Ganglia is a graphing and monitoring solution for large clusters. It can aggregate information in convenient overview displays.
The word ganglia is the plural form of ganglion, which is a nerve cell cluster in anatomy. The analogy implies that Ganglia can be the sensory nerve cell network in your cluster.
Like Munin, Ganglia also uses RRDs for database storage and graphing, so the graphs will look similar to the previous Munin graphs. Code reuse is a good thing!
Ganglia's start page shows an overview of the available data in graph format. You can drill down in the graphs to get other perspectives:
If you drill down one of the application cluster graphs for CPU load, for example, you get a detailed view of the individual servers in the cluster. You can click on a server node to drill down further.
If you have a large cluster, you can clearly see the attraction of...