On Fast and Coordinated Data Backup in Geo-Distributed Optical Inter-Datacenter Networks

Yao, Jingjing; Lu, Ping; Gong, Lei; Zhu, Zuqing

doi:10.1109/jlt.2015.2425303

Cited by 61 publications

(53 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For the researches about selection of backup transmission paths, several studies [18][19][20][21] have not considered how to allocate bandwidth for multiple concurrent backup transfers (especially in shared links) during data transmission. Some studies 1,22,23 have constructed a mutual backup model but not considered fault-tolerant performance and fair load distribution to backup DCs. In our early work, 24,25 we leverage unicast multipath routing for rapid and fair backup data transmission under constraints of receiving capacity and redundancy guarantee but have not considered bandwidth cost optimization.…”

Section: Disaster Backup Among Mutiple Dcsmentioning

confidence: 99%

Cost‐efficient disaster backup for multiple data centers using capacity‐constrained multicast

Wang

Shen

et al. 2019

Concurrency and Computation

View full text Add to dashboard Cite

To leverage periodic disaster backup in a cloud data center (DC) network, previous studies employ disjoint unicast paths for bulk data transfers among multiple geographically distributed DCs, causing massive unnecessary traffic duplication. This not only adds the overhead but also may result in severe network congestion. With flexible network resource management in software-defined networks and powerful traffic aggregation capability of multicast, we propose capacity-constrained multicast to realize cost-efficient disaster backup. First, considering limited backup storage capacity and essential redundancy guarantee, we construct a capacity-constrained multicasting backup model. Then, we formulate the disaster backup problem as capacity-constrained multiple Steiner tree problem, which is NP-hard. To solve this problem, we design a new multicasting backup ant colony optimization algorithm based on requirement-aware growth. It directly optimizes every disaster-backup multicast tree (DBMT) from its root node to cover enough destination nodes guaranteeing sufficient redundancy and then expands them into the forest under the guidance of a multicast tree shared degree, the ratio of available storage capacity, and backup load distribution offset. We introduce unique edge fitness evaluation and pheromones for every DBMT to reduce mutual influences among multiple trees. Extensive simulations demonstrate that our strategy performs with less bandwidth consumption cost and relatively good backup load distribution fairness simultaneously. KEYWORDScapacity-constrained, cost-efficient, disaster backup, multiple Steiner tree INTRODUCTIONWith more and more frequent occurrence of natural disasters and human-made intentional attacks on application data centers (DCs), 1,2 a large number of data face the risk of failure. Therefore, people need to leverage periodic backup among multiple geographically distributed (geo-distributed) DCs. 3 To improve the fault-tolerant performance and obtain sufficient data redundancy, terabytes to petabytes of data produced during a period are regularly replicated and assigned to three or more other remote DCs, 4 called disaster backup. Due to the huge amount of data transfers, disaster backup activity always consumes huge bandwidth and imposes a heavy load to the links between DCs and may even cause network congestion. Therefore, to support quality-of-service requirements of various real-time applications in network, the traffic engineering problem for disaster backup among multiple DCs is of significant importance.It is noteworthy that in the research area of traffic engineering for multiple disaster-backup transfers, unicast routing has been studied widely, while multicast routing attracts much less attention. Multicast is designed to deliver contents to a large number of destinations and benefits DC group communications in three aspects. By saving network traffic, it can increase the throughput of bandwidth-hungry computations. By releasing the sender from sending multiple copies of packets to diffe...

show abstract

Section: Disaster Backup Among Mutiple Dcsmentioning

confidence: 99%

Cost‐efficient disaster backup for multiple data centers using capacity‐constrained multicast

Wang

Shen

et al. 2019

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…The classical protection schemes do not provide protection against dominant failure scenario (disaster) especially when it covers a region which affects both the primary and backup paths. Protection schemes include dedicated path protection (DPP), shared path protection (SPP), dedicated link protection and shared link protection while restoration schemes include path restoration and link restoration [31][32][33][34][35][36][37][38][39][40][41][42][43]. An organization of these schemes is shown in …”

