Deep‐Q learning‐based heterogeneous earliest finish time scheduling algorithm for scientific workflows in cloud

2020 ◽  
Author(s):  
Avinash Kaur ◽  
Parminder Singh ◽  
Ranbir Singh Batth ◽  
Chee Peng Lim
Keyword(s):  
Scheduling Algorithm ◽  
Scientific Workflows ◽  
Q Learning ◽  
Finish Time ◽  
Time Scheduling ◽  
Earliest Finish Time

scholarly journals HPEFT for Hierarchical Heterogeneous Multi-DAG in a Multigroup Scan UPA System

Electronics ◽  
2019 ◽  
Vol 8 (5) ◽  
pp. 498
Author(s):  
Yuzhong Li ◽  
Wenming Tang ◽  
Guixiong Liu
Keyword(s):  
Signal Processing ◽  
Scheduling Algorithm ◽  
Frame Rate ◽  
Hierarchical Processing ◽  
Finish Time ◽  
System A ◽  
Time Period ◽  
Upa System ◽  
Ultrasonic Phased Array ◽  
Earliest Finish Time

Multidirected acyclic graph (DAG) workflow scheduling is a key problem in the heterogeneous distributed environment in the distributed computing field. A hierarchical heterogeneous multi-DAG workflow problem (HHMDP) was proposed based on the different signal processing workflows produced by different grouping and scanning modes and their hierarchical processing in specific functional signal processing modules in a multigroup scan ultrasonic phased array (UPA) system. A heterogeneous predecessor earliest finish time (HPEFT) algorithm with predecessor pointer adjustment was proposed based on the improved heterogeneous earliest finish time (HEFT) algorithm. The experimental results denote that HPEFT reduces the makespan, ratio of the idle time slot (RITS), and missed deadline rate (MDR) by 3.87–57.68%, 0–6.53%, and 13–58%, respectively, and increases relative relaxation with respect to the deadline (RLD) by 2.27–8.58%, improving the frame rate and resource utilization and reducing the probability of exceeding the real-time period. The multigroup UPA instrument architecture in multi-DAG signal processing flow was also provided. By simulating and verifying the scheduling algorithm, the architecture and the HPEFT algorithm is proved to coordinate the order of each group of signal processing tasks for improving the instrument performance.

scholarly journals EHEFT-R: multi-objective task scheduling scheme in cloud computing

2021 ◽  
Author(s):  
Honglin Zhang ◽  
Yaohua Wu ◽  
Zaixing Sun
Keyword(s):  
Cloud Computing ◽  
Task Scheduling ◽  
Scheduling Algorithm ◽  
Solution Space ◽  
Finish Time ◽  
Task Scheduling Algorithm ◽  
Matching Efficiency ◽  
Earliest Finish Time ◽  
Core Issues

AbstractIn cloud computing, task scheduling and resource allocation are the two core issues of the IaaS layer. Efficient task scheduling algorithm can improve the matching efficiency between tasks and resources. In this paper, an enhanced heterogeneous earliest finish time based on rule (EHEFT-R) task scheduling algorithm is proposed to optimize task execution efficiency, quality of service (QoS) and energy consumption. In EHEFT-R, ordering rules based on priority constraints are used to optimize the quality of the initial solution, and the enhanced heterogeneous earliest finish time (HEFT) algorithm is used to ensure the global performance of the solution space. Simulation experiments verify the effectiveness and superiority of EHEFT-R.

scholarly journals A generic approach to scheduling and checkpointing workflows

2019 ◽  
Vol 33 (6) ◽  
pp. 1255-1274 ◽  
Author(s):  
Li Han ◽  
Valentin Le Fèvre ◽  
Louis-Claude Canon ◽  
Yves Robert ◽  
Frédéric Vivien
Keyword(s):  
Large Scale ◽  
Rare Events ◽  
Soft Errors ◽  
Scientific Workflows ◽  
Significant Gain ◽  
Trade Off ◽  
Finish Time ◽  
Dependence Graphs ◽  
Earliest Finish Time ◽  
Special Classes

This work deals with scheduling and checkpointing strategies to execute scientific workflows on failure-prone large-scale platforms. To the best of our knowledge, this work is the first to target fail-stop errors for arbitrary workflows. Most previous work addresses soft errors, which corrupt the task being executed by a processor but do not cause the entire memory of that processor to be lost, contrarily to fail-stop errors. We revisit classical mapping heuristics such as Heterogeneous Earliest Finish Time and MinMin and complement them with several checkpointing strategies. The objective is to derive an efficient trade-off between checkpointing every task (CkptAll), which is an overkill when failures are rare events, and checkpointing no task (CkptNone), which induces dramatic re-execution overhead even when only a few failures strike during execution. Contrarily to previous work, our approach applies to arbitrary workflows, not just special classes of dependence graphs such as minimal series-parallel graphs. Extensive experiments report significant gain over both CkptAll and CkptNone for a wide variety of workflows.

Sign in / Sign up

Export Citation Format

Share Document