docs-v2/content/flux/v0.x/stdlib/testing/benchmark.md

title	description	menu	aliases	weight	flux/v0.x/tags	introduced
testing.benchmark() function	The `testing.benchmark()` function executes a test case without comparing test output with the expected test output. This lets you accurately benchmark a test case without the added overhead of comparing test output that occurs in [`testing.run()`](/flux/v0.x/stdlib/testing/run/).	flux_0_x_ref name parent testing.benchmark testing	/influxdb/v2.0/reference/flux/stdlib/testing/benchmark/ /influxdb/cloud/reference/flux/stdlib/testing/benchmark/	301	tests transformations	0.49.0

The testing.benchmark() function executes a test case without comparing test output with the expected test output. This lets you accurately benchmark a test case without the added overhead of comparing test output that occurs in testing.run().

import "testing"

testing.benchmark(case: exampleTestCase)

Parameters

case

Test case to benchmark.

Examples

Define and benchmark a test case

The following script defines a test case for the sum() function and enables profilers to measure query performance.

import "testing"
import "profiler"

option profiler.enabledProfilers = ["query", "operator"]

inData = "
#datatype,string,long,string,dateTime:RFC3339,string,double
#group,false,false,true,false,true,false
#default,_result,,,,,
,result,table,_measurement,_time,_field,_value
,,0,m,2021-01-01T00:00:00Z,t,1.2
,,0,m,2021-01-02T00:00:00Z,t,1.4
,,0,m,2021-01-03T00:00:00Z,t,2.2
"

outData = "
#datatype,string,long,dateTime:RFC3339,dateTime:RFC3339,string,string,double
#group,false,false,true,true,true,true,false
#default,_result,,,,,,
,result,table,_start,_stop,_measurement,_field,_value
,,0,2021-01-01T00:00:00Z,2021-01-03T01:00:00Z,m,t,4.8
"

t_sum = (table=<-) => table
    |> range(start: 2021-01-01T00:00:00Z, stop: 2021-01-03T01:00:00Z)
    |> sum()

test _sum = () => ({input: testing.loadStorage(csv: inData), want: testing.loadMem(csv: outData), fn: t_sum})

testing.benchmark(case: _sum)