Loading...
Please wait, while we are loading the content...
Similar Documents
N.: Termination Detection in an Asynchronous Distributed System with Crash-Recovery Failures (2006)
| Content Provider | CiteSeerX |
|---|---|
| Author | Freiling, Felix C. Majuntke, Matthias Mittal, Neeraj |
| Abstract | In practice, it cannot easily be detected whether a computation running in a distributed system has terminated or not. Thus, suitable observing algorithms are required to solve this problem of termination detection. A termination detection algorithm involves a computation of its own and the computation it observes without interfering it. Additionally, it satisfies two properties: (1) it should never announce termination unless the underlying computation has in fact terminated. (2) If the underlying computation has terminated, the termination detection algorithm should eventually announce termination. For the definition of termination, the states of processes are mapped to just two distinct states: active and passive. An active process still actively participates in the computation while a passive process does not participate anymore unless it is activated by an active process. In message-passing systems, which we also assume here, activation can only be done by receiving a message. A widely accepted definition of termination is that (1) all processes are passive and (2) all channels are empty. |
| File Format | |
| Language | English |
| Publisher Date | 2006-01-01 |
| Access Restriction | Open |
| Subject Keyword | Termination Detection Crash-recovery Failure Asynchronous Distributed System Termination Detection Algorithm Active Process Message-passing System Distributed System Distinct State Passive Process |
| Content Type | Text |
| Resource Type | Technical Report |
