality application and the QACCOUNT dimension.
You can see for which dimension, there is an invalid member. This member might be getting created through one of your script logics.
Hope this helps.

SSAS Cube Total Amount showing correctly but monthly split up is wrong

Dear All,
I have created two dimensions "Detail Fields" and "Time Period" from named query.Report Amount is the measure group.
Our report is generated By Dropping the Prime Supplier in row fields,
Report Year in Column Fields with Report Amount in the Data Region. Please see the below output.
Report Year
2011 2012 2013
2014 Grand Total
Prime Supplier Report Amount Report Amount Report Amount Report Amount Report Amount
Test01 $1500.00 $2500.00 $1520.50 $2540.00
$10,000.00
Test02 $2980.00 $1240.00 $5210.50 $5410.00
$15,000.00
The issue here is the Grand Total for every supplier is fetching correctly (i.e $10,000 and $15,0000 is correct for each supplier), but the split up for every year is incorrect (i.e 1500,2500,1520,2540 for the Supplier Test01) . I don't understand why its
showing the wrong split up. May be Using of complex named query or any relationship to be created if we are using named query or any other settings or any other reason.
Also if the record count is less we are not facing any issue, the amount is splitting perfectly for every year, but in production there we have millions of records hence the amount is splitting wrongly or the problem with the complex
named query if needed i can paste the query used.

Okay guys, let me share what went wrong. The problem is with cube-dimension relationship. After fixing it the value showing correctly.

Business days in month

Hello,
I need to create a new report with some external data, from a formula or table.
This report is just a Total Amount value multiplied by the current business day in the total business days in the current month. This is just to preview the total amount expected in the end of the month. For example:
The current sales ammount is 100, and today is the business day 15. The month have 22 business days so: 100 * (22/15) = 146,66. This is just an a approximate value in the end of the month.
The problem is: How can I get in a column the value of the business day of today?, and, how can I get the total business days in the current month?.
I've tried to get those values froma a sql function (actually this is working) but i can't use this inside Oracle BI.
Also, I've tried to make a table with all the values of this columns. Date - Business day in month - Total business days in month. Then import at the repository with date dimension relationship, but in BI it gets the date of the last Sale (maybe the last sale was 2 days ago, and the relationship with my table was with Date, not today.
It's hard to explain, but essentialy I need to get just to values: Today Business Day and Total business day. I can generate this values with any other tecnology (Oracle, any language, table...) but unable to import or use it inside oracle BI.
Thanks.

Since you raised this under a E-business suite subforum, I assume you have EBS.
Normally in Oracle EBS, each organization is associated with a calendar and the calendar specifies which day is a working day and which is not.
That data is stored in bom_calendar_dates.
Sandeep Gandhi

Connect a Date Dimension to a cube without relationship

Hi everybody,
I would like to answers to one business requirements.
I create a cube that models the following event : a customer send a product from an agency to another customer who receives it in another agency.
So I have a fact table with only two measures
Amount
Count
which is connected to these dimensions
Product
Sending Date
Receiving Date
Sender (Customer)
Receiver (Customer)
Sender (Agency)
Receiver (Agency)
The users would like to analyse the following KPI, at a specific date :
Number of transactions sent, the amount
Number of transactions received, the amount
Number of transactions pending, the amount
To answer this business requirement, I have added a new date dimension in the cube with no relation, so that the user can select a date from this independent dimension and get the different KPIs.
But I don't get any results.
Is it a good model ? How to make it possible for the user to use the independent Date Dimension to perform analyzes of different KPI ?

Is it a good model ? How to make it possible for the user to use the independent Date Dimension to perform analyzes of different KPI ?
Hi Meal,
According to your description, you want to know if is it possible for the user to use the independent Date Dimension to perform analyzes of different KPI, right?
As per my understanding, we cannot do this without relationship between dimension tables and the independent Date table. However, we can add the relationship between the added date table and the fact table Sending Date and Receiving Date column. Please refer
to the link below to see the details.
http://msdn.microsoft.com/en-us/library/ms175427.aspx
Regards,
Charlie Liao
TechNet Community Support