Section: Fault Management In Wdm Networkmentioning

confidence: 99%

“…In [35], the authors address the problem of fast and coordinated data backup in geographically distributed optical inter-DC networks in order to improve the data-transfer efficiency of the regular DC backup. In order to prevent data loss in an optical inter-datacenter (inter-DC) network, a cloud system usually leverages multiple datacenters (DCs) for obtaining sufficient data redundancy.…”

mentioning

confidence: 99%

Dynamic provisioning of fault tolerant optical networks for data centers

Das

Bandyopadhyay

Mamoori

et al. 2017

2017 IEEE 30th Canadian Conference on Electrical and Computer Engineering (CCECE)

View full text Add to dashboard Cite

“…To ensure effective service provisioning, the time required for lightpath configuration should be at least guaranteed within TS's duration [29,30]. On the other hand, to adapt to the changing traffic load, network operations should be adjusted according to the dynamics of traffic demands.…”

Section: Problem Description a Network Modelmentioning

confidence: 99%

On Hybrid IR and AR Service Provisioning in Elastic Optical Networks

Lü

Zhu

Mukherjee

2015

J. Lightwave Technol.

Self Cite

View full text Add to dashboard Cite

This paper investigates hybrid immediate reservation (IR) and advance reservation (AR) service provisioning in elastic optical networks (EONs), with the objective to minimize IR/AR service conflicts. We design algorithms to coordinate service provisioning of IR and AR requests. Specifically, both proactive and reactive IR provisioning schemes are considered to minimize IR service failures. Our AR scheduling algorithm can coordinate AR service provisioning with various IR traffic patterns, balance spectrum utilization in both time and spectral domains, and reduce IR/AR service conflicts. Simulation results verify that our proposed IR + AR schemes can significantly reduce IR service failures as well as routing and spectrum allocation (RSA) reconfigurations in IR service provisioning. Moreover, the results indicate that the proposed IR + AR schemes can achieve more performance gain from AR flexibility when compared with two existing benchmarks.AR requests in an EON, where spectrum resources are fully shared among IR and AR requests. Specifically, we consider a dynamic EON, in which the IR requests' service durations are unknown while the AR ones use the specific start time and specific duration (STSD) reservation model [6]. For the STSD model, we consider both the STSD-fixed and STSD-flexible cases. The STSD-fixed case means that the service start time of an AR request is fixed in the future, while STSD-flexible reservation allows the service start time of an AR request to slide within a pre-defined time window. Regarding the relation between IR and AR requests, we assume that the AR ones possess higher priority, which means that, whenever a service conflict occurs, an in-service IR request has to compromise to the AR request which shares resources with it. Based on this model, we design both proactive and reactive IR provisioning algorithms to minimize IR service failures (i.e., including both service blocking and interruption), and also propose an AR scheduling algorithm that considers different IR traffic patterns to coordinate AR provisioning, to balance resource utilization in both time and spectral domains, and to eventually relieve IR/AR service conflicts in the EON.The rest of the paper is organized as follows. Section II reviews related work. In Section III, we describe the network model and define the problem of hybrid IR/AR service provisioning in an EON. To minimize IR/AR service conflicts, novel provisioning algorithms are proposed in Section IV. Section V discusses the simulations for performance evaluation. Finally, we summarize the paper in Section VI. 0733-8724 (c)

show abstract

On Fast and Coordinated Data Backup in Geo-Distributed Optical Inter-Datacenter Networks

Cited by 61 publications

References 28 publications

Cost‐efficient disaster backup for multiple data centers using capacity‐constrained multicast

Cost‐efficient disaster backup for multiple data centers using capacity‐constrained multicast

Dynamic provisioning of fault tolerant optical networks for data centers

On Hybrid IR and AR Service Provisioning in Elastic Optical Networks

Contact Info

Product

Resources

About