This document is an introduction to the Scyld cluster computing concept as modified for use within the NCSU environment. Although the NCSU modifications are minor and do not impact basic cluster functionality, it is important for both the user of a cluster and for the administrator of a cluster to understand those modifications, for they imapct the way an individual (user or administrator) interacts with the cluster.
This document is not meant to be an introduction to cluster computing concepts, not is it meant to be an introduction even to Scyld cluster concepts. In order to get the most out of this document, you must be comfortable with the UNIX environment, and especially with the Red Hat Linux 6.2 environment (on which Scyld's distribution is based). It also assumes a familiarity with the concept of clustered computing, specifically with operating in an MPI-based network cluster environment.
If you are not yet familiar with Scyld's distribution or with clustered computing in general, there is a list of recommended reading material in Appendix A.
| Next >>> | ||
| Why cluster |