Many to many relationship does not work on certain role played dimension

Hi,
I have a cube with role playing Account dimension, used once as “Account” and once as “Other Account” dimension.
I’ve added Trip dimension related to the account dimension by a bridge table AccountTrip
making a many to many relationship (many accounts can be in one trip and one account can go to many trips).
When I added the bridge measure group to the cube, I saw in the dimension usage tab that the relationship between the bridge and the Account dimension was done automatically.
I added the trip dimension with regular relationship to the bridge measure group and many to many relationship to the main fact table using the bridge group.
All compiled fine and gave correct results. Then I noticed that when slicing the fact table by the Other account dimension I get correct results,
but when I add the Trip dimension I get the results as if I was slicing by the Account dimension, not by the “Other account” dimension.
I checked the cube and noticed I did not set the relationship between the bridge to the “Other Account” dimension (role playing). I set it now (regular relationship, same as the account dimension), but still getting the same results.
Conclusion- when slicing by the new Trip (M2M) dimension and the “Other Account”
(role played) dimension I get the results of the “Account” dimension, not those of the “Other Account” dimension.
I checked the relationships of the “Other Account” dimension in the cube but it looks correctly set (to the external account of the main fact table and to account of the bridge table).
(Just to note I have two other bridges on the cube which are not related and don’t look like they need to be related, plus two other measure groups of the main fact table used for distinct count which are related appropriately).
What else should be done???
I would greatly appreciate your help!
Thanks
Namnami

The update server is down; try this temporary workaround
App Store>Purchased>Select "All"
Note: Look out for apps that have the word "Update"
http://i1224.photobucket.com/albums/ee374/Diavonex/9c256282736869f322d4b3071bbb2 a82_zps51a6f546.jpg

1:N relationships within a dimension

We are using OBIEE 10.1.3.3
I have a request from users where they want attributes within a dimension where those attributes have a 1:N relationship.
Example, Sales is dimensioned by Customer and Time.
The granularity of the Sales Fact is Day and Customer.
Now each customer can be know by different names and that attribute is known as Customer Synonym. There is a 1:N relationship between Customer and Customer Synonym. The users want to see both Customer and Customer Synonym and also query sales. When I add the Customer Synonym table within the same logical table source and run a query,the Sales numbers are doubled counted where there are more than one Synonyms. To avoid confusion we have trained users about this issue. However, I would like to figure out a better way to avoid the double counting. I cannot create a dimension out of Customer Synonym because the grain of the sales data is customer.
Is there a better design pattern within OBIEE to address this issue ?
Any suggestions would be appreciated.

Ya got your requirement,I got an idea or think its a suggestion lets give a try
If you know SQL coding and as you said they are alternative names,you can do 1 thing instead of displaying both names as different columns.You combine them in SQL it is concatenated like this
customer_name||' '||customer_synonym so result=AB Corporation Alpha Beta Corporation Barclay Corporation
so end result AB Corporation Alpha Beta Corporation Barclay Corporation 100
but this can be written in the query itself or may be as a condition you should give a try and play around.
Hope it helps you.
Cheers,
Kranthi

How to design many to many relationship in the fact and dimension

There is a problem in my project what is the subject.And i wanna know how to implement in owb.I use the warehouse builder 10. Thanks.

You may design and load whatever db model you want to.
But If you set a unique key, you may find some integrity issues. I wouldn't do a many to many relationship between facts and dimensions. This could cause you lots of headaches when users start to submit queries using this tables. You'll probably face performance issues.
Regards,
Marcos

Dimension usage relationship between dimensions

Hi,
please I have a case where an attribute of a dimension is part of the measure in a report, How can I establish a dimension usage relationship between this table and other dimension that is related as below( whereby the supposed fact table primary key is
appearing as foreign key in the dimension table). Thanks
ACCOUNT TABLE
accid -PK
EXTRAS TABLE
accid -FK

Hi 14QR1A,
According to your description, you want to build the relationship between dimension tables. Right?
In Analysis Services, it can have relationship between dimension like some snowflake schema. In BIDS, you can direct build the relationship between dimension tables in Data Source View. For example, the ProductCategory has relationship with
ProductSubCategory in AdventureWorks sample.
If you have any question, please feel free to ask.
Best Regards,
Simon Hou
TechNet Community Support

Multiple referenced relationships over the same intermediate dimension - generated processing query

