Fault tolerance in cloud computing - survey

Cloud computing datacenter hosts hundreds of thousands of servers that coordinate users´ tasks in order to deliver highly available computing service. These servers consist of multiple memory modules, network cards, storage disks, processors etc., each of these components while capable of failing. At such a large scale, hardware component failure is the norm rather than an exception. Hardware failure can lead to performance degradation to users and can result in losses to the business. Fault tolerant is one of efficient modules that keep hardware in operational mode as much as possible. In this paper, we survey the most famous fault tolerance technique in cloud computing, and list numerous FT methods proposed by the research experts in this field.