How to run 3000 Thread from GitHub

[mukul miglani]

Hi everyone,

I've encountered some scenarios where I need to run 3000 users from Github. Is it possible to do so from a single client machine, or do we require multiple clients?

Thanks & Regards

Mukul Miglani 

[DT]

It depends on the nature of your test and parameters of your Github Actions runner.

No one can give you the answer, it's you who needs to measure it.

Normally process looks as follows:

1. Set up monitoring of the JMeter load generator resources consumption (CPU, RAM, Network and Disk IO, etc). If you don't have better options you can use JMeter PerfMon Plugin for this
   
2. Make sure to follow JMeter Best Practices
   
3. Start with 1 virtual user and gradually increase the load until any of monitored metrics from step 1 starts exceeding i.e. 85% of total available capacity. Mention how many threads were active at this stage. This is how many users you can simulate from particular this machine for particular this test. Basing on the information you get you decide whether to continue with standalone mode or go for distributed testing
   

[mukul miglani]

Thanks for the quick response.

Can you please guide me how we can do distributed load tests on github.

--
You received this message because you are subscribed to the Google Groups "JMeter Forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ptgram24+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ptgram24/bd171f34-aac7-48e5-bfdd-688849e25223n%40googlegroups.com.

--

Thanks & Regards

Mukul Miglani || 8504978530

[DT]

The same way as you run it anywhere else:

1. Get the desired number of load generator machines somehow
   
2. Install JMeter on them
3. Make sure that load generator machines can communicate with Github Actions runner where you will be executing JMeter master in terms of networking and RMI configurations (machines are in the same network and necessary ports are open), Github actions runner should be able to connect to the slaves to send the .jmx test script and test started event and the slaves need to be able to connect to the master machine to send back results and inform that test execution ended.
   

More information: How to Perform Distributed Testing in JMeter