One of the terms used to describe server systems that are able to resist failures is hardened. While hardening is most associated with closing unused ports and other vulnerabilities, to achieve a truly hardened system usually requires the application of fault-tolerance techniques.
When one component fails, it may affect the function of another component, which then affects the functions of yet another component. This chain reaction of failing components creates a condition of cascading failure, which is not a good thing. There are several methods through which a server can be fault (fail) tolerant, meaning the server is able to withstand a component failure and stay available, achieving high availability.
