Post by ACM WAP Moderator on Mar 28, 2021 12:50:00 GMT 8
Title: An Efficient Resource Prediction of Heterogeneous Workload in Cloud Hosted Datacentres Using Neural Networks
Authors: Asma Ali and Atif Ali
Country: Pakistan
Abstract:
Virtualized data centers are increasingly hosted as cloud Datacenters for Business-critical workloads web servers, mail servers, app servers, etc. Understanding how business-critical workloads demand and use resources is vital in capacity sizing, infrastructure operation and testing, and application performance management. However, relatively little is currently known about these workloads because the knowledge is intricate, large-scale, heterogeneous, shared-clusters and because datacenter operators remain reluctant to share such information. Moreover, the few operators that did share data (e.g., Google and several other
Supercomputing centers) have enabled studies in business intelligence (MapReduce), search, and scientific computing (HPC), but not in business-critical workloads. To alleviate this example, in this work, we conduct a comprehensive study of business-critical workloads hosted in cloud data centers. We collect largescale and long-term workload traces like requested and used resources during a distributed datacenter servicing business-critical workloads. We perform an in-depth analysis of workload traces. Our study sheds light on the workload of cloud datacenter hosting business-critical workloads. This work's results are often used to develop efficient resource management mechanisms for data centers. Moreover, the traces we released during this work are often used for workload verification, modeling and evaluating resource scheduling policies, etc.
Authors: Asma Ali and Atif Ali
Country: Pakistan
Abstract:
Virtualized data centers are increasingly hosted as cloud Datacenters for Business-critical workloads web servers, mail servers, app servers, etc. Understanding how business-critical workloads demand and use resources is vital in capacity sizing, infrastructure operation and testing, and application performance management. However, relatively little is currently known about these workloads because the knowledge is intricate, large-scale, heterogeneous, shared-clusters and because datacenter operators remain reluctant to share such information. Moreover, the few operators that did share data (e.g., Google and several other
Supercomputing centers) have enabled studies in business intelligence (MapReduce), search, and scientific computing (HPC), but not in business-critical workloads. To alleviate this example, in this work, we conduct a comprehensive study of business-critical workloads hosted in cloud data centers. We collect largescale and long-term workload traces like requested and used resources during a distributed datacenter servicing business-critical workloads. We perform an in-depth analysis of workload traces. Our study sheds light on the workload of cloud datacenter hosting business-critical workloads. This work's results are often used to develop efficient resource management mechanisms for data centers. Moreover, the traces we released during this work are often used for workload verification, modeling and evaluating resource scheduling policies, etc.