An efficient fault-tolerant scheduling algorithm for real-time tasks with precedence constraints in heterogeneous systems
read more
Citations
Risk-resilient heuristics and genetic algorithms for security-assured grid job scheduling
Scheduling security-critical real-time applications on clusters
Performance implications of failures in large-scale cluster scheduling
A novel fault-tolerant scheduling algorithm for precedence constrained tasks in real-time heterogeneous systems
A dynamic and reliability-driven scheduling algorithm for parallel real-time jobs executing on heterogeneous clusters
References
Introduction to Probability
Introduction to Probability
Deadline Scheduling for Real-Time Systems: EDF and Related Algorithms
Task allocation for maximizing reliability of distributed computer systems
Related Papers (5)
Scheduling algorithms for multiprogramming in a hard real-time environment
Fault-tolerance through scheduling of aperiodic tasks in hard real-time multiprocessor systems
Frequently Asked Questions (10)
Q2. What is the description of the proposed algorithm?
To the best of their knowledge, the proposed algorithm is the first of its kind reported in the literature, in that it most comprehensively addresses the issues of fault-tolerance, reliability, real-time, task precedence constraints, and heterogeneity.
Q3. What is the significance of the experimental result?
This experimental result validates the use of the proposed FRCD and eFRCD algorithm to enhance the reliability of the system, especially when tasks either have loose deadlines or no deadlines.
Q4. Why is vj B more complex than ESTi P?
The computation of ESTi B(v) is more complex than that of ESTi P(v), due to the need to judiciously overlap some backup copies on the same processors.
Q5. What are the conditions that each backup copy has to satisfy?
In addition to these conditions, each backup copy has three extra conditions to satisfy, namely, (i) it is allocated on the processor that is different than the one assigned for its primary copy, (ii) its start time is later than the finish time of its primary copy plus the fault detection time δ and (iii) it is allowed to overlap with other backup copies on the same processor if their primary copies are allocated to different processors.
Q6. What is the difference between eFRCD and FGLS?
In addition, Fig.8 indicates that performability of FGLS increases much more rapidly with heterogeneity level than that of eFRCD, implying that FGLS is more sensitive to the change in computational heterogeneity than eFRCD.
Q7. What is the simplest way to determine if viB is a strong primary copy?
Given two tasks vi and vj, (vi, vj)∈ E, if viB is not schedule-preceding vj P, and vi P is a strong primary copy, then vj B and viP can not be allocated to the same processor.
Q8. What is the difference between FGLS and eFRCD?
FGLS and FRCD require more computing resources than eFRCD, which is likely to lead to a relatively low SC when the number of processors is fixed.
Q9. What is the definition of a measure of computational heterogeneity?
A measure of computational heterogeneity is modeled by a function, C: V×P→ Z+, which represents the execution time of each task on each processor.
Q10. Why do eFRCD and FRCD have better reliability?
The FRCD and eFRCD algorithms have much better reliability simply because OV and FGLS do not consider reliability in their scheduling schemes while both FRCD and eFRCD take reliability into account.