Skip to main content

Star schemas and aggregate (or summary) fact tables


Aggregate tables can further improve query performance by reducing the number of rows over which higher-level metrics must be aggregated. 
However, the use of aggregate tables with dimension tables is not a valid physical modeling strategy. Whenever aggregation is performed over fact data, it is a general requirement that tables joined to the fact table must be at the same attribute level or at a higher level. If the auxiliary table is at a lower level, fact rows will be replicated prior to aggregation and this will result in inflated metric values (also known as "multiple counting").

With the above Time dimension table, a fact table at the level of Day functions correctly because there is exactly one row in DIM_TIME for each day. To aggregate the facts to the level of Quarter, it is valid to join the fact table to the dimension table and group by the quarter ID from the dimension table.

Sql
select DT.QUARTER_ID,
   max(DT.QUARTER_DESC) Quarter_Desc
   sum(FT.REVENUE) Revenue
from DAY_FACT_TABLE FT
   join DIM_TIME DT
      on (FT.DAY_ID = DT.DAY_ID)
group by DT.QUARTER_ID

If, however, there is an aggregate fact table already at the level of Quarter, the results will not be correct. This is because the query must join on Quarter ID, but the quarter ID is not a unique key of the dimension table. Because any given quarter of a year contains 90, 91 or 92 days, the dimension table will contain that many rows with the same quarter ID. Thus fact rows will be replicated prior to taking the sum, and the sum will be too high.

Sql
select FT.QUARTER_ID,
   max(DT.QUARTER_DESC) Quarter_Desc
   sum(FT.REVENUE) Revenue
from QTR_FACT_TABLE FT
   join DIM_TIME DT
      on (FT.QUARTER_ID = DT.QUARTER_ID)
group by FT.QUARTER_ID

This is a generally recognized problem with star schemas, and is not strictly a MicroStrategy limitation.

Star schemas will function correctly with MicroStrategy SQL Generation Engine 8.x as long as they obey the general data warehousing principle that fact tables should not be at a higher level than the dimension tables to which they are joined.

If aggregate tables are required, it is necessary to provide higher-level lookup tables with unique rows corresponding to each aggregate table's key. Logical views are a way to do this without adding tables or views to the warehouse. For example, LWV_LU_QUARTER may be defined using the following SQL statement:

Sql
select distinct QUARTER_ID, QUARTER_DESC, YEAR_ID
from DIM_TIME

 

With this logical view, it becomes possible for MicroStrategy SQL Generation Engine 8.x to query the quarter-level fact table as follows. Since the logical view has distinct rows per quarter, multiple counting will not occur in this query.

Sql
select FT.QUARTER_ID,
   max(LQ.QUARTER_DESC) Quarter_Desc
   sum(FT.REVENUE) Revenue
from QTR_FACT_TABLE FT
   join (select distinct QUARTER_ID, QUARTER_DESC, YEAR_ID
         from DIM_TIME) LQ
      on (FT.QUARTER_ID = LQ.QUARTER_ID)
group by FT.QUARTER_ID

For more information on the use of logical views in MicroStrategy SQL Generation Engine 8.1.x and 9.x, consult the MicroStrategy Project Design Guide manual, Appendix B: Logical Tables, "Creating logical tables."

Aggregate tables store pre-summarized totals at a higher level of aggregation than the most granular fact table. They allow reports to be generated from small, rather than large, tables; therefore, performance is enhanced. A successful aggregation strategy seeks to choose aggregate tables that will have the most impact while taking the least amount of space.

Aggregation decisions are driven by the following factors:

  • Usage patterns: Build aggregate tables that are likely to be used the most.
  • Compression ratios: The compression ratio between two tables is defined as the size of the aggregate compared to the size of the smallest table from which the aggregate can be derived.
  • Volatility: Changes in hierarchies over time impact the accuracy of aggregate tables. Sometimes aggregate tables must be rebuilt as a result of changes in dimensions.
A good candidate for aggregation should have at most 10-15 percent of the size of the smallest table from which it is derived.

EXAMPLE:
The MicroStrategy Tutorial project uses aggregate tables by default. A simple metric sum (Revenue) will go to different aggregate tables depending on the attributes on the template.

