Quantitative Comparison of Job-level Dynamic Scheduling Policies in Parallel Real-time Systems

Abstract

Scheduling algorithms play an important role in design of real-time systems. Due to high processing power and low price of multiprocessors, real-time scheduling in such systems is more interesting; however, more complicated. Earliest Deadline First (EDF) and Least Laxity First (LLF) are two well-known and extensively applied dynamic scheduling algorithms on which many researches have already been done.However, to the best of our knowledge, the efficiency of aforementioned algorithms has not been compared under similar conditions. Perhaps the main reason is that LLF algorithm is fully dynamic and impractical to implement. In this research, we have used a job-level dynamic and practical version of LLF which is called Modified Least Laxity First (MLLF) algorithm instead of the traditional LLF and have compared its performance with EDF algorithm from many different aspects. The success ratio has been chosen as the key factor for evaluation of the algorithms.

Cost-Minimizing Scheduling of Workflows on a Cloud of Memory Managed Multicore Machines

“…Formulations of realistic scheduling problems are typically found to be NPcomplete, hence heuristic scheduling policies are generally employed to provide acceptable scheduling solutions, e.g., refer to [7,8,9,10,11,12]. The scheduling evaluations conducted in the present paper account for the impact that garbage collection has on a machine's performance.…”

Investigating the effect of processor affinity on uniform parallel machine scheduling

“…If the execution requirement of tasks is known at their arrival, it is possible to predict whether a task will meet its deadline or not. In a real-time system if a task can not complete its execution and misses its deadline for any reason, it not only wastes the CPU time but also lessens the chance of the other tasks waiting for execution to be completed [4]. Apparently, a work conserving algorithm should not execute jobs which will finally miss their deadline.…”

“…This evaluation can be conducted in two major ways [4]: Examining the effect of a specific parameter (e.g. load factor) on a variety of performance metrics as dependent variables.…”

Deadline Scheduling with Processor Affinity and Feasibility Check on Uniform Parallel Machines

“…In a real-time system if a task can not complete its execution and misses its deadline for any reason, it not only wastes the CPU time but also lessens the chance of the other tasks waiting for execution to be completed [4]. Apparently, a work conserving algorithm should not execute jobs which will finally miss their deadline.…”

“…4. For all j, ( min( , )) j k m ≤ if no task is assigned to the j th processor, select the task with earliest deadline from remaining tasks of A and assign it to the j th processor.…”