Perfect failure detection in timed asynchronous systems
- 19 February 2003
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 52 (2) , 99-112
- https://doi.org/10.1109/tc.2003.1176979
Abstract
Perfect failure detectors can correctly decide whether a computer is crashed. However, it is impossible to implement a perfect failure detector in purely asynchronous systems. We show how to enforce perfect failure detection in timed asynchronous systems with hardware watchdogs. The two main system model assumptions are: 1) each computer can measure time intervals with a known maximum error and 2) each computer has a watchdog that crashes the computer unless the watchdog is periodically updated. We have implemented a system that satisfies both assumptions using a combination of off-the-shelf software and hardware. To implement a perfect failure detector for process crash failures, we show that, in some systems, a hardware watchdog is actually not necessary.Keywords
This publication has 15 references indexed in Scilit:
- Approximate real-time clocks for scheduled eventsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2003
- Simulating fail-stop in asynchronous distributed systemsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Rejuvenation and failure detection in partitionable systemsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- The timed asynchronous distributed system modelIEEE Transactions on Parallel and Distributed Systems, 1999
- Fail-aware datagram serviceIEE Proceedings - Software, 1999
- Unreliable failure detectors for reliable distributed systemsJournal of the ACM, 1996
- Understanding the limitations of causally and totally ordered communicationPublished by Association for Computing Machinery (ACM) ,1993
- Leases: an efficient fault-tolerant mechanism for distributed file cache consistencyPublished by Association for Computing Machinery (ACM) ,1989
- Impossibility of distributed consensus with one faulty processJournal of the ACM, 1985
- Distributed snapshotsACM Transactions on Computer Systems, 1985