I have recently started developing a simple Monitoring server, using Zookeeper as the main distributed controller.
Over the next few weeks I will be describing how the system hangs together and can be used to generate arbitrary alerts (with a plugin architecture) when a server fails.
“babysitter“ comprises of:
- a Python client library to register and update server state;
- a Java Monitor Server;
- a Java (plugin) Alert Manager;
- a simple REST API to manage the server.