Continuous Delivery with Docker and Jenkins

Scalability testing

Scalability testing explains how latency and throughput change when we add more servers or services. The perfect characteristic would be linear, which means if we have one server and the average request-response time is 500 ms when used by 100 parallel users, then adding another server would keep the response time the same and allow us to add another 100 parallel users. In reality, it's often hard to achieve thsi because of keeping data consistency between servers.

Scalability testing should be automated and should provide the graph presenting the relationship between the number of machines and the number of concurrent users. Such data is helpful in determining the limits of the system and the point at which adding more ...