Forum Discussion
Report Server Caching Ability
rwmnau wrote:
... However, viewing a report does mean that PBIRS has to load the data model into memory - I suspect that if another user loads the same report before too long, the data model may still be loaded into memory on the server. I'm not sure if PBIRS works like that, but that's the only kind of "caching" I would see happening.
PBIRS works exactly like this. After a period of inactivity (60 minutes by default) the data model behind a report will get unloaded from memory. When the next user hits that report there may be a short delay (depending on the size of the data model) while the data is streamed from the SQL Server database and loaded into memory on the internal SSAS instance hosted by PBIRS. And I think there is also a mechanism where if there is memory pressure on the PBIRS server that it may unload the least recently used model from memory.
You can alter the default 60 minute timeout in the server settings when you connect to Report Server in SSMS then right click on the server and choose the "Properties" option
I don't think PBI Server has a cache like what you're talking about - the data refresh runs on whatever schedule you've set in the report properties, and there's not another layer of caching that I'm aware of. However, viewing a report does mean that PBIRS has to load the data model into memory - I suspect that if another user loads the same report before too long, the data model may still be loaded into memory on the server. I'm not sure if PBIRS works like that, but that's the only kind of "caching" I would see happening.
The only way to allow models to stay loaded longer would be adding more memory to the server or changing the settings to give the PBIRS service additional memory.
If I'm way off here, I'll leave it to somebody else to correct me and I'll happily admit it 🙂
rwmnau wrote:
... However, viewing a report does mean that PBIRS has to load the data model into memory - I suspect that if another user loads the same report before too long, the data model may still be loaded into memory on the server. I'm not sure if PBIRS works like that, but that's the only kind of "caching" I would see happening.
PBIRS works exactly like this. After a period of inactivity (60 minutes by default) the data model behind a report will get unloaded from memory. When the next user hits that report there may be a short delay (depending on the size of the data model) while the data is streamed from the SQL Server database and loaded into memory on the internal SSAS instance hosted by PBIRS. And I think there is also a mechanism where if there is memory pressure on the PBIRS server that it may unload the least recently used model from memory.
You can alter the default 60 minute timeout in the server settings when you connect to Report Server in SSMS then right click on the server and choose the "Properties" option