Hello!
I’m searching for a clarity about two specific QS CloudWatch metrics - DashboardLoadTime and VisualLoadTime.
My main question is - which metric represents the real wait time for the user to interact with the dashboard? Is it the DashboardLoadTime or max(VisualLoadTime) across all the visualisations on the specific sheet?
I read that DashboardLoadTime includes not only loading the visualisations but also rendering layot, parameters, filters, etc - > so it’s the slowest visual + overhead. Can the user start interacting with the dashboard before everything is ready?
The reason I’m asking is that for the max(VisualLoadTime) we have around 1.5 min → the slowest visualisation takes 1.5min, while DashboardLoadTime varies from minimum 4mins and sometimes it happened that the load time was 16 mins:
- Does this mean the user waited from 4 mins and 16 mins for the dashboard to be fully ready or he waited for 1.5min (the slowest visualisation)?
- What could be the reason why DashboardLoadTime is that long, especially this 16 mins one? What are the recomendations to improve that?
Also, I saw the case where for a specific dashboard VisualLoadTime is present, but on that same date DashboardViewLoadTime is not present, how can it be?
For the context - dashboard is embedded, SPICE dataset is used, dashboard has 2 sheets with pivot tables only + a lot of caclulated fields.
Best regards,
Kseniya