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.
| Tags | Networking Monitoring Clustering/Distributed Networks Logging Hardware Watchdog Text Processing Markup HTML/XHTML XML Utilities |
|---|---|
| Licenses | BSD Original GPL |
| Operating Systems | POSIX BSD FreeBSD Linux Solaris AIX HP-UX IRIX Mac OS X |
| Implementation | C Perl PHP |
Recent releases


Changes: A modular metric interface for C and Python (DSO support) and scriptable metric module support with Python. All pre-existing metrics (CPU, network, disk, memory, etc.) were converted to metric modules. New metric modules were introduced: multicpu, multidisk, and tcp_conn status. Modular front-end graph support was added. Metric groups that can be viewed or hidden as desired were implemented. Additional scaling capacity was provided for systems with memory greater than 4TB. Native metric support for Windows was improved (Built with CygWin).


Changes: Host view metric graphs now work correctly. "Show Hosts" toggled works again. A memory leak in network metrics on Linux has been fixed. A spoof memory leak on gmond has been fixed.


Changes: This release fixes a bug that caused XML port output to be truncated, fixes FreeBSD compilation errors, makes gmetad more resilient to round-robin database problems, makes "gmond -t" output more complete, updates the underlying Apache runtime library, expands the maximum size of gmetric messages, provides various minor PHP bugfixes in the Web frontend, and adds 3D pie graph effects.


Changes: New AIX metrics code. NetBSD support has been added. It is now possible to specify the PID file with gmond and gmetad. Bugfixes and backwards compatibility support for gmond. Compile fixes for MacOS Tiger. PHP fixes in the Web frontend. Various other bugfixes and code cleanups.


Changes: A serious gmond communication bug was fixed, the gmond.conf conversion code was fixed, gmond.conf now allows for include() statements, Linux 2.6.x machines now correctly collect network metrics, and some FreeBSD metric collection bugs were fixed.
A tool to summarize and search configuration data.
- All comments
Recent commentsA short Ganglia overview for sysadmins new to Linux clusters.
If somebody is interested I made a general overview of cluster monitoring topic with Ganglia as a main example (including adding your own metrics).
http://clusteradmin.blogspot.com/2008/02/cluster-monitoring.html
Quick read + links for first-timers.
Thanks,
-marek