The faker datasource uses faker.js to create random data for panels.
In a single request, the time range is divided up, and a request for random data is generated. The datasource uses moustace templates to generate data, so:
{{name.lastName}}, {{name.firstName}} {{name.suffix}}
{{address.city}}
{{database.column}}
{{random.number({"min": 1, "max": 255})}}
Will generate a table of data with time, and 3 columns: a random name, a city, a database column, and a random integer between 1 and 255.
For a chart, the following:
{{random.number({"min": 1, "max": 255})}}
{{random.number({"min": 255, "max": 512})}}
Will generate 2 series from a single request (A). If you move the second line to it's own request, you get the same dataset, but over two queries.
Feedback and comments welcome.