ramup | A command to backup your work on RAMDisk | Continuous Backup library

What you are trying to do will highly depend on the response times of the requests you send and the number of requests you have under a thread.

For example, with the config, you have suggested we can expect several outcomes.

100 Ramup period : 1 Loop :30

• We will have 100 Threads up at the beginning of the test
• They will loop 30 times
• You will have 3000 hits, ONLY if you have one request under the thread group
• For example, if a request has a response time of 2 seconds. It will take a minimum (30*2 = 60) seconds to complete the run if the ramp-up is 0. But the response times can be lower higher. Considering that, we can't expect to generate a load of 3000 per minute with 100 threads looping 30 times.

You can have different configs to serve you need and some solutions could have some drawbacks.

If you really need to use 100 Threads, to generate a load of 3000 hits per minute, you need to do the following.

1. You must control the throughput
2. You must control the request sending frequency ( send requests without waiting for the response

Controlling throughput is not a problem. You can simply use the Constant Throughput Timer

But the problem is controlling the request sending frequency. In this case, if you are planning to monitor the server behavior with a load of 3000 hits per minute, this method will work. If you have multiple requests in a thread. Those requests will execute in a sequence (In most of the real-world scenarios, those requests are interdependent). In looping, for a new loop to be started, the current requests in the thread should be sent.

With the above behavior, we can't achieve a thread frequency of 3000 requests per minute. So we have to for the responses of the requests. To do that, move to the Advanced tab of the HTTP request sampler. and change the response time out 1ms

Now what will happen is, once the request is sent, JMeter will wait only 1 ms for the response. And will start sending the next request in line which means we can create a frequency that can serve 3000 requests per minute. With this most probably all of your requests will get "Timed out" errors but your server will get hits. If your threads have requests which depend on the previous request, this setup will not work for you. Which is why I do not recommend it.

Note: If you want to send hits to the server with a frequency of 3000 per minute. What I suggest is to use a

• Constant Throughput Timer
• Increase the number of threads and loops to create more than 3000 requests per minute

for example, 1000 threads, 10 Rampup 20 Loops with a constant throughput timer of 3000 hits per minute will help you to generate the load you need. But if you want to stick with the 100 threads option we must reduct the response timeout time and control the request frequency