Dashboards: Refresh "Auto" setting leads to 100% CPU Utilization on MSSQL Server

### What happened? I have got some large dashboards containing some costly quarries (nothing too crazy though). I set the refresh intervals to 15min (dashboard is used by about 5 clients at all times). If the dashboard/tab sets its refresh intervall to "Auto", sometimes after a refresh of the browser tab, and keeps running for some time, the dashboard floods my MSSQL Server with a lot of the "expensive" quarries. This results in 100% CPU Utilization. All together acting like a "death-spiral". High usage -> longer response -> more quarries in queue. ### What did you expect to happen? I would expect the dashboard to not quarry new data every 2-5s, instead at a max rate of 1x all Panels every 15min (configured in dashboard settings). ### Did this work before? As the dashboards grew, it got worse. ### How do we reproduce it? 1. Create a dashboard with a somewhat expensive quarry. 2. Set the intervall to (only) 15min in dashboard settings. 3. Set the refresh interval to "Auto" 4. Leave the dashboard running for some days. ### Is the bug inside a dashboard panel? No, its the whole dashboard that causes the problem. ### Environment (with versions)? Grafana: Grafana v12.0.2+security-01 (76ea754dbb) OS: Windows 10/11 Browser: Chrome (latest) ### Grafana platform? Other (Windows Server) ### Datasource(s)? MSSQL