How to run benchmarks without aggregation

[Aleksey Shipilev]

On 1/19/22 12:52 PM, Toine Hartman wrote:
> I am running some JMH benchmarks and want to process and visualize the results using external tools. For that, I want to have access to the individual benchmark results. However, I am currently using multiple single shot runs, which are then aggregated to an average and error score in my CSV output.
> 
> Is there a way to have JMH output each result separately, without performing this kind of aggregation?

"-rf json" would dump JSON with "rawData".

Otherwise, you can run the JMH from Java API (every JMH Sample had a sample entry point), and 
extract raw data programmatically.

-- 
Thanks,
-Aleksey