Monitoring with Ganglia

Chapter 9. Ganglia Case Studies

Daniel Pocock

Bernard Li

Alex Dean

Peter Phaal

The Ganglia project started out in 1999 with the aim of monitoring grid computing infrastructure: largely homogeneous clusters of similar compute nodes, typically in the academic and research community. The project founders (including Matt Massie) designed the system to be lightweight and efficient.

In this chapter, we present a range of case studies that demonstrate just how widely respected Ganglia has become—not just within the original audience, but in the wider world of industry.

The SARA case study is just one example of Ganglia at home in the environment it was designed for, albeit on the other side of the Atlantic.

The fact that Ganglia is being used to monitor 24 x 7 e-commerce enterprises is a sign of just how robust it is. Some of these include Etsy and Quantcast, both of which have shared an insight into just how Ganglia keeps their business running. The social networking trend is one of the most widely talked about revolutions in communications today, and it is no surprise to find Ganglia has had its finger in that pie, too, as it is the tool of choice at Tagged.

Stepping back from extremes of multicore CPU deployments, Ganglia has also proven itself to be truly adaptable and versatile in the face of dramatic change. It is estimated that more people will be accessing the Web from smartphones than from desktop PCs by the time you are reading this book. In this new world, CPUs spend more time sleeping ...