Oliveira R., Solving Asynchronous Consensus with the Crash and Recovery of Processes. PhD Thesis, EPFL D'epartement d'Informatique, 1999 (to appear).  Home/Search   Document Not in Database   Summary   Related Articles

....distinguish between ENC and EO processes. Since there is no bound for the number of times a ENC process may crash and recover, this distinction would mean predicting the crash pattern of the process. For the Crash Recovery model we consider the 3S r class of failure detectors defined in [23]. 3S r differs from 3S in the completeness property. Any failure detector of class 3S r satisfies Eventual weak accuracy and ffl Recurrent strong completeness: Every bad process is infinitely often suspected by every good process. As with 3S failure detectors, completeness can be realized by ....

....on the use processes make of stable storage. The algorithms of Oliveira, Guerraoui and Schiper [22] and Hurfin, Mostefaoui and Raynal [20] were designed using failure detectors satisfying strong completeness and can be proved correct with failure detectors satisfying Recurrent strong completeness [23]. The algorithm of Aguilera, Chen and Toueg [1] uses a hybrid failure detector which satisfies strong completeness regarding EC processes and handles the detection of EO processes by providing an estimate count of the number of recoveries of all processes. With regards to stable storage, these ....

Oliveira R., Solving Asynchronous Consensus with the Crash and Recovery of Processes. PhD Thesis, EPFL D'epartement d'Informatique, 1999 (to appear).

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC