Scalable and fault tolerant failure detection and consensus

Future extreme-scale high-performance computing systems will be required to work under frequent component failures. The MPI Forum's User Level Failure Mitigation proposal has introduced an operation, MPI_Comm_shrink, to synchronize the alive processes on the list of failed processes, so that applications can continue to execute even in the presence of failures by adopting algorithm-based fault tolerance techniques. This MPI_Comm_shrink operation requires a fault tolerant failure detection and consensus algorithm. This paper presents and compares two novel failure detection and consensus algorithms. The proposed algorithms are based on Gossip protocols and are inherently fault-tolerant and scalable. The proposed algorithms were implemented and tested using the Extreme-scale Simulator. The results show that in both algorithms the number of Gossip cycles to achieve global consensus scales logarithmically with system size. The second algorithm also shows better scalability in terms of memory and network bandwidth usage and a perfect synchronization in achieving global consensus.

Timing and nature of the penultimate deglaciation in a high alpine stalagmite from Switzerland

The timing and nature of the penultimate deglaciation, also known as Termination II (T-II), is subject of controversial discussions due to the scarcity of precisely-dated palaeoclimate records. Here we present a new precisely-dated and highly-resolved multi-proxy stalagmite record covering T-II from the high alpine Schafsloch Cave in Switzerland, an area where climate is governed by the North Atlantic. The inception of stalagmite growth at 137.4 ± 1.4 kyr before present (BP) indicates the presence of drip water and cave air temperatures of above 0°C, and is related to a climate-induced change in the thermal state (from cold- to warm-based) of the glacier above the cave. The cessation of stalagmite growth between 133.1 ± 0.7 and 131.9 ± 0.6 kyr BP is most likely related to distinct drop in temperature associated with Heinrich stadial 11. The resumption of stalagmite growth at 131.9 ± 0.6 kyr BP is accompanied by an abrupt increase in temperature and precipitation as indicated by distinct shifts in the oxygen and carbon isotopic composition as well as in trace element concentrations. The mid-point of T-II is around 131.8 ± 0.6 kyr BP in the Schafsloch Cave record is significantly earlier compared to the age of 129.1 ± 0.1 kyr BP in the Sanbao Cave record from China. The different ages between both records can be best explained by the competing effects of insolation and glacial boundary forcing on seasonality and snow cover extent in Eurasia.

Mobile sink based fault diagnosis scheme for wireless sensor networks

Network diagnosis in Wireless Sensor Networks (WSNs) is a difficult task due to their improvisational nature, invisibility of internal running status, and particularly since the network structure can frequently change due to link failure. To solve this problem, we propose a Mobile Sink (MS) based distributed fault diagnosis algorithm for WSNs. An MS, or mobile fault detector is usually a mobile robot or vehicle equipped with a wireless transceiver that performs the task of a mobile base station while also diagnosing the hardware and software status of deployed network sensors. Our MS mobile fault detector moves through the network area polling each static sensor node to diagnose the hardware and software status of nearby sensor nodes using only single hop communication. Therefore, the fault detection accuracy and functionality of the network is significantly increased. In order to maintain an excellent Quality of Service (QoS), we employ an optimal fault diagnosis tour planning algorithm. In addition to saving energy and time, the tour planning algorithm excludes faulty sensor nodes from the next diagnosis tour. We demonstrate the effectiveness of the proposed algorithms through simulation and real life experimental results.

Fault Coverage-Driven Incremental Test Generation

In this paper, we consider a classical problem of complete test generation for deterministic finite-state machines (FSMs) in a more general setting. The first generalization is that the number of states in implementation FSMs can even be smaller than that of the specification FSM. Previous work deals only with the case when the implementation FSMs are allowed to have the same number of states as the specification FSM. This generalization provides more options to the test designer: when traditional methods trigger a test explosion for large specification machines, tests with a lower, but yet guaranteed, fault coverage can still be generated. The second generalization is that tests can be generated starting with a user-defined test suite, by incrementally extending it until the desired fault coverage is achieved. Solving the generalized test derivation problem, we formulate sufficient conditions for test suite completeness weaker than the existing ones and use them to elaborate an algorithm that can be used both for extending user-defined test suites to achieve the desired fault coverage and for test generation. We present the experimental results that indicate that the proposed algorithm allows obtaining a trade-off between the length and fault coverage of test suites.

Transparent Fault Tolerance for Web Services based Architectures

Service-based architectures enable the development of new classes of Grid and distributed applications. One of the main capabilities provided by such systems is the dynamic and flexible integration of services, according to which services are allowed to be a part of more than one distributed system and simultaneously serve different applications. This increased flexibility in system composition makes it difficult to address classical distributed system issues such as fault-tolerance. While it is relatively easy to make an individual service fault-tolerant, improving fault-tolerance of services collaborating in multiple application scenarios is a challenging task. In this paper, we look at the issue of developing fault-tolerant service-based distributed systems, and propose an infrastructure to implement fault tolerance capabilities transparent to services.

Alpine Ski Areas in Maine

http://digitalcommons.colby.edu/atlasofmaine2009/1018/thumbnail.jpg

Fault Lines and Earthquakes of Maine 1568-2005

http://digitalcommons.colby.edu/atlasofmaine2008/1006/thumbnail.jpg

EFFICIENT FAULT LOCATION IN UNDERGROUND DISTRIBUTION SYSTEMS THROUGH OPTIMUM-PATH FOREST

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Fault Current Limiter Using YBCO Coated Conductor-The Limiting Factor and Its Recovery Time

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Test of a Modular Fault Current Limiter for 220 V Line Using YBCO Coated Conductor Tapes With Shunt Protection

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

A New Approach for Real Time Fault Diagnosis in Induction Motors Based on Vibration Measurement

Condition monitoring is used to increase machinery availability and machinery performance, reducing consequential damage, increasing machine life, reducing spare parts inventories, and reducing breakdown maintenance. An efficient real time vibration measurement and analysis instruments is capable of providing warning and predicting faults at early stages. In this paper, a new methodology for the implementation of vibration measurement and analysis instruments in real time based on circuit architecture mapped from a MATLAB/Simulink model is presented. In this study, signal processing applications such as FIR filters and fast Fourier transform are treated as systems, which are implemented in hardware using a system generator toolbox, which translates a Simulink model in a hardware description language - HDL for FPGA implementations.

Robust fault diagnosis in power distribution systems based on fuzzy ARTMAP neural network-aided evidence theory

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)