2. Fault tolerance techniques

In this section, we present some of the techniques used to achieve fault tolerance. After an introduction to the families of techniques, we stress the need for a certain amount of redundancy within the system to be protected. This is followed by a description of error recovery methods. We then turn to communication faults and techniques for fault tolerance at the level of the nodes or machines forming a distributed system.

Our main focus is on physical faults caused by adverse physical phenomena, as opposed to human-induced faults. Fault tolerance is then implemented by two families of complementary techniques: error processing and fault handling ...