Hi, I know the title is a bit confusing so let me try to explain it:
I have added 3 new dimensions do the AdventureWorks cube - RefA, RefB and RefC:
They are all related to the DimReseller table, and finally, to the FactResellerSales table. In the dimension usage chart, you can see that there are 4 referenced relationships using the Reseller dimension as the intermediate dimension. The Geography
was the existing one, and the other 3 were added.
Now, if I process a partition from the Reseller Sales measure group, the Profiler catches this generated query on the server:
SELECT [dbo_FactResellerSales].[dbo_FactResellerSalesSalesAmount0_0] AS [dbo_FactResellerSalesSalesAmount0_0]
,[dbo_FactResellerSales].[dbo_FactResellerSalesOrderQuantity0_1] AS [dbo_FactResellerSalesOrderQuantity0_1]
,[dbo_FactResellerSales].[dbo_FactResellerSalesExtendedAmount0_2] AS [dbo_FactResellerSalesExtendedAmount0_2]
,[dbo_FactResellerSales].[dbo_FactResellerSalesTaxAmt0_3] AS [dbo_FactResellerSalesTaxAmt0_3]
,[dbo_FactResellerSales].[dbo_FactResellerSalesFreight0_4] AS [dbo_FactResellerSalesFreight0_4]
,[dbo_FactResellerSales].[dbo_FactResellerSalesDiscountAmount0_5] AS [dbo_FactResellerSalesDiscountAmount0_5]
,[dbo_FactResellerSales].[dbo_FactResellerSalesUnitPrice0_6] AS [dbo_FactResellerSalesUnitPrice0_6]
,[dbo_FactResellerSales].[dbo_FactResellerSalesUnitPriceDiscountPct0_7] AS [dbo_FactResellerSalesUnitPriceDiscountPct0_7]
,[dbo_FactResellerSales].[dbo_FactResellerSalesTotalProductCost0_8] AS [dbo_FactResellerSalesTotalProductCost0_8]
,[dbo_FactResellerSales].[dbo_FactResellerSalesProductStandardCost0_9] AS [dbo_FactResellerSalesProductStandardCost0_9]
,[dbo_FactResellerSales].[dbo_FactResellerSales0_10] AS [dbo_FactResellerSales0_10]
,[dbo_FactResellerSales].[dbo_FactResellerSalesEmployeeKey0_11] AS [dbo_FactResellerSalesEmployeeKey0_11]
,[dbo_FactResellerSales].[dbo_FactResellerSalesPromotionKey0_12] AS [dbo_FactResellerSalesPromotionKey0_12]
,[dbo_FactResellerSales].[dbo_FactResellerSalesSalesTerritoryKey0_13] AS [dbo_FactResellerSalesSalesTerritoryKey0_13]
,[dbo_FactResellerSales].[dbo_FactResellerSalesResellerKey0_14] AS [dbo_FactResellerSalesResellerKey0_14]
,[dbo_FactResellerSales].[dbo_FactResellerSalesProductKey0_15] AS [dbo_FactResellerSalesProductKey0_15]
,[dbo_FactResellerSales].[dbo_FactResellerSalesCurrencyKey0_16] AS [dbo_FactResellerSalesCurrencyKey0_16]
,[dbo_FactResellerSales].[dbo_FactResellerSalesOrderDateKey0_17] AS [dbo_FactResellerSalesOrderDateKey0_17]
,[dbo_FactResellerSales].[dbo_FactResellerSalesShipDateKey0_18] AS [dbo_FactResellerSalesShipDateKey0_18]
,[dbo_FactResellerSales].[dbo_FactResellerSalesDueDateKey0_19] AS [dbo_FactResellerSalesDueDateKey0_19]
,[dbo_DimReseller_6].[GeographyKey] AS [dbo_DimResellerGeographyKey5_0] -- 1
,[dbo_DimReseller_22].[RefAKey] AS [dbo_DimResellerRefAKey7_0] -- 2
,[dbo_DimReseller_23].[RefBKey] AS [dbo_DimResellerRefBKey9_0] -- 3
,[dbo_DimReseller_24].[RefCKey] AS [dbo_DimResellerRefCKey11_0] -- 4
FROM (
SELECT [SalesAmount] AS [dbo_FactResellerSalesSalesAmount0_0]
,[OrderQuantity] AS [dbo_FactResellerSalesOrderQuantity0_1]
,[ExtendedAmount] AS [dbo_FactResellerSalesExtendedAmount0_2]
,[TaxAmt] AS [dbo_FactResellerSalesTaxAmt0_3]
,[Freight] AS [dbo_FactResellerSalesFreight0_4]
,[DiscountAmount] AS [dbo_FactResellerSalesDiscountAmount0_5]
,[UnitPrice] AS [dbo_FactResellerSalesUnitPrice0_6]
,[UnitPriceDiscountPct] AS [dbo_FactResellerSalesUnitPriceDiscountPct0_7]
,[TotalProductCost] AS [dbo_FactResellerSalesTotalProductCost0_8]
,[ProductStandardCost] AS [dbo_FactResellerSalesProductStandardCost0_9]
,1 AS [dbo_FactResellerSales0_10]
,[EmployeeKey] AS [dbo_FactResellerSalesEmployeeKey0_11]
,[PromotionKey] AS [dbo_FactResellerSalesPromotionKey0_12]
,[SalesTerritoryKey] AS [dbo_FactResellerSalesSalesTerritoryKey0_13]
,[ResellerKey] AS [dbo_FactResellerSalesResellerKey0_14]
,[ProductKey] AS [dbo_FactResellerSalesProductKey0_15]
,[CurrencyKey] AS [dbo_FactResellerSalesCurrencyKey0_16]
,[OrderDateKey] AS [dbo_FactResellerSalesOrderDateKey0_17]
,[ShipDateKey] AS [dbo_FactResellerSalesShipDateKey0_18]
,[DueDateKey] AS [dbo_FactResellerSalesDueDateKey0_19]
FROM (
SELECT [dbo].[FactResellerSales].[ProductKey]
