Creating load for app engine testing
I'm implementing a tool to generate data and load for our app engine based application. The requirement is to post data to an endpoint. I implemented initial python version based on solution presented in Google's load testing article and I'm currently running the script on compute engine virtual machine. At this point I have been able to get ~1800 requests / second using 50 threads when posting to an empty endpoint. It seems that increasing threads beyond 50 doesn't increase request / second rate. Currently in order to create more requests / second, I'm creating more compute engine instances and running the script simultaneously from multiple virtual machines. My question is that is there a better way for this. What is the recommended or optimal way to achieve this?
Multi-user timer with Google App Engine
Certificate parse error when upload new SSL certificates in AppEngine dash
Can't Deploy Go app Error 422 to AppEngine
measuring performance golang
GAE Authentication and user data
Add Cloud SDK to Eclipse with GPE
Set Java bytecode version in Android Studio for AppEngine
How do I use environment variables in app.yaml file?
How to set version when deploying gae java?
google cloud app engine deploy failed
issue with local package imports: golang, appengine, testing
Google cloud is throwing the error saying updating service failed when I try to run mvn appengine:deploy
Google App Engine - what is the meaning of each x-appengine header?
app.yaml to send params to Dockerfile
Ioncube Loader with Google App Engine
Why is the user_id returned from Google Account API different from the user_id on Google plus and how do I access the former with oauth 2