DMCA
Optimistic Recovery in Multi-threaded Distributed Systems
Cached
Download Links
- [www.ece.utexas.edu]
- [www.ece.utexas.edu]
- [users.ece.utexas.edu]
- DBLP
Other Repositories/Bibliography
| Citations: | 6 - 1 self |
BibTeX
@TECHREPORT{Damani_optimisticrecovery,
author = {Om P. Damani and Ashis Tarafdar and Vijay K. Garg},
title = {Optimistic Recovery in Multi-threaded Distributed Systems},
institution = {},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
The problem of recovering distributed systems from crash failures has been widely studied in the context of traditional non-threaded processes. However, extending those solutions to the multi-threaded scenario presents new problems. We identify and address these problems for optimistic logging protocols. There are two natural extension to optimistic logging protocols in the multi-threaded scenario. The first extension is process-centric, where the points of internal nondeterminism caused by threads are logged. The second extension is thread-centric, where each thread is treated as a separate process. The process-centric approach suffers from false causality while the thread-centric approach suffers from high causality tracking overhead. By observing that the granularity of failures can be different from the granularity of rollbacks, we design a new balanced approach which incurs low causality tracking overhead and also eliminates false causality. 1.
Keyphrases
multi-threaded distributed system optimistic recovery multi-threaded scenario false causality optimistic logging protocol crash failure second extension first extension new problem high causality separate process low causality internal nondeterminism natural extension traditional non-threaded process process-centric approach suffers thread-centric approach suffers
