Root Server is the base of the system. It manages who gets access to the system through a series of databases. To control network traffic it passes much of the responsibility to the Node Servers, but it has final control of the network.
As Projects enter the system, they will be registered and passed off to an available Node Server.
Similarly, as a Screen Saver first enters the system it will register with Root which will then pass it off to an available Node Server which will then pass it off to a Project Server balance loads and resources dynamically.
This is the primary mechanism of scalability and fault tolerance. By implementing this strategy I reduce the network load of the primary server to registering resources and fault handling. The server could be off-line for long periods and the system will still function! Additionally if it detects a Node server failure it can dynamically transfer the load to a different node while maintaining the structure of the resources the failed Node server was managing.
This page was last updated 11/20/99 09:28:12 PM