Online Failure Forecast For Fault Tolerant Data Stream Processing

1843 Words8 Pages
Chapter 5 Online Failure Forecast for Fault-Tolerant Data Stream Processing 5.1 Introduction 5.1.1 Historical background of Failure management in Data Stream Processing In the recent times Data Stream Management Systems (DSMSs) have been developed to support critical applications that must quickly and continuously process data as soon as it becomes available. A few of the example applications include financial stream analysis and network intrusion detection. Fault tolerance and high availability are the most important for these applications because faults can lead to losses that are quantifiable. Therefore a DSMS must be equipped with fault tolerance techniques to handle both node and network failures in order to support such applications. All of the basic techniques that help to cope with failures involve some kind of replication. Typically the state of a system’s computation is replicated onto independently failing nodes. The system should then coordinate with the replicas for accurate recovery from failures. Fault-tolerant techniques are most usually designed to tolerate up to a pre-defined number, say k, of simultaneous failures. If such methods are used, the system is then said to be k-fault tolerant. Replication and coordination have 2 general approaches. Both approaches rely on the assumption that computation can be modeled as a deterministic state machine. The implication of this assumption is that two non-faulty computations that receive the same input in the
Get Access