We covered quite a lot of ground in this chapter. We learned how we can distribute a load using different techniques when executing test plans. We learned how to have JMeter work in a master/node configuration. With the help of tools such as Vagrant, we made a daunting task really easy. We learned how to spin off several node machines on the same physical box (or different boxes) and use a master node to control them all from a JMeter GUI. While convenient, we saw that this method was limiting in terms of scalability. As the number of slave nodes grew, the master quickly became a bottleneck due to high I/O generated from several nodes trying to report progress to it. To overcome such restrictions and ultimately achieve infinite scalability, we learned how to run several test machines in parallel to execute our test plans. In the process, we leveraged the AWS infrastructure and saw how we can use the cloud to aid more efficient testing, thereby helping us reach our goals.
In the last...