Schedule refresh not refreshing on time

[jpnayak]

I have 130 scheduled refreshes in my report server, all data refreshes are facing some sort of lag or delay when I check the last run time. while investigating I noticed some new refreshes never start or are stuck at refreshing so I checked the report server database event table and there were millions of event records piled up. I cleaned them all and it started working, but after some days it again started piling up events. I need to know if there are any Report server data refresh limitations or any event-related settings to be set somewhere. Please guide.
This is a single server with 48GB RAM and 8VCPU VM

[josef78]

I have a production environment with about 1000 plans, I have no problem with the event table, but many other problems.

But the first and most important step is to move to scale-out deployment. I think it's absolutely nessenary for any larger environment to separate front-end servers and refresh background servers as described by v-caitlyn-mstf.

But that still doesn't solve all the problems, because there will still be issues like stuck refsresh mashups on background servers, or stuck versions of models on the front end servers. I recommend setting up an automatic restart of the PBIRS service once a day (at night).

Thanks to separate frontend/background, if the refresh often runs during the day, there will be a new problem with reloading the model from DB to frontend server after the background refresh is completed (and saved to DB). But these problems are much smaller than the problems on a single server

[v-xiaoyan-msft · Accepted solution]

Hi @jpnayak ,

Do you have a lot of measures in your reports that cause complex calculations and take up too much machine memory resulting in incomplete refreshes?
If you optimize them and they do not improve, consider whether it is a memory pressure.

Memory pressure can occur when reports require more memory to process and render. Schedule refresh on reports may demand a significant amount of memory on the machine. Especially for larger reports. Memory pressure can result in report failures as well as a potential crash of the report server itself.

If you are encountering memory pressure consistently, it may be worth looking at a scaled out deployment of the report server in order to spread the load of resources. You can also define that a given report server is used for data refresh with the IsDataModelRefreshService setting within rsreportserver.config. With this setting, you could define one or more servers to be the front end server to handle on demand reports, and have another set of servers to only be used for scheduled refresh.

For more you can refer to：Troubleshoot scheduled refresh in Power BI Report Server 

Hope it helps,

Community Support Team _ Caitlyn

If this post helps then please consider Accept it as the solution to help the other members find it more quickly.

[jpnayak]

@v-xiaoyan-msft thank you so much for the insight. will consider scaling up the report server if needed.

[jpnayak]

Scale-up did not help. I had to scale it out to a 2 server deployment. Now one is only used for scheduled refresh. Thank you @v-xiaoyan-msft 

[lbendlin]

If you have a Pro license you can consider raising a Pro ticket at https://powerbi.microsoft.com/en-us/support/pro/