Fault-Tolerant Clock Synchronization in Environments with High Message Delay Variation
|
Abstract:
Abstract--- Fault-tolerant clock synchronization is an important requirement in many distributed systems, especially in time-critical and safety-critical applications. Frequently, interactive convergence algorithms are used for fault-tolerant clock synchronization, providing advantages such as fully-distributed operation, low message exchange overhead and simplicity of implementation. This paper presents the measured performance of three interactive convergence, fault-tolerant clock synchronization algorithms. Our experiments were conducted in a distributed UNIX environment featuring high message delay variation, which poses severe constraints on the clock synchronization tightness that may be achieved. The algorithms that were tested in this environment are: FTMA (Fault-Tolerant Midpoint Algorithm) [3], AEFTMA (Adaptive Exponential Averaging Fault-Tolerant Midpoint Algorithm) [4] and SWA (Sliding Window Algorithm) [1], [2]. Our experimental results indicate that SWA outperforms the other algorithms in this environment, being able to achieve tighter synchronization under different simulated fault conditions. The superiority of SWA can be attributed to its high degree of fault tolerance, combined with its ability to treat messages with much longer than expected delays as faults.
Citations
| 81 | An upper and lower bound for clock synchronization – Lundelius, Lynch - 1984 |
| 77 | Reaching Approximate Agreement in the Presence of Faults – Dolev, Lynch, et al. - 1986 |
| 68 | The MAFT architecture for distributed fault tolerance – Kieckhafer, Walter, et al. - 1988 |
| 48 | The accuracy of clock synchronization achieved by tempo in berkeley unix 4.3bsd – Gusella, Zatti - 1989 |
| 10 | Clock synchronization in distributed real time systems – Kopetz, Ochsenreiter - 1987 |
| 9 | A new and improved algorithm for fault-tolerant clock synchronization – Pfluegl, Blough - 1995 |
| 8 | Clock synchronization in MAFT – Thambidurai, Finn, et al. - 1989 |
| 8 | Measured Performance of the Network Time – Mills - 1989 |
| 5 | Fault-Tolerant Clock Synchronization for Distributed Systems with High Message Delay Variation – Azevedo, Blough - 1994 |
| 3 | Clock Synchronization in Fault-Tolerant Systems – Pfluegl - 1992 |
| 3 | Distributed Fault-Tolerant Real Time Systems: the MARS Approach – Kopetz, Damm, et al. - 1989 |
