Hello everybody,
I have to make a performance test involving 500 virtual users,
My simulation setup implies to inject 250 users in 10 min, nothing for 10 min, inject the next 250 user in 10 min, nothing for 10 min. My problem is that the final report (first statistics chart) does not reflect exactly my simulation since it takes all the request together and does not partition them in accordance to the number of parallel user injected.
I have made a workaround, but since I have 50 different requests, it results to a heavy file and I suppose that the way I am doing it is not really optimal.
For the moment, I split request like that:
My User request (the 250_to_500_users_scenario is exactly the same, only the name changed)
`
val 1_to_250_users_scenario =
group(“1_to_250_users”) {
exec(http(“1_to_250_users”)
.get(“users”)
.header(HttpHeaderNames.Authorization,headerAdminAuth)
.headers(headers_12)
.check(status.is(200)))
}
`
My scenario
val scn_1_to_250 = scenario("GlobalScenario1")
.during (20 minutes){
randomSwitch(
50d -> 1_to_250_users_scenario,
50d -> 1_to_250_account_scenario
)
}
.forever{
randomSwitch(
50d -> 250_to_500_users_scenario,
50d -> 250_to_500_account_scenario
)
}
val scn_250_to_500 = scenario("GlobalScenario2")
.forever{
randomSwitch(
50d -> 250_to_500_users_scenario,
50d -> 250_to_500_account_scenario
)
}
My setup