On-Line Multicast Scheduling with Bounded Congestion in Fat-Tree Data Center Networks

Zeng, Guo; Duan, Jun; Yang, Yuanyuan

doi:10.1109/jsac.2014.140110

Cited by 60 publications

(21 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most of these researches focus on the one‐