how can send bulk request

[hemedani]

Problem to solve

I want to send a bulk HTTP request for example 1000 req and get a benchmark from it.

I want to get results like the average time for each request and hardware resource usage.

[fabricereix]

Hi @hemedani,
we don't have an option yet to run the same request several times.
For the time-being, you could generate the Hurl file with the 1000 requests.

$ for i in $(seq 0 1000);do echo GET http://google.com >> test.hurl; done

And run the file with the JSON output which contains the duration for each request.

$ cat test.hurl | hurl --json >results.json
$ cat results.json  | jq '.entries|.[]|.time'
75
28
27
...

[fabricereix]

But that's something we'd like to add in the future, especially with the possibility to run the requests concurrently.
What kind of options would you like to see?

[hemedani]

tanks @fabricereix ,
one good example is wrk.
some good options are time duration, active connection, and thread use.
and it's very nice if we have a hardware resource usage report also

[jcamiel]

Initialisation of an architecture document here => /docs/spec/runner/parallel.md

[jcamiel]

@hemedani with Hurl 5.0.0, we've added a --paralleloption and a --repeat option. Parallel run is activated by default with --test so if you have one file foo.hurl, you can run now:

$ hurl --test --repeat 1000 foo.hurl

That will runs 1000 foo.hurl in parallel (numbers of threads can be set with --jobsoption)