| Ok, the old post history doesn’t make much sense with the new move back to Wordpress, so I’ll summarize. Tinnac (This is not Nagios and Cacti) is a personal project I started to take care of issues I have with the usability of Nagios, and Cacti, integrating them together into a more unified system monitor/management platform. Note: The Google Code page is severely out of date. Well it turned into a full rewrite, starting in C++, supporting NRPE, and has since evolved a bit to a Java application with a federation of nodes, each a roughly standalone node but loosely coupled to eachother (think, a non-reliant cluster), each node handles itself, and there is one (or several if you choose) node that handles the overall babysitting and alerting. Configs are stored locally in JavaDB and synced with the babysitter node and a Mysql (or whatever DB you want to use, its JDBC man) Database for longterm statistical gathering and to update their own configs from a global setting. Nodes check-in with the parent based on a schedule, if they fail to checkin on that schedule the parent fires an alert, or if the data in the checkin is out of scope, the parent fires an alert. Nodes can be Parents or specific groups of Nodes or classes of Alerts, Multi-Parent setups can be configured to babysit eachother easily. I’ll be publishing a bit more on it as it comes together, the overall design and a fairly decent amount of the underlying code is done. Oh, and I’m using my buddy Bret’s Java server Niche for parts of the frontend, with a webservice and additional PHP frontend to interface with nodes over XML-RPC in the works. |
June 24th, 2009
Tinnac!
by Daniel | Posted in Uncategorized | No Comments » |
0 Comments »
No comments yet.
RSS feed for comments on this post. TrackBack URI