Unlocking the Power of Attributes in OneStream: Balancing Potential with Performance
If you're currently grappling with consolidation performance issues or encountering slow-rendering Cube Views, fear not. We've got you covered with valuable insights on assessing your application and making the necessary adjustments to steer it back on the path of optimal performance. Let's embark on this voyage through the world of OneStream Attributes and discover how to harness their power while keeping your financial systems running smoothly.
Impact of IsConsolidated Property on Attribute of Entity or Scenario
Below is an example of when an Attribute is related to an Entity or Scenario, and how the IsConsolidated property affects the behavior of the data and performance.
Impact of IsConsolidated Property on Attribute of Account level Dimension
Below is an example of when an Attribute is related to an Account level Dimension (Account, UD1-8), how the IsConsolidated property affects the behavior of the data and performance.
Consideration #1 - Number of Attribute Members used in an application
As previously stated, Attributes can significantly enhance the depth and context of your financial data; however, managing them becomes a delicate balancing act. It’s essential to keep a close eye on their usage, especially when related to Entities or Scenarios.
Consider the 500-Attribute member threshold. This is a critical point to remember because once your application crosses the threshold of 500 Attribute members, you will have to be prepared for a noticeable impact on consolidation times across the entire application. Yes, you read that right – all cubes within the same application will experience a performance hit. Said another way, Cubes that are not using attributes but are part of the same application will be see a performance impact.
But why does this happen? The key lies in the way the system handles Attribute Members during the consolidation process. When attributes are involved, the system must evaluate and calculate the intersections with these additional Attribute Members. This can significantly inflate the in-memory data unit size (number of intersections being processed). The result? Significant memory overhead that can slow down consolidation times considerably.
A prudent approach to the 500-Atrribute Member threshold is to consider limiting the use of Attribute Members to 500 or fewer within your application. Yes, within an application not by cube. This threshold is a practical guideline to help maintain optimal system performance. However, it’s essential to understand that even with a reduced number of Attributes in play, you must continue monitoring performance. This ongoing assessment ensures that your system remains in peak condition and that no further degradation is occurring.
Now, you might be wondering if there are alternatives to harness the power of Attributes without suffering from consolidation slowdowns. The answer is yes, and they come in the form of XFBR strings and Dynamic Calculations.
XFBR strings offer a flexible solution for dynamically building specific lists or calculations for reporting on different data segments. They do not impact consolidation times or lead to data unit explosions. This makes them a valuable tool for handling complex reporting needs without sacrificing system performance.
Dynamic Calculations provide another avenue to perform calculations without the associated performance overhead of excessive Attribute usage. By using Dynamic Calculations strategically, you can achieve the desired results while maintaining the efficiency of your consolidation process.
In summary, while Attributes are a powerful tool in OneStream, they must be used with care, especially when they relate to Entity or Scenarios. To keep your system performing its best, consider limiting the use of Attribute Members, monitor performance diligently, and explore the alternatives like XFBR strings and Dynamic Calculations when complex reporting demands arise. Balancing functionality with performance is the key to unlocking the full potential of OneStream’s Attributes.
Consideration #2 – Attribute Members not constrained by Cube
It was mentioned in the previous section, that Attributes are not constrained by Cube, but they are considered as part of the whole application and therefore have a direct impact on consolidations and cube view retrieval times.
The impact on consolidation processing times on Cubes that are not explicitly using Attributes is because the system looks for matches between Attributes and their assigned dimensions across all Cubes, meaning that even if a Cube is not actively utilizing Attributes, these will still be triggered during the consolidation process.
The impact on cube view retrieval time is very similar to that of the consolidation process in that when you render a cube view from a Cube that is not associated with Attribute Members, the cube view will look to match the Attribute Member to its assigned dimension in the Cube; which will significantly impact the time it takes for the cube view to render and display its results.
In conclusion, Attribute Members are not confined to the Cubes that actively employ them; but rather extend to other Cubes within the same application. We stress the importance of understanding their impact on consolidation and cube view retrieval times as this will enable you to strike the right balance between functionality and performance within your OneStream application.
We understand this to be an issue for many of our partners and customers and have created an enhancement request that can be found by following the link below https://community.onestreamsoftware.com/t5/Platform/Attributes-Performance-Constraint-by-ScenarioType-and-Cube/idi-p/18837
Consideration #3 – Extensibility on Related Dim Type
When using extensibility on Cubes there will be an impact to how data aggregates to Members in the Inherited Dimension when the IsConsolidated setting is False or Conditional vs True.
