A Cooperative Approach for Distributed Task Execution in Autonomic Clouds

Abstract

Abstract-Virtualization and distributed computing are two key pillars that guarantee scalability of applications deployed in the Cloud. In Autonomous Cooperative Cloud-based Platforms, autonomous computing nodes cooperate to offer a PaaS Cloud for the deployment of user applications. Each node must allocate the necessary resources for customer applications to be executed with certain QoS guarantees. If the QoS of an application cannot be guaranteed a node has mainly two options: to allocate more resources (if it is possible) or to rely on the collaboration of other nodes. Making a decision is not trivial since it involves many factors (e.g. the cost of setting up virtual machines, migrating applications, discovering collaborators). In this paper we present a model of such scenarios and experimental results validating the convenience of cooperative strategies over selfish ones, where nodes do not help each other. We describe the architecture of the platform of autonomous clouds and the main features of the model, which has been implemented and evaluated in the DEUS discrete-event simulator. From the experimental evaluation, based on workload data from the Google Cloud Backend, we can conclude that (modulo our assumptions and simplifications) the performance of a volunteer cloud can be compared to that of a Google Cluster.

Analyzing real cluster data for formulating allocation algorithms in cloud platforms

“…Di et al used it in [10] to test fault-tolerance optimization techniques based on a checkpointing/restart mechanism, and in [11] to predict host load using a Bayesian model. Amoretti et al [12] carried out an experimental evaluation of an application-oriented QoS model. On a different context, job burstiness is used as an input for a queue-based model of an elasticity controller in [13].…”

Trusted, Heterogeneous, and Autonomic Mobile Cloud

“…This approach can also be useful against cheating nodes. A model for validating the convenience of cooperative cloud node strategies over selfish ones, where nodes do not help each other, is introduced in [4], describing the architecture of autonomous cloud platforms in a discrete-event simulator. Dautov et al [24] present a vision of cloud application platforms as sensor networks, based on the similarities between the problem domain of cloud application platform monitoring and sensor-enabled domains as traffic surveillance, environmental monitoring or home automation.…”

“…3 An overview of mobile GPU evolution related to OpenCL language support [74] Fig. 4 Architecture of an hardware virtualization supporting mobile CPU [7] Google's Renderscript [26] and NVIDIA's CUDA [64]. The final winner of this contest will shape the future mobile-GPU-as-a-service offerings.…”

AntiCheetah: Trustworthy computing in an outsourced (cheating) environment

“…This work is interesting as their approach can also be of help against cheating. Amoretti et al [42] present a model for validating the convenience of cooperative cloud node strategies over selfish ones, where nodes do not help each other. The authors describe the architecture of the platform of autonomous clouds and the model implemented in a discrete-event simulator.…”