# Regarding usage of RampUsersPerSec and Throttle usage

**URL:** <https://community.gatling.io/t/regarding-usage-of-rampuserspersec-and-throttle-usage/1306>\
**Category:** Gatling (Open-Source)\
**Created:** [August 20, 2014, 1:40pm UTC](https://community.gatling.io/t/regarding-usage-of-rampuserspersec-and-throttle-usage/1306 "2014-08-20T13:40:51Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Perf\_Tester](https://avatars.discourse-cdn.com/v4/letter/p/ce73a5/32.png) [@Perf\_Tester](https://community.gatling.io/u/Perf_Tester)\
**Post date:** [August 20, 2014, 1:40pm UTC](https://community.gatling.io/t/regarding-usage-of-rampuserspersec-and-throttle-usage/1306/1 "2014-08-20T13:40:51Z")

</div>

Hello Gatling Experts,

I am looking to simulate 50 concurrent users for period of 1 hour, to perform Load Testing. Each users expected to perform transactions for around 3 minutes.

Could you please advise best way to simulate this scenario.

I have tried different combinations as listed below (one of them), does not seem this approach is correct.

setUp(

testscenario.inject(constantUsersPerSec(1) during (3 minutes)) // presume this would cause around 60 users/minute for  
or  
rampUsersPerSec(1) to (5) during (3 minutes))

.throttle(reachRps(60) in (20 seconds), holdFor(40 seconds), jumpToRps(20), holdFor(150 seconds)) ===\> This may not be needed, not sure what would be the best scenario this can be used along with RampUsers.

).protocols(httpConf)  
.maxDuration(60 minutes)

Thans,  
Srujana

---

<div class="post-metadata">

**Author:** ![Michelle\_Burrows](https://avatars.discourse-cdn.com/v4/letter/m/a6a055/32.png) [@Michelle\_Burrows](https://community.gatling.io/u/Michelle_Burrows)\
**Post date:** [August 20, 2014, 1:48pm UTC](https://community.gatling.io/t/regarding-usage-of-rampuserspersec-and-throttle-usage/1306/2 "2014-08-20T13:48:30Z")

</div>

You might want 50 users per 180 seconds = constantUsersPerSec(0.28) during(3600)?