Ganglia is a scalable distributed monitoring system for high-performance computing systems such as clusters and grids. It is based on a hierarchical design targeted at federations of clusters. Ganglia is currently in use on over 500 clusters around the world and has scaled to handle clusters with 2000 nodes.
RRDutil is a a tool to collect statistics (typically every 5 minutes) from multiple servers, store the values in RRD databases (using RRDtool), and plot out pretty graphs to a Web server on demand. The graph types shown include CPU, memory, disk (space and I/O), Apache, MySQL queries and query types, email, Web hits, and more.
Cammgr manages a collection of web cameras. It will bring cameras online or offline, and initiate or suspend image capture operations. Any camera can be controlled provided a driver is available. Cammgr supports multiple images per camera, default images for inactive cameras, per- user/host/camera notifications, and an easy-to- use configuration file. It can scan Apache log files to auto-initiate image capture, and tune capture frequency can to the minimum needed to ensure that clients receive a new image on request.
ICPLD (Internet Connection Performance Logging Daemon) is a connection monitor that sends ICMP requests to IP addresses of your choice and monitors if your machine has a working network connection. It logs failed attempts to reach the hosts, and will stamp a log as soon as a reply is received. It keeps track of when and for how long the connection was unavailable and records both total down time and each occasion of interrupted connection. It supports IPv6 and can execute a command whenever a connection goes up or down, which is useful for alerting users.
mod_backhand is a load balancing module for Apache. It provides per-request HTTP redirection within a heterogeneous Apache server cluster. Each request is processed and run through a set of "candidacy functions" to determine which server is best suited to respond. The request is then proxied to that server. Facilities are in place to allow you to write your own dynamically loadable decision making algorithms. Everything about the request and the current availability of resources can be used in the decision-making process.
skill sends signals to processes given any combination of user names, ttys, commands, and pids. snice changes the priority of processes (given the same). They are similar to kill(1) and renice(8), but the commandline is order-independent. There are also verbose, search, and interactive modes of operation.