  1. Create a report with Year on the rows and Revenue on the columns.
  2. Execute the report and view the SQL:

  3. Drill from Year to Item and view the SQL:

The query will go from using ORDER_FACT to ORDER_DETAIL. When Year is on the template, the engine selects the smaller table and the fact is calculated as:

sum(a11.ORDER_AMT)
    instead of:

    sum((a11.QTY_SOLD * (a11.UNIT_PRICE - a11.DISCOUNT)))
    Labels:

    Comments

    1. hinaFebruary 5, 2021 at 12:19 PM

      Is there any solution now to use aggregate tables with star schema without creating logical tables?

      ReplyDelete

    Post a Comment

    Popular posts from this blog

    MicroStrategy URL API Parameters

    MicroStrategy URL Structure The following table summarizes the root URL structure used for every request to MicroStrategy Web. Environment Main Application URL Administration URL J2EE http://webserver/MicroStrategy/servlet/mstrWeb http://webserver/MicroStrategy/servlet/mstrWebAdmin .NET http://webserver/MicroStrategy/asp/Main.aspx http://webserver/MicroStrategy/asp/Admin.aspx Every request sent to MicroStrategy Web calls a central controller. Parameters are appended to  Main.aspx  or  mstrWeb  (in a .NET and J2EE environment, respectively) to indicate to the controller how the request should be internally forwarded and handled. The following examples show a URL for accessing a MicroStrategy folder when the user does not have an existing session. The URL contains not only the parameters needed to connect to MicroStrategy Web, but also the parameters needed to log on and create a session. J2EE environment: <a href="http:...
    2 comments
    Read more

    Create an alert-based subscription in MicroStrategy Distribution Services

    Create an alert-based subscription in MicroStrategy Distribution Services on Web Subscription to a report or Report Services document which will be executed when a certain conditional threshold is met based on another executing report. For example, a scheduled report executes which shows the Revenue by day for the past week. If the Revenue on any one day falls below a certain value, a subscription to another report or Report Services document can be triggered and delivered to a recipient. An alert based subscription can only be created directly on a report; however, another report or Report Services document can be delivered when the alert based subscription is triggered. Note: you need a grid report to create an alert and you cannot create if you want to create on a document with text boxes. The following example will walk through the basic steps on how to setup a subscription based on an alert like this: Follow the brief  steps bel...
    3 comments
    Read more

    Non Aggregate metrics Beginning lookup, Ending lookup, Beginning fact., Ending fact

    Non Aggregate metrics  Beginning lookup,  Ending lookup,  Beginning fact .,  Ending fact A nonaggregatable metric, such as an inventory metric, is one that should not be aggregated across an attribute.   For example, if you have monthly inventory numbers in your data warehouse and want to calculate the yearly inventory, adding the monthly numbers together does not provide a useful business measure. Instead, you may want to use the end-on-hand and beginning-on-hand inventory numbers to see how the total inventory changed during the year.  The following options are available: • To use the first value in the lookup table, select  Beginning lookup . • To use the last value in the lookup table, select  Ending lookup . • To use the first value in the fact table, select  Beginning fact . • To use the last value in the fact table, select  Ending fact .
    Post a Comment
    Read more

    Microstrategy Dossiers explained

    Microstrategy  Dossiers With the release of MicroStrategy 10.9, we’ve taken a leap forward in our dashboarding capabilities by simplifying the user experience, adding storytelling, and collaboration.MSTR has  evolved dashboards to the point that they are more than dashboards - they are  interactive, collaborative analytic stories . Ultimately, it was time to go beyond dashboards, both in concept and in name, and so  the've  renamed VI dashboards to  ‘ dossiers ’.  Dossiers can be created by using the new Desktop product or Workstation or simply from the Web interface which replaces Visual Insights. All the existing visual Insights dashboards will be converted to Dossiers   With MicroStrategy 10.9, there was an active focus on making it easier to build dashboards for the widest audience of end users. To achieve this, some key new capabilities were added that make it easier to author, read, interact and collaborate on dashboards ...
    7 comments
    Read more

    Bursting file subscriptions Microstartegy

    Bursting file subscriptions: Delivering  parts of reports across multiple files: Large MicroStrategy reports and documents are often broken up into separate pages by attributes. In a similar way, with Distribution Services, you can split up, or burst, a report or document into multiple files. When the subscription is executed, a separate file is created for each element of each attribute selected for bursting. Each file has a portion of data according to the attributes used to group data in the report (page-by axis) or document (group-by axis). Ex:, you may have a report with information for all regions. You could place Region in the page-by axis and burst the file subscription into the separate regions. This creates one report file for each region. As a second ex:, if you choose to burst your report using the Region and Category attributes, a separate file is created for each combination of Region and Category, such as Central and Books as a report, Central and Ele...
    1 comment
    Read more

