Provide a mechanism for datasource as a variable

FR1: Datasource as a global filter
I have a usecase where I have central UI and multiple datasources for logs, metrics and traces connected to the HyperDx UI. When a dashboard is created, at this moment the datasource has to be selected at the start of creating filters/dashboard panels.

This essentially means that I cannot repurpose same dashboard for a different datasource and start from scratch. I am requesting for a feature where datasource can be a dropdown filter and then rest of the filters and dashboard panels can repopulate the data based on how datasource changes.

FR2: Repeat option based on a filter
There is another advanced usecase which I currently use from Grafana where panel/row will repeat based on the variable(in HyperDx it is called as a filter). Example, for my application I have 10 instances or hosts based cluster. I want to see a graph panel for all the 10 instances in separate panels, then I will have to manually create 10 of them. As against that if there is a setting through which if I enable repeat for a specific filter, it will create new panels accordingly.

Single selection:

Multiple/All selection:

Start by tracing the dashboard, filter, and datasource-selection flows described in the issue, including how panels currently obtain their datasource and filter values. Done means a datasource can be changed through a global filter and panels or rows can repeat for selected filter values, with both single and multiple/all selection behavior covered.