Abstract
The energy consumption of underlying cloud hardware has dramatically increased. The cloud service providers need to adopt some cost-effective and energy-Aware job scheduler without compromising the quality of service (QoS) specified in the service level agreement (SLA). Based on a rigorous mathematical model, we formulate an energy efficient problem to improve the resource utilisation for high system throughput. A multiple-procedure heuristic workflow scheduling and consolidation strategy is proposed with objectives to maximise the resource utilisation and minimise the power. Several techniques have been utilised including dynamic voltage and frequency scaling (DVFS) with task module migration for workload balance and task consolidation for virtual machine (VM) overhead reduction. The simulation results illustrate that our approach consistently achieves a lower power consumption and higher resource utilisation rate within the execution time bound compared with other similar scheduling algorithms as well as our previous algorithm without the task migration based on VM threshold.
| Original language | English |
|---|---|
| Pages (from-to) | 268-284 |
| Number of pages | 17 |
| Journal | International Journal of Computational Science and Engineering |
| Volume | 13 |
| Issue number | 3 |
| DOIs | |
| State | Published - 1 Jan 2016 |
Fingerprint
Keywords
- Cloud computing
- Dag
- Directed acyclic graph
- Energy-efficient
- Scientific workflow scheduling
- Task consolidation
Cite this
}
Energy-efficient task scheduling and consolidation algorithm for workflow jobs in cloud. / Khaleel, Mustafa; Zhu, Michelle.
In: International Journal of Computational Science and Engineering, Vol. 13, No. 3, 01.01.2016, p. 268-284.Research output: Contribution to journal › Article › Research › peer-review
TY - JOUR
T1 - Energy-efficient task scheduling and consolidation algorithm for workflow jobs in cloud
AU - Khaleel, Mustafa
AU - Zhu, Michelle
PY - 2016/1/1
Y1 - 2016/1/1
N2 - The energy consumption of underlying cloud hardware has dramatically increased. The cloud service providers need to adopt some cost-effective and energy-Aware job scheduler without compromising the quality of service (QoS) specified in the service level agreement (SLA). Based on a rigorous mathematical model, we formulate an energy efficient problem to improve the resource utilisation for high system throughput. A multiple-procedure heuristic workflow scheduling and consolidation strategy is proposed with objectives to maximise the resource utilisation and minimise the power. Several techniques have been utilised including dynamic voltage and frequency scaling (DVFS) with task module migration for workload balance and task consolidation for virtual machine (VM) overhead reduction. The simulation results illustrate that our approach consistently achieves a lower power consumption and higher resource utilisation rate within the execution time bound compared with other similar scheduling algorithms as well as our previous algorithm without the task migration based on VM threshold.
AB - The energy consumption of underlying cloud hardware has dramatically increased. The cloud service providers need to adopt some cost-effective and energy-Aware job scheduler without compromising the quality of service (QoS) specified in the service level agreement (SLA). Based on a rigorous mathematical model, we formulate an energy efficient problem to improve the resource utilisation for high system throughput. A multiple-procedure heuristic workflow scheduling and consolidation strategy is proposed with objectives to maximise the resource utilisation and minimise the power. Several techniques have been utilised including dynamic voltage and frequency scaling (DVFS) with task module migration for workload balance and task consolidation for virtual machine (VM) overhead reduction. The simulation results illustrate that our approach consistently achieves a lower power consumption and higher resource utilisation rate within the execution time bound compared with other similar scheduling algorithms as well as our previous algorithm without the task migration based on VM threshold.
KW - Cloud computing
KW - Dag
KW - Directed acyclic graph
KW - Energy-efficient
KW - Scientific workflow scheduling
KW - Task consolidation
UR - http://www.scopus.com/inward/record.url?scp=84987652851&partnerID=8YFLogxK
U2 - 10.1504/IJCSE.2016.078933
DO - 10.1504/IJCSE.2016.078933
M3 - Article
VL - 13
SP - 268
EP - 284
JO - International Journal of Computational Science and Engineering
JF - International Journal of Computational Science and Engineering
SN - 1742-7185
IS - 3
ER -