Most of the configuration, when building a cluster, is similar to what is done on a single machine. Setting the IP address and changing some startup environment are basically the same on all Linux machines. No specialized knowledge is required. This is different for large installations that implement a cluster file system. The software glue between all the cluster nodes is the cluster file system. Highly optimized for the cluster or large installation, a cluster file is not contained in the standard Linux distributions. The choice for a certain cluster file system depends on the file access pattern and the costs.