CPU issue in LGs during TruCilent testing

LG CPU issue is common issue in TruClinet protocol.

Regarding to recent conversation with HP team and footprint calculation, I can assign 3 vusers for each LGs which have 4 CPU and 16 GB Ram each. It means if I want to test my application with 100 vusers, I need at least 33 LGs. In enterprise level assigning these numbers of LGs for testing will block resources from other team.

I will be happy, if there is any way to manage this issue?

 

  • Hello Ali,

     

    As you know the TruClient protocol is one of the more resource demanding protocols that we have and is going to require a lot of memory and CPU to work properly, this to avoid the error of 80% exceed.

     

    Based on this, unfortunatley there is no a particular solution for this kind of issue like just to change some settings on your script, as this issues are more related to the resources of your enviroment.

     

    Thank you and regards,

  • I agree with Dani's reply.

     

    I also advice using stronger HW.  4 CPU and 16 GB RAM is not strong enough. I am not sure why you can run only 3 Vusers, I guess the bottleneck is CPU as you mentioned.

     

    Consider adding wait steps in the script to relieve pressure from CPU. Some pacing time between iterations can help as well. Lastly, consider a more moderate ramp-up. It might be that during the ramp up the CPU is high and later on it will stabilize.