Forum Discussion
Advocate IIIAzure AS Migration Migration Experience from the Fields, Open Questions
Hi, jochenj
Based on the information you provided, you can refer to the following statement.
1. memory management, AzureAS uses a lot of memory to process and query data. If the memory limit is reached, you may need to consider vertically scaling the AzureAS instance or optimizing the model to use less memory. However, as you point out, memory in MS Fabric SKUs is more expensive. You may need to balance the cost of additional memory with the performance benefits it provides.
Facing memory issues with the Azure analysis services - Microsoft Q&A
Analysis Services memory properties | Microsoft Learn
2. In DirectLake mode, a "Capacity Limit" reach error was encountered. Whether or not you can create partitions in DirectLake mode in MS Fabric, you can refer to the following link.
A Comprehensive Guide to Direct Lake Semantic Model Migration for Power BI – Hitachi Solutions (hitachi-solutions.com)
3. DirectQuery mode sometimes causes a timeout. This may be due to the complexity of converting DAX to SQL. You may need to optimize your DAX queries or consider using a different query mode.
4. In ImportMode, even with small table permissions and F256-SKUs, an Out of Memory error is encountered. This indicates that your model may be too large for the available memory. You may want to consider ways to reduce the size of your model, such as removing unnecessary columns or rows, or using more efficient data types.
AAS: How can I optimize the memory usage of an Azure Analysis Services instance? - Stack Overflow
5. Regarding the cost of using MS Fabric, you may want to review the Azure Cost Optimization recommendations or consider using the Azure Cost Management tool to better understand and manage your costs.Microsoft Cost Management | Microsoft Azure
Tutorial - Reduce Azure costs with recommendations - Microsoft Cost Management | Microsoft Learn
6. you mentioned that doing a large scale re-model of an existing model to improve performance is not an option. However, there are smaller optimizations you can make that could have a significant impact on performance. For example, you might consider optimizing memory settings in Analysis ServicesOptimizing memory settings in Analysis Services - SQLBI
I hope this helps you in some way.Best Regards,
Yang
Community Support TeamIf there is any post helps, then please consider Accept it as the solution to help the other members find it more quickly.
If I misunderstand your needs or you still have problems on it, please feel free to let us know. Thanks a lot!How to get your questions answered quickly -- How to provide sample data in the Power BI Forum
Hi, jochenj
Based on the information you provided, you can refer to the following statement.
1. memory management, AzureAS uses a lot of memory to process and query data. If the memory limit is reached, you may need to consider vertically scaling the AzureAS instance or optimizing the model to use less memory. However, as you point out, memory in MS Fabric SKUs is more expensive. You may need to balance the cost of additional memory with the performance benefits it provides.
Facing memory issues with the Azure analysis services - Microsoft Q&A
Analysis Services memory properties | Microsoft Learn
2. In DirectLake mode, a "Capacity Limit" reach error was encountered. Whether or not you can create partitions in DirectLake mode in MS Fabric, you can refer to the following link.
A Comprehensive Guide to Direct Lake Semantic Model Migration for Power BI – Hitachi Solutions (hitachi-solutions.com)
3. DirectQuery mode sometimes causes a timeout. This may be due to the complexity of converting DAX to SQL. You may need to optimize your DAX queries or consider using a different query mode.
4. In ImportMode, even with small table permissions and F256-SKUs, an Out of Memory error is encountered. This indicates that your model may be too large for the available memory. You may want to consider ways to reduce the size of your model, such as removing unnecessary columns or rows, or using more efficient data types.
AAS: How can I optimize the memory usage of an Azure Analysis Services instance? - Stack Overflow
5. Regarding the cost of using MS Fabric, you may want to review the Azure Cost Optimization recommendations or consider using the Azure Cost Management tool to better understand and manage your costs.
Microsoft Cost Management | Microsoft Azure
Tutorial - Reduce Azure costs with recommendations - Microsoft Cost Management | Microsoft Learn
6. you mentioned that doing a large scale re-model of an existing model to improve performance is not an option. However, there are smaller optimizations you can make that could have a significant impact on performance. For example, you might consider optimizing memory settings in Analysis Services
Optimizing memory settings in Analysis Services - SQLBI
I hope this helps you in some way.
Best Regards,
Yang
Community Support Team
If there is any post helps, then please consider Accept it as the solution to help the other members find it more quickly.
If I misunderstand your needs or you still have problems on it, please feel free to let us know. Thanks a lot!
How to get your questions answered quickly -- How to provide sample data in the Power BI Forum