,[dbo].[FactResellerSales].[OrderDateKey]
,[dbo].[FactResellerSales].[DueDateKey]
,[dbo].[FactResellerSales].[ShipDateKey]
,[dbo].[FactResellerSales].[ResellerKey]
,[dbo].[FactResellerSales].[EmployeeKey]
,[dbo].[FactResellerSales].[PromotionKey]
,[dbo].[FactResellerSales].[CurrencyKey]
,[dbo].[FactResellerSales].[SalesTerritoryKey]
,[dbo].[FactResellerSales].[SalesOrderNumber]
,[dbo].[FactResellerSales].[SalesOrderLineNumber]
,[dbo].[FactResellerSales].[RevisionNumber]
,[dbo].[FactResellerSales].[OrderQuantity]
,[dbo].[FactResellerSales].[UnitPrice]
,[dbo].[FactResellerSales].[ExtendedAmount]
,[dbo].[FactResellerSales].[UnitPriceDiscountPct]
,[dbo].[FactResellerSales].[DiscountAmount]
,[dbo].[FactResellerSales].[ProductStandardCost]
,[dbo].[FactResellerSales].[TotalProductCost]
,[dbo].[FactResellerSales].[SalesAmount]
,[dbo].[FactResellerSales].[TaxAmt]
,[dbo].[FactResellerSales].[Freight]
,[dbo].[FactResellerSales].[CarrierTrackingNumber]
,[dbo].[FactResellerSales].[CustomerPONumber]
FROM [dbo].[FactResellerSales]
WHERE OrderDateKey >= '20080101'
AND OrderDateKey <= '20081231'
) AS [FactResellerSales]
) AS [dbo_FactResellerSales]
/*** One JOIN per Referenced Relationship ***/
,[dbo].[DimReseller] AS [dbo_DimReseller_6] -- 1
,[dbo].[DimReseller] AS [dbo_DimReseller_22] -- 2
,[dbo].[DimReseller] AS [dbo_DimReseller_23] -- 3
,[dbo].[DimReseller] AS [dbo_DimReseller_24] -- 4
WHERE (
([dbo_FactResellerSales].[dbo_FactResellerSalesResellerKey0_14] = [dbo_DimReseller_6].[ResellerKey]) -- 1
AND ([dbo_FactResellerSales].[dbo_FactResellerSalesResellerKey0_14] = [dbo_DimReseller_22].[ResellerKey]) -- 2
AND ([dbo_FactResellerSales].[dbo_FactResellerSalesResellerKey0_14] = [dbo_DimReseller_23].[ResellerKey]) -- 3
AND ([dbo_FactResellerSales].[dbo_FactResellerSalesResellerKey0_14] = [dbo_DimReseller_24].[ResellerKey]) -- 4
As you can see in the comments, the generated query contains 1 join on the referenced table
per referenced relationship. This can slow down the processing if both the Fact table and the Referenced Dimension table are big (and they usually are, otherwise the design would be different). The thing is, all 4 tables are filtered in the
WHERE clause on the same condition, which makes perfect sense. This leads to the conclusion that all 4 columns could have been read from a single join.
I was quite surprised by this kind of query generation and I can't seem to find a reason why this would be so, or how it could be avoided. Naturally, the RefAKey, RefBKey, RefCKey columns could be added to the Fact table, eliminating the need for referenced
relationships, but this would add 3*4 bytes per row to the Fact table. The Fact table in my real example contains around 3 billion rows, so adding extra bytes is painful. (Yes, it's partitioned but still ... )
The referenced relationships could also be "dematerialized" by removing the famous Materialize check in the relationship settings. Then again, this would issue a join to the referenced tables at query time, slowing down the query.
Has anyone been profiling processing queries like this one and stopped to think about it?

I don't believe it's on the Essbase side, as it uses a fairly standard API to allow string input to such things, and it's limitation is well beyond 512 chars even for a direct formula input. I instead suspect that a carriage return/line feed is being inserted by the ODBC driver as part of it's output formatting, and this is causing the API to consider the input record "completed." I could be wrong, of course, but if you examine the Memo field closely, you may find the formatting to have migrated to the Access database, and even if not, you are still using ODBC to access it on the output side, so we're back to the same issue. Perhaps it really is an issue related to the ODBC drivers in Essbase, but it's not unique to dim build, to formulas, or Essbase for that matter. Allowing a single field to output more than 512 chars in a single field is where I believe you will find the real issue. Again, we need an ODBC expert to weigh in on this. <img src="i/expressions/face-icon-small-smile.gif" border="0">

Dimension relationship

Similar Messages

Maybe you are looking for