Dynamic Object Replica Placement using Underlying Routing Protocols: Ensuring the Reliability (original) (raw)

Dynamic Object Replica Placement Using Underlying Routing Protocols: Ensuring Reliability

Object replication is an unavoidable phenomenon in distributed systems to address the issues of size, administrative and resources scalability. It also guarantees the availability of system as more users can be accommodated and failures can be tolerated. If the location of object replica is not proper, the problem may arise when it is requested by many entities at once. The response time may also increase due to the network latency although network is considered to be steadfast in the distributed systems but the latency can never be avoided. Efficient replica placement, both client initiated and server initiated, can be ensured using the dynamic replica placement mechanism in conjunction with the underlying routing protocols that provides the reliability of contents distribution and minimizes the network latency. In this paper, it is discussed that reliability can be assured if we use replica placement algorithm along with the EIGRP (Enhanced Interior Gateway Routing Protocol). The issues of choosing the best path for placing the replica that provides the minimum latency and reliability of contents transmission can be overcome by combining the EIGRP features that includes internetwork delay, bandwidth consumption, reliability of packet transmission and traffic load, with dynamic replica placement algorithm.

Dynamic replica placement for scalable content delivery

2002

In this paper, we propose the dissemination tree, a dynamic content distribution system built on top of a peer-to-peer location service. We present a replica placement protocol that builds the tree while meeting QoS and server capacity constraints. The number of replicas as well as the delay and bandwidth consumption for update propagation are significantly reduced. Simulation results show that the dissemination tree has close to the optimal number of replicas, good load distribution, small delay and bandwidth penalties for update multicast compared with the ideal case: static replica placement on IP multicast.

Continuous Replica Placement schemes in distributed systems

Proceedings of the 19th annual international conference on Supercomputing - ICS '05, 2005

The Replica Placement Problem (RPP) aims at creating a set of duplicated data objects across the nodes of a distributed system in order to optimize certain criteria. Typically, RPP formulations fall into two categories: static and dynamic. The first assumes that access statistics are estimated in advance and remain static, and, therefore, a one-time replica distribution is sufficient (1RPP). In contrast, dynamic methods change the replicas in the network potentially upon every request. This paper proposes an alternative technique, named Continuous Replica Placement Problem (CRPP), which falls between the two extreme approaches. CRPP can be defined as: Given an already implemented replication scheme and estimated access statistics for the next time period, define a new replication scheme, subject to optimization criteria and constraints. As we show in the problem formulation, CRPP is different in that the existing heuristics in the literature cannot be used either statically or dynamically to solve the problem. In fact, even with the most careful design, their performance will be inferior since CRPP embeds a scheduling problem to facilitate the proposed mechanism. We provide insight on the intricacies of CRPP and propose various heuristics.

Design and analysis of an adaptive object replication algorithm in distributed network systems

Computer Communications, 2008

In this paper, we propose an adaptive object replication algorithm for distributed network systems, analyze its performance from both theoretical and experimental standpoints. We first present a mathematical cost model that considers all the costs associated with servicing a request, i.e., I/O cost, control-message transferring cost, and data-message transferring cost. Using this cost model, we develop an adaptive object replication algorithm, referred to as Adaptive Distributed Request Window (ADRW) algorithm. Our objective is to dynamically adjust the allocation schemes of objects based on the decision of ADRW algorithm, i.e., whether the system is read-intensive or write-intensive, so as to minimize the total servicing cost of the arriving requests. Competitive analysis is carried out to study the performance of ADRW algorithm theoretically. We then implement our proposed algorithm in a PC based network system. The experimental results convincingly demonstrate that ADRW algorithm is adaptive and is superior to several related algorithms in the literature in terms of the average request servicing cost.

Replica placement in ring based content delivery networks

Computer Communications, 2006

The recent introduction of Content Distribution Networks (CDNs) enhances the delivery of high quality multimedia content to end users. In a CDN architecture, the content is replicated to so-called surrogate servers, generally at the edge of the transport network, to improve the quality of service (QoS) of streaming multimedia delivery services. By using peer-to-peer (P2P) technologies, these edge servers can co-operate and provide a more scalable and robust service in a self-organizing CDN.