    "System Prompt" and its uses in MicroStrategy

    System Prompt and its uses in MicroStrategy WHAT IS A "SYSTEM PROMPT"? "System Prompt" is a system object that was introduced back in version 8.0.0. The object is named as "User Login" and is implemented as a prompt object. The object can be found under Public Objects > Prompts > System prompts, as shown below: Unlike ordinary prompt objects, system prompts don't require any answers from the user. When a report containing a system prompt runs, the prompt is answered automatically with the login of the user who runs the report. On the other hand, like other prompt objects, answers to system prompts are used to match caches. Therefore, users don't share caches for reports that contain system prompts. For details on how caches are matched, refer to the following MicroStrategy Knowledge Base document: KB5300-7X0-0147 - How are caches matched in MicroStrategy Intelligence Server 7.x? WHEN ARE SYSTEM PROMPTS USED?    System pr...
    Post a Comment
    Read more

    Control the display of null and zero metric values

    Show   Control the display of null and zero metric values in a grid report You can determine how to display or hide rows and columns in a grid report that consist only of null or zero metric values. You can have MicroStrategy hide the rows and columns in the following ways: Hide rows and columns that consist only of null metric values Hide rows and columns that consist only of zero metric values Hide rows and columns that consist only of null or zero metric values (default) Once you have defined how MicroStrategy hides null and zero metric values in the grid, you can quickly show or hide the grid using the Hide Nulls/Zeros option in the Data menu, as described below, or by clicking the  Hide Nulls/Zeros  icon  in the Data toolbar. To determine how null and zero metric values are displayed or hidden in a grid report Open the report in Edit mode. From the  Tools  menu, select  Report Options . The Report Options...
    Post a Comment
    Read more

    Retrieve a list of user groups and the associated users in MicroStrategy Developer 9.x / 10.x

    Retrieve a list of user groups and the associated users in MicroStrategy Developer Follow the steps below to create a list of all groups and the users in each group: In MicroStrategy Developer 9.x, select 'Project Documentation' from the Tools menu to start the wizard. Select any project that is in the project source that contains the users and groups and click Next. Select only Configuration Objects for documentation. Uncheck the 'Basic Properties' object category from the next screen, as shown below: Then select only 'User Group' under the Configuration Objects section and only 'Groups' and 'Members' under the Definition section, as shown below: Go through the rest of the wizard, and open the resulting documentation. After navigating down to the User Groups, the documentation should look similar to the following image: This page shows every group, any child groups, and all members of each group.
    Post a Comment
    Read more

    Microstrategy Removing sections that do not have metric data

    Removing sections that do not have metric data This is an interesting feature which might not be explored by many of us and it comes us handy. A  cross join between datasets can result in rows or Group Header/Footer sections that do not have metric data. For example, a document contains two datasets. Dataset 1 contains Year and Revenue, with data for three years (2007-2009). Dataset 2 contains Year and Profit, filtered to return data for only two years (2008 and 2009). If you place Year and Profit in the Details and execute the document, it displays three rows, although no profit data exists for 2007. This is a product of the cross join between the two datasets. You do not want to see the blank line for 2007 since it does not give you any data for profit. You can select the  Trim sections for which no metric value data is available  check box. This removes the row for 2007, since no metric data for Profit is available for 2007. The results are shown below: ...
    Post a Comment
    Read more

    Client Rendering Optimizations for Dashboard Performance Optimizations

      The amount of data retrieved and objects being used in a Report Services Dashboard have a direct impact in the size of the final Dashboard. The bigger the Dashboard size the longer it will take to be prepared, be sent to the client, and render.   Client Rendering Once the data reaches the end user's browser window the data has to be formatted according to the definition of the Dashboard as specified in the formatting set by the architect. To do so the browser will have to either build the HTML page in DHTML mode or initialize the flash container and parse the XML.   Client rendering greatly varies depending on the hardware used. More powerful machines will render dashboard faster for a list of recommended client hardware specifications please refer to the Readme File for the specific version of MicroStrategy.   Optimization Techniques common to DHTML and Flash Client rendering time greatly relies in the amount of XML that needs to be parsed. In order to ensure that...
    Post a Comment
    Read more