Exploiting peer group concept for adaptive and highly available services
Muhammad Asif Jan, Fahd Ali Zahid, Mohammad Moazam Fraz, Arshad Ali
Abstract
This paper presents a prototype for redundant, highly available and fault tolerant peer to peer framework for data management. Peer to peer computing is gaining importance due to its flexible organization, lack of central authority, distribution of functionality to participating nodes and ability to utilize unused computational resources. Emergence of GRID computing has provided much needed infrastructure and administrative domain for peer to peer computing. The components of this framework exploit peer group concept to scope service and information search, arrange services and information in a coherent manner, provide selective redundancy and ensure availability in face of failure and high load conditions. A prototype system has been implemented using JXTA peer to peer technology and XML is used for service description and interfaces, allowing peers to communicate with services implemented in various platforms including web services and JINI services. It utilizes code mobility to achieve role interchange among services and ensure dynamic group membership. Security is ensured by using Public Key Infrastructure (PKI) to implement group level security policies for membership and service access.
Create a lesson
Related papers
PixelFlow: Token-Level Workload Management for Efficient Distributed DiT Serving
Zhexiang Zhang, Minchen Yu, Yifan Sun et al.
A Kubernetes-Native Request Router for Quality-Aware Inference Serving in the Computing Continuum
Ignjat Karanovic, Pantelis A. Frangoudis, Ivan Čilić et al.
Accelerating Sharded Data Parallelism at Scale with Federated Learning
Gianluca Mittone, Marco Aldinucci
Distributed Edge Inference: an Experimental Study on Multiview Detection
Gianluca Mittone, Giulio Malenza, Marco Aldinucci et al.
P-GADMM: Parallel Group-Based ADMM for Asynchronous Optimization in Heterogeneous Edge Networks
Gaiguo Wei, Qingying Zhang, Heqiang Wang et al.
VERA: Reinforcement Learning for Dynamic Memory Scaling of HPC Workloads in Kubernetes
Ade Pramono, Jie Ren, Ivy Peng