Efficient resource management in the cloud: From simulation to experimental validation using a low‐cost Raspberry Pi testbed

Abstract

Within the context of cloud computing, efficient resource management is of great importance as it can result in higher scalability and significant energy and cost reductions over time. Because of the high complexity and costs of cloud environments, however, newly developed resource allocation strategies are often only validated by means of simulations, for example, by using CloudSim or custom-developed simulation tools.This article describes a general approach for the validation of cloud resource allocation strategies, illustrating the importance of experimental validation on physical testbeds. Furthermore, the design and implementation of Raspberry Pi as a Service (RPiaaS), a low-cost embedded testbed built using Raspberry Pi nodes, is presented. RPiaaS aims to facilitate the step from simulations toward experimental evaluations on larger cloud testbeds and is designed using a microservice architecture, where experiments and all required management services are running inside containers.The performance of the RPiaaS testbed is evaluated using several benchmark experiments. The obtained results not only illustrate that the overhead of both using containers and running the required RPiaaS services is minimal but also provide useful insights for scaling up experiments between the Raspberry Pi testbed and a larger more traditional cloud testbed.The introduced validation approach is then illustrated using a case study focusing on the allocation of hierarchically structured tenant data. The results obtained through simulations are compared to the experimental results. The RPiaaS testbed proved to be a very useful tool for the initial experimental validation before moving the experiments to a large-scale testbed.

Impact of etcd deployment on Kubernetes, Istio, and application performance

“…The load generation cluster is configured to generate a consistent load level in a closed-loop manner, as defined in Section 2.2. The number of initial Pods used to start each scenario for autoscaling varies in the range (J ∈ [1,2,3,4,5,10,15,20]). To reduce the overall running time of the entire suite of experiments and because we expect the successful request rate to plateau with larger numbers of initial Pods, the experiment's resolution is reduced when J > 5.…”

“…To ensure that results are stable and persistent, it makes extensive use of disk I/O. It is known in the etcd community that poor I/O per second performance will negatively affect etcd 5 . However, to our knowledge there is no published information on how such poor performance affects Kubernetes and applications deployed onto it.…”

“…Like many research experiments conducted today, [5][6][7][8] we used a privately owned but shared cloud infrastructure to carry out our experiments.…”

Resource Management in a Containerized Cloud: Status and Challenges

“…Resource management strategies are often only validated by means of simulations [98], for example by using CloudSim [42], in which the whole cloud computing environment is modeled and simulated in software. This is mainly because of the nature of the research, as resource allocation strategies for example are often designed for managing large sets of applications within large cloud environments.…”

A self‐learning fuzzy approach for proactive resource provisioning in cloud environment

“…Within the context of cloud computing, autonomous resource management is of great importance as it can result in higher scalability and cost reductions over time . Today, most of the autonomous resource provisioning approaches can be categorized as either being reactive or proactive (predictive).…”