Complex event processing is a concept and a group of technologies that realize real-time event processing. An input data stream of events are being processed by a CEP system in order to extract patterns and information about the data source. The volume of this data stream can be enormous and the need for a mechanism to ease the CEP system's workload is needed. For this, a load-shedder mechanism is deployed on the system. But in order for the load-shedder to have a clear insight of the CEP system, a monitoring mechanism must be applied, to feed the load shedder with the information needed. But this monitoring solution must not impose additional overhead to the already overloaded system. In this thesis, multiple monitoring strategies are being assessed, by carrying out experiments using a monitoring mechanism with different configurations for sampling, deviation threshold, and scope and slide time frames for sampling, with the help of two projects, i.e. CoDaEFC and Precept. The goal of these experiments is to analyze the monitoring solutions' approach. A mechanism's goal should be to give insight to the system, yet without adding too much workload on the system. The result metrics of the experiments are then being logged and presented in a visual form. Conclusions are being drawn, based on the results of the experiments and how these configurations affect the CEP system's performance.