Performance evaluation of different replica placement algorithms

International Journal of Grid and Utility Computing, 2009

One of the challenges for data replication in a Data Grid is to select the candidate sites that will host the replicas of datasets. In our earlier research (Rahman et al., 2006, May; Rahman et al., 2007, May), we propose different replica placement algorithms based on three mathematical models, i.e. p-centre, p-median and a multi-objective models. We also present a replica maintenance algorithm to relocate replicas to different sites if the performance of the replica host sites degrades significantly. In this research, we validate our replica placement algorithms with different performance metrics, e.g. total file transfer time, the number of local and remote file access, with accuracy. To evaluate our replica placement algorithms, we use a Data Grid simulator called OptorSim (Bell et al., 2003). This paper presents detailed execution flows of the replication algorithms in the simulator. The study of our replica placement algorithms is carried out using a model of the EU Data Grid Testbed 1 (Bell et al., 2003) sites and the associated network geometry. Jobs are based on the CDF use-case as described in Huffman et al. (2002).

Load Balancing Using Dynamic Replication Scheme for the Distributed Object Group

2006

CORBA is the most widely used middleware for implementing distributed application. Currently, object implementations facilitate object group model to organize the object services. In addition to the complexity of designing the object groups, researchers also seek to improve the quality of service (QoS) by various means such as implementing load balancing to the system. This paper deals with the proposed load balancing service of distributed object groups. The proposed load balancing service uses the dynamic replication scheme which is mechanized by flow balance assumption (FBA) that derives from the arrival and service rate to execute request forwarding to new objects. The proposed on-demand replication scheme adjusts the number of replicated objects based on the arrival rate to minimize the waiting time of clients. It consists of procedures such as intercepting the request and executing on-demand activation of objects. The result of the simulation shows the improvement of the total mean client request completion time of the system as compared to other load balancing schemes.

Insuring Locality: A load balancing protocol for fully replicated system

The distribution of data via replication within or between organizations is crucial for many applications. Such a distribution may be transparent, as it is embodied in distributed database systems to increase their performance, reliability and availability. Many distributed object systems based on fully replication is suffering from bottleneck problem. The bottleneck problem comes from the fact that there is no load-balance between the nodes in the system. In such system only one node is managing the whole replicated object store. This manager node is responsible about the store consistency and transaction serialization. This work investigates the one node manager problems and presents a new protocol for insuring locality and solves the bottleneck problem, by having more than one manager the system increasing No. of manager in the system would solve the load balancing issue in the system.

Reliable Communication Infrastructure for Adaptive Data Replication

Lecture Notes in Computer Science, 2009

In this paper, we propose a reliable group communication solution dedicated to a data replication algorithm to adapt it to unreliable environments. The data replication algorithm, named Adaptive Data Replication (ADR), has already an adaptiveness mechanism encapsulated in its dynamic replica placement strategy. Our extension of ADR to unreliable environments provides a data replication solution that is adaptive both in terms of replica placement and in terms of request routing. At the routing level, this solution takes the unreliability of the environment into account, in order to maximize reliable delivery of requests. At the replica placement level, the dynamically changing origin and frequency of read/write requests are analyzed, in order to define a set of replica that minimizes communication cost. Performance evaluation shows that this original combination of two adaptive strategies makes it possible to ensure high request delivery, while minimizing communication costs in the system.

Latency-Driven Replica Placement

The 2005 Symposium on Applications and the Internet, 2005

This paper presents HotZone, an algorithm to place replicas in a wide-area network such that the client-to-replica latency is minimized. Similar to the previously proposed HotSpot algorithm, HotZone places replicas on nodes that along with their neighboring nodes generate the highest load. In contrast to HotSpot, however, HotZone provides nearly-optimal results by considering overlapping neighborhoods. HotZone relies on a geometric model of Internet latencies, which effectively reduces the cost of placing K replicas among N potential replica locations from O(N 2 ) to O(N ยท max(log N, K)).