Piranha is a collection of programs that interact with each other to provide a clustering solution. It is vital to note that cluster computing consists of two distinct branches:
Compute clustering (such as beowulf) uses multiple machines to provide greater computing power for computationally-intensive tasks. This type of clustering is not addressed by Piranha.
High Availability (or HA) clustering uses various technologies to gain an extra level of reliability for a service. HA clustering is the focal point for Piranha.
Let's take a look at some sample configurations, using both FOS and LVS as the base technology. Note that the boxes in the following diagrams (and the terms used to describe them) designate roles rather than specific systems. Also keep in mind that, due to the variety of ways in which the underlying technologies may be deployed, you will find that various terms may be used interchangably throughout this and other cluster-related documents. Although a cluster may be configured such that each role is carried out by a dedicated system, there is no technological requirement for this. However, capacity planning or system administration-related issues may dictate that dedicated systems be used for the various roles in a given cluster.
![]() | Please Note |
---|---|
Due to the variety of ways in which the underlying technologies may be deployed, you will find that various terms may be used interchangably throughout this and other documents. While every attempt has been made to define each term as it is first used, |