Skip to content
hyperthunk edited this page Jun 14, 2011 · 5 revisions

Nodewatch is designed to provide a practical degree of segregation between end-user modules, referred to hereafter as plugins, and core system functionality. This separation is applied by keeping the following concerns internal to Nodewatch:

This is an obvious one - Nodewatch manages process supervision internally. User defined plugins and other extensibility mechanisms can participate in supervision trees, by exporting the relevant process configuration.

Remote code loading

Nodewatch has its own mechanism for this, which is not (currently) customisable.

Scatter/Gather

Nodewatch deals with gathering data from sensors and publishing to relevant sink-interfaces. Nodewatch also provides APIs to manage both these elements.

Authoring Plugins

The only non-functional concern that plugin authors must deal with, is the need to export the requisite configuration data upon which Nodewatch relies.

Getting Help

All Nodewatch’s internal sensors and interfaces are implemented as plugins. The source code for built in modules is therefore a perfect reference for implementors.

There is an FAQ page maintained at https://github.com/hyperthunk/nodewatch/wiki/FAQ, and any concerns, questions, bugs or other issues can be raised on https://github.com/hyperthunk/nodewatch/issues.

Clone this wiki locally