The Attribute Member will need to be set to True for the data to properly aggregate to the Members in the Inherited Dimension. See illustration below:
- Extensibility on Related Member does not rollup to Parent Related Member
- Houston Drivers Mach5 Premium
- Houston Drivers Mach10 Pro
IsConsolidated = False or Conditional
IsConsolidated = True
Always remember to check the Attribute setting if your data is not aggregating properly to the Parent Member.
This issue has been logged as an enhancement and can be found by following the link below:
Consideration #4 – Non-Standard Translation and/or Overrides
Non-standard translation and overrides face implications in the presence of Attribute Members as well.
Specifically, consider the combination of Entity and Account. When the IsConsolidated setting is True, the Attribute Member behaves differently. In this case the override value is not displayed because the value is dynamically calculated.
Illustration #1
Entity and Account Attribute
IsConsolidated = True
Attribute does not display USD Override Value
Dynamic Translation
When an Account Attribute IsConsolidated setting is False or Conditional, the Attribute correctly shows the override value, but does not aggregate to the Parent Account when extensibility is used on the Source Dimension.
Illustration #2
Account Attribute
IsConsolidated = False or Conditional
Attribute does not display USD Override Value
Does not rollup to Parent Account when extensibility is used on source dimension
Lastly, when an Entity Attribute IsConsolidated setting is False or Conditional, the Attribute correctly shows the override value, but does not consolidate or aggregate to the Parent Entity.
Illustration #3
Entity Attribute
IsConsolidated = False or Conditional
Attribute does display USD Override Value
Does not rollup to Parent Entity
This has been logged as an enhancement here:
Consideration #5 – Varying Attribute Members by Time
When an Attribute changes its classifications based on Time, the only behavior is that the balance will move on a YTD basis, regardless of Entity and Scenario settings.
Illustration #1
If Hybrids moved its classification from Pro Products to Premium Products in 2022M2, OS defaults that the entire YTD balance moved, instead of having the 1,852,093 exist in Pro Products in M1, and any new activity M2 forward in Premium Products.
This has been logged as an enhancement here:
https://community.onestreamsoftware.com/t5/Platform/Attribute-Members-Periodic-vs-YTD/idi-p/12807
Clearing Stored Attribute Data
If you are currently working with Attributes and have or are experiencing performance impacts, or your data strategy has simply change, you may be wondering if there’s a way to remove data associated with their use, well there is, and this step-by-step guide will walk you through the process for effectively cleaning up this data from your OneStream application.
Step 1: Diagnostic Check
Before making any changes, it’s essential to pinpoint the exact issue and understand what needs to be addressed. Start by running a system diagnostic check. This will help you identify the specific data related to Attributes that you want to remove.
Step 2: Adjust the IsConsolidated Setting
Once the problematic data has been identified, it’s time to make some adjustments. Change the “IsConsolidated” setting to False for the relevant dimension or dimensions. This step is crucial, as it alters how the system handles consolidation and ensures that the data related to Attributes is treated differently.
Step 3: Disable the Attribute
To remove the Attribute-related data, you’ll need to disable the Attribute. Change the “Is Attribute Member” setting to False. This action effectively turns off the Attribute and prevents it from influencing data behavior within your OneStream application.
Step 4: Clear Data Units with Force Consolidate
To clear out the data related to Attributes effectively, you’ll need to use the “Force Consolidate” option. This powerful tool is your go-to for purging the data units impacted by the Attribute changes you’ve made. It ensures that any Attribute-related data are thoroughly removed from your system.
Note: It is very important to understand that the Data Management step “ClearData” will not be able to clear this specific data. “Force Consolidate” is the recommended method for this purpose.
By following these steps, you will efficiently remove all data associated with the use of Attributes in your OneStream application. Follow these steps as detailed for the successful execution of the data cleanup and return your application to its most optimal state.
Conclusion
As we have explored in this blog, Attributes are a powerful tool that can offer different avenues to enrich the financial model. They provide valuable context and flexibility to your data, but they also come with a great deal of responsibility.
We have discussed how Attributes not only impact the dimensions directly associated with them but also the broader application performance, consolidation times, and Cube View retrieval times. Understanding these implications is crucial in striking the perfect balance between functionality and performance.
Whether you are considering the use of Attributes during your implementations or if they are already part of an existing application, the insights shared here should help guide your decision-making process.
Carefully assess your organization’s needs, monitor system performance, follow best practices so you can harness the full potential of Attributes while ensuring your OneStream application runs smoothly and efficiently.
