Azure Analysis Services Memory Management

[ThreadRider2000]

Hi Team,

I am working on an Azure Analysis Services project and noticed that after clear and full rebuild, if I pause the AAS instance and restart it, I can reclaim a good portion of memory dropping almost by half from ~60 GBs to ~30 GBs without impacting user performance or data. 

I have found little information as why that is and whether that is a recommended practice.

Any information, white papers or links about this would be greatly appreciated.

Thanks,

Kelly

[#2]

Hi Kelly,

When you clear and rebuild an Azure Analysis Services instance, all of the data is loaded into memory. The amount of memory used by the instance will depend on the size of the data, the complexity of the model, and the number of queries being executed.

Pausing the instance releases all of the memory used by the instance, so it is not surprising that restarting the instance after a clear and full rebuild results in a lower memory usage. However, it is not generally recommended to pause and restart an Azure Analysis Services instance as it can impact user performance and data.

Instead, you can consider using the scale-up or scale-out options in Azure Analysis Services to manage memory usage. Scale-up allows you to increase the resources allocated to the instance, while scale-out allows you to add more instances to the server. Additionally, you can monitor memory usage using Azure Monitor to determine if additional resources are needed.

I hope this helps! Let me know if you have any further questions.

Kind regards,

Luke Madden