F. Cristian and C. Fetzer. Probabilistic internal clock synchronization. In Proceedings of the 13th Symposium on Reliable Distributed Systems, pages 22--31, Dana Point, California, USA, October 1994. IEEE Computer Society Press.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....and that there is a bound f on the number of failed processors throughout the lifetime of the system. Many such works are based on local convergence functions. An early overview of this approach can be found in Schneider s report [26] A very partial list of results along this line includes [13, 7, 8, 9, 21, 2, 20]. The Network Time Protocol, designed by Mills [21] allows recoveries, but without analysis and proof. Furthermore, while authenticated versions of [21] were proposed, so far these do not attempt recovery from malicious faults. Our algorithm uses a convergence function similar to that of Fetzer ....

....for the clock values of other processors, using a fixed, simple convergence function. No rounds. As mentioned in section 1. 1, one notable differences between our protocol and other protocols that have been proposed in the literature is that many convergence function protocols (for example [8, 9]) proceed in rounds, where each processor keeps a different logical clock for each round. A round is the time between two consecutive synchronization protocols. In these protocols, if a processor is asked for a round i clock when this processor is already in its i 1 st round, it would return ....

F. Cristian and C. Fetzer, Probabilistic Internal Clock Synchronization, Proceedings of the 13th Symposium on Reliable Distributed Systems, Oct. 1994, Dana Point, CA. Pages 22--31.

....other. This value will be approximated by 2aeT ffi because drift rates are very small. For most quartz clocks available in commodity workstations, the drift rates are in the order of 10 Gamma5 or 10 Gamma6 , and for high precision clocks ae is in the order of 10 Gamma7 or 10 Gamma8 [3]. The value of the scheduling delay is normally in order of tens of milliseconds, however, we will only require that ffi is smaller than the checkpoint period. Processes communicate by exchanging messages. Messages may be lost while in transit, arrive out of order, be duplicated, or may be ....

F. Cristian and C. Fetzer. Probabilistic internal clock synchronization. In Proceedings of the 13th Symposium on Reliable Distributed Systems, pages 22--31, October 1994.

....used methods are an exact synchronization and drift estimation before the beginning of the application [8, 9] or better before and after the application [20] with a linear interpolation while the application is running. The exact synchronization can be done by deterministic [28] or probabilistic [2, 3] methods. Continuous synchronization with little resource usage (e.g. xntp [22] is normally not sufficiently exact due to the large jitter of the message delays in local area networks. The trace based synchronization is another alternative. The differences between the clocks are computed by the ....

F. Cristian and C. Fetzer. Probabilistic internal clock synchronization. Technical Report CS94-367, University of California, San Diego, May 18 1995. ftp://cs.ucsd.edu/pub/team/internalProbClockSync.ps.Z, ftp://cs.ucsd.edu/pub/cfetzer/CS94-367.ps.Z.

....used methods are an exact synchronization and drift estimation before the beginning of the application [8, 9] or better before and after the application [20] with a linear interpolation while the application is running. The exact synchronization can be done by deterministic [28] or probabilistic [2, 3] methods. Continuous synchronization with little resource usage (e.g. xntp [22] is normally not sufficiently exact due to the large jitter of the message delays in local area networks. The trace based synchronization is another alternative. The differences between the clocks are computed by the ....

F. Cristian and C. Fetzer. Probabilistic internal clock synchronization. In Proceedings. 13th Symposium on Reliable Distributed Systems, Dana Point, CA, USA, Oct. 25-27, 1994, pages 22--31. IEEE Computer Society Press, 1994.

....computes a clock correction term using a fault tolerant averaging convergence function. Examples of ICV algorithms include the original interactive convergence algorithm (CNVA) 8] the fault tolerant midpoint algorithm (FTMA) 18] and the differential fault tolerant midpoint algorithm (DFTMA) [3]. We refer to these algorithms as unistep ICV algorithms, because in each round a TSP will compute a clock correction term using a single call to a convergence function. This research was supported in part by Conselho Nacional de Desenvolvimento Cientfico e Tecnologico (CNPq Brazil) under ....

.... services such as message authentication [6, 16] atomic broadcast [2] or agreement protocols [8, algorithms COM and CSM] Clock dissemination protocols used by ICV algorithms can stagger clock synchronization messages (hereby referred to simply as messages) over time to reduce network contention [3], and they can also limit the number of TSPs that must be informed of a given clock value, which reduces the number of messages per round significantly [4, 11] ICV algorithms can be classified according to the type of clock dissemination protocol they use. Three communication models can be used, ....

[Article contains additional citation context not shown here]

F. Cristian and C. Fetzer, "Probabilistic Internal Clock Synchronization, " Proc. 13th Symp. Rel. Dist. Sys., Oct. 1994, pp. 22-31.

.... of ICV algorithms include the original interactive convergence algorithm (CNVA) 12] the fast convergence algorithm (FCA) 18] the fault tolerant average algorithm (FTAA) 7, 11] the fault tolerant midpoint algorithm (FTMA) 7, 37] the differential fault tolerant midpoint algorithm (DFTMA) [4], and the sliding window algorithm (SWA) 23] We refer to these algorithms as unistep ICV algorithms, because in each round a TSP will compute a clock correction term using a single call to a convergence function. We believe that ICV algorithms are an attractive alternative for synchronizing ....

.... or multicast primitives, in which a TSP disseminates its clock value to a number of destination TSPs using clock synchronization messages (hereby referred to simply as messages) Clock dissemination protocols used by ICV algorithms can stagger messages over time to reduce network contention [4, 5, 21, 25], and they can also limit the number of TSPs that must be informed of a given clock value, which reduces the number of messages per round significantly [20] ICV algorithms can be classified according to the type of clock dissemination protocol they use. Three communication models can be used, ....

[Article contains additional citation context not shown here]

F. Cristian and C. Fetzer, "Probabilistic Internal Clock Synchronization," Proceedings of the 13th Symposium on Reliable Distributed Systems, October 1994, pp. 22-31.

....into which convergence functions are more suitable (if any) to NDCSA. Another open issue is how to use the information gathered during the phase of message interchange to compute a new value for the clock correction, and how it aoeects the algorithm s performance. As Cristian pointed out in [4], the rationale employed for deterministic clock synchronisation may not always be valid for NDCSA. ....

Flaviu Cristian and Christof Fetzer. Probabilistic internal clock synchronization. In Proc. 13th Symposium on Reliable Distributed Systems, pages 2231. IEEE, 1994.

No context found.

F. Cristian and C. Fetzer. Probabilistic internal clock synchronization. In Proceedings of the 13th Symposium on Reliable Distributed Systems, pages 22--31, Dana Point, California, USA, October 1994. IEEE Computer Society Press.

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