Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userA framework for proving the self-organization of dynamic systems
84
0
0.0
(
0
)
Added by
Maria Potop-Butucaru
Publication date
2010
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
This paper aims at providing a rigorous definition of self- organization, one of the most desired properties for dynamic systems (e.g., peer-to-peer systems, sensor networks, cooperative robotics, or ad-hoc networks). We characterize different classes of self-organization through liveness and safety properties that both capture information re- garding the system entropy. We illustrate these classes through study cases. The first ones are two representative P2P overlays (CAN and Pas- try) and the others are specific implementations of Omega (the leader oracle) and one-shot query abstractions for dynamic settings. Our study aims at understanding the limits and respective power of existing self-organized protocols and lays the basis of designing robust algorithm for dynamic systems.
rate research
Read More
To ensure uninterrupted services to the cloud clients from federated cloud providers, it is important to guarantee an efficient allocation of the cloud resources to users to improve the rate of client satisfaction and the quality of the service provisions. It is better to get as more computing and storage resources as possible. In cloud domain several Multi Agent Resource Allocation methods have been proposed to implement the problem of dynamic resource allocation. However the problem is still open and many works to do in this field. In cloud computing robustness is important so in this paper we focus on auto-adaptive method to deal with changes of open federated cloud computing environment. Our approach is hybrid, we first adopt an existing organizations optimization approach for self organization in broker agent organization to combine it with already existing Multi Agent Resource Allocation approach on Federated Clouds. We consider an open clouds federation environment which is dynamic and in constant evolution, new cloud operators can join the federation or leave this one. At the same time our approach is multi criterion which can take in account various parameters (i.e. computing load balance of mediator agent, geographical distance (network delay) between costumer and provider...).
As networks and their structure have become a major field of research, a strong demand for network visualization has emerged. We address this challenge by formalizing the well established spring layout in terms of dynamic equations. We thus open up the design space for new algorithms. Drawing from the knowledge of systems design, we derive a layout algorithm that remedies several drawbacks of the original spring layout. This new algorithm relies on the balancing of two antagonistic forces. We thus call it {em arf} for attractive and repulsive forces. It is, as we claim, particularly suited for a dynamic layout of smaller networks ($n < 10^3$). We back this claim with several application examples from on going complex systems research.
As distributed systems grow in scale and complexity, the need for flexible automation of systems management functions also grows. We outline a framework for building tools that provide distributed, scalable, declarative, modular, and continuous automation for distributed systems. We focus on four points of design: 1) a state-management approach that prescribes source-of-truth for configured and discovered system states; 2) a technique to solve the declarative unification problem for a class of automation problems, providing state convergence and modularity; 3) an eventual-consistency approach to state synchronization which provides automation at scale; 4) an event-driven architecture that provides always-on state enforcement. We describe the methodology, software architecture for the framework, and constraints for these techniques to apply to an automation problem. We overview a reference application built on this framework that provides state-aware system provisioning and node lifecycle management, highlighting key advantages. We conclude with a discussion of current and future applications.
In recent years, many DHT-based P2P systems have been proposed, analyzed, and certain deployments have reached a global scale with nearly one million nodes. One is thus faced with the question of which particular DHT system to choose, and whether some are inherently more robust and scalable. Toward developing such a comparative framework, we present the reachable component method (RCM) for analyzing the performance of different DHT routing systems subject to random failures. We apply RCM to five DHT systems and obtain analytical expressions that characterize their routability as a continuous function of system size and node failure probability. An important consequence is that in the large-network limit, the routability of certain DHT systems go to zero for any non-zero probability of node failure. These DHT routing algorithms are therefore unscalable, while some others, including Kademlia, which powers the popular eDonkey P2P system, are found to be scalable.
High-performance computing (HPC) is undergoing significant changes. Next generation HPC systems are equipped with diverse global and local resources, such as I/O burst buffer resources, memory resources (e.g., on-chip and off-chip RAM, external RAM/NVRA), network resources, and possibly other resources. Job schedulers play a crucial role in efficient use of resources. However, traditional job schedulers are single-objective and fail to efficient use of other resources. In this paper, we propose ROME, a novel multi-dimensional job scheduling framework to explore potential tradeoffs among multiple resources and provides balanced scheduling decision. Our design leverages genetic algorithm as the multi-dimensional optimization engine to generate fast scheduling decision and to support effective resource utilization.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
