Analysis of Congestion Control Virtualization on Execution of Hadoop MapReduce Application
Vilson Moro, Maurício A. Pillon, Charles C. Miers, Guilherme Koslovski · 2018
Cloud providers host multiple virtual machines (VMs) configured based on specific versions of operating systems and libraries. The diversity of algorithms and parameters related to TCP constitutes a heterogeneous communication scenario. Legacy congestion control algorithms compromise the performance of VM-hosted applications. Due to total control in the data center, providers can apply the virtualization of congestion control (VCC) to generate optimized algorithms. From the tenant's perspective, virtualization is a transparently performed. Although promising, the application of VCC requires a deep analysis of the impact on the final applications. Thus, the present work dissects the execution of Hadoop MapReduce atop VCC-based scenarios. The experimental analysis discusses the execution time of Hadoop MapReduce and the behavior of intermediate switches queues, highlighting some VCC limitations.