| | |
Summary: Tolerating Faults While Maximizing Reward
Hakan Aydin, Rami Melhem, Daniel Moss´e
Department of Computer Science
University of Pittsburgh
Pittsburgh, PA 15260
(aydin, melhem, mosse)@cs.pitt.edu
Abstract
The imprecise computation(IC) model is a general schedul-
ing framework, capable of expressing the precision vs. timeli-
ness trade-off involved in many current real-time applications.
In that model, each task comprises mandatory and optional
parts. While allowing greater scheduling flexibility, manda-
tory parts in the IC model have still hard deadlines and hence
they must be completed before the task's deadline even in the
presence of faults. In this paper, we address fault tolerant (FT)
scheduling issues for IC tasks. First, we propose two recovery
schemes, namely Immediate Recovery and Delayed Recovery.
These schemes can be readily applied to provide fault toler-
ance to mandatory parts by scheduling optional parts appro-
priately for recovery operations. After deriving the necessary
|
