Concurrent User is one of the key performance metrics in a system. Many performance issues that can only be exposed under load are related to the increased number of concurrent users. In the context of load test, we often hear phrases such as “test the system under x concurrent users”. This is a common misunderstanding of the term. In most cases, Concurrent User is a Random Variable. We propose an analytical alternative to the classic queuing theory for estimating Concurrent User. This model helps us determine whether the simulated workload is a proper representation of the expected production scenario.

Xiaosong Lou
Senior Performance Engineer
Blackline Inc
Woodland Hills, CA

Performance Engineering and DevOps