Abstract
Cloud workloads are increasingly heterogeneous such that a single Cloud job may encompass one to several tasks, and tasks belonging to the same job may behave distinctively during their actual execution. This inherent task heterogeneity imposes increased complexities in achieving an energy efficient management of the Cloud jobs. The phenomenon of a few proportions of tasks characterising increased resource intensity within a given job usually lead the providers to over-provision all the encompassed tasks, resulting in majority of the tasks incurring an increased proportions of resource idleness. To this end, this paper proposes a novel analytics framework which integrates a resource estimation module to estimate the resource requirements of tasks a priori, a straggler classification module to classify tasks based on their resource intensity, and a resource optimisation module to optimise the level of resource provision depending on the task nature and various runtime factors. Performance evaluations conducted both theoretically and through practical experiments prove that the proposed methodology performs better than the compared statistical resource estimation methods and existing models of straggler mitigation, and further demonstrate the effectiveness of the proposed methodology in achieving energy conservation by postulating appropriate level of resource provisioning for task execution.
Highlights
IN the Cloud Computing concept, user request arrives in duration, resource intensity etc
This paper proposes a novel analytics framework for mitigating the energy-aware stragglers within jobs to optimise the level of resource provisioning to reduce energy expenditures incurred in the form of idle resource proportions during task execution
The weight assignment is executed in three cascaded phases, Phase I verifies the Similarity Scores assigned by the Profile Information (PI) table, Phase II verifies the correctness of the samples depending on the availability of the actual execution profile and Phase III verifies the usage rate-duration trade-off for every task execution
Summary
IN the Cloud Computing concept, user request arrives in duration, resource intensity etc. Giving a special emphasis to this inherent task heterofor the LXCs (Linux Containers) or VMs (Virtual Ma- geneity, this paper postulates the phenomenon of a few chines) to consume of the physical resources This prede- tasks within a single job exhibiting a resource intensivefined level of resource provision is usually the maximum ness as an increased multiple of majority of the remaining level of allowed resources for the LXCs or VMs to con- co-located tasks as energy-aware straggling behaviour of sume. Node-level long tail stragglers usually affect the overall completion time of the entire job but energy-aware stragglers, posing a significant energy implication, have not gained suffice importance so far To this end, this paper proposes a novel analytics framework for mitigating the energy-aware stragglers within jobs to optimise the level of resource provisioning to reduce energy expenditures incurred in the form of idle resource proportions during task execution.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have