Structure of a distributed system

To ensure optimum performance, all sub-servers should use their own system resources (CPU, RAM, hard disk). In addition, the master server should be connected to the sub-servers via a high-speed network. To ensure fault-free communication between the PPM servers in scenarios with a high analysis volume, the network used by the master server to communicate with the sub-servers should be uncoupled from the rest of the network.

Example: Dedicated high-speed network

ppm_scaled-system

In principle, installation of multiple sub-servers on a multi-processor system is possible. However, when designing the scaling concept, you need to make sure that the scaling effect is not reduced by shared resources. If a shared database instance is used for multiple sub-servers, each sub-server should be assigned individual data files on physically different data carriers.