Forum Discussion
Helper IPower BI Ram Usage
Hi Experts,
Currently we using Power BI Reporting Server Via Citrix and it has 32 GB. We using Direct query and performance is very slow. So when we convert to Import, it hits RAM and its like 80% usage and also after Refreshing in Import, it will still sit in RAM Correct? So we aiming to switching back to Service becaue of very slow performance.
I am trying to understand RAM Processing in Service.
We currently having data On permises which is SSMS and we don't maintain in Cloud.
So in Power BI Desktop, let's say I connected to SQL Server (On Permises) and did reports and then published. But this with Power BI Pro License and non citrix.
So I want to please understand how RAM Processing will be done here?
And also after we refresh, where does the data sits? Is is Reporting Server RAM or Power BI Server RAM?
Please help as it is very important to underdstand.
If it is cloud, I think it will use Azure and RAM from there isn't it?
Sudhakar510 wrote:Hi Shahfaisal ,
Regards to Point 3, So in Service, after we publish and while refreshing, it uses Server Memory?
But (After Import) when user starts to use the report it won't use server memory but it does use Power BI Service's Memory? Correct? Please confirm me. I think this is the main point I am trying to understand.
If you are using Power BI Service with import mode, DURING the refresh process, the local server's resources (memory, etc.) will be used to extract the data. AFTER the refresh finishes, users connect to the model in Power BI Service (all the data sits in Power BI service) so users won't be touching the local server.
When users interact with reports/dashboards, they will be querying the model in Power BI service and are completely disconnected from the local server (meaning local server resources such as memory, CPU, won't be consumed).
4 Replies
ShahfaisalSolution Sage
1. If you are having performance issues with direct query, I suggest that you troubleshoot what is causing it. Maybe you need to optimize the queries you are using, or maybe you will need to add new indexes to your underlying tables.
2. Yes, with import option, the data will be in memory after the import is completed.
3. If you are using Power BI Service, with import option, Power BI Service's memory will be used when users are running reports, etc. However, during the refresh process, you will see the memory being utilized on the database server where your data stored.
Sudhakar510Hi Shahfaisal ,
Thank you for your promt response.
Sorry one more question
Regards to your Point 1, So when we run Import Method in Power BI Reporting Server, it stores in Memory while refreshing and after Refresh(After Import Completed).. Thanks for making me understanding this
Regards to Point 3, So in Service, after we publish and while refreshing, it uses Server Memory?
But (After Import) when user starts to use the report it won't use server memory but it does use Power BI Service's Memory? Correct? Please confirm me. I think this is the main point I am trying to understand.
Currently we use Import Process in Reporting server and it is depending on RAM size
In Reporting Server we have 32 GB Ram so while refresh and import is always stores and consuming Reporting Server RAM. This is our main problem currently.
But when we publish in Service and For End users to view the Reports in Power BI Service is super quick because it uses Power BI Service's memory(Not Consuming Reporting Server Memory) . Is that Right? (Offcourse depends upon Data Model, DAX etc)
Please help,, I think I almost understood but please confirm me or Please correct me if I am wrong.
- Shahfaisal
Sudhakar510 wrote:Hi Shahfaisal ,
Regards to Point 3, So in Service, after we publish and while refreshing, it uses Server Memory?
But (After Import) when user starts to use the report it won't use server memory but it does use Power BI Service's Memory? Correct? Please confirm me. I think this is the main point I am trying to understand.
If you are using Power BI Service with import mode, DURING the refresh process, the local server's resources (memory, etc.) will be used to extract the data. AFTER the refresh finishes, users connect to the model in Power BI Service (all the data sits in Power BI service) so users won't be touching the local server.
When users interact with reports/dashboards, they will be querying the model in Power BI service and are completely disconnected from the local server (meaning local server resources such as memory, CPU, won't be consumed).
- Sudhakar510
You are star Shahfaisal . Perfectly make sends and understood. Thanks for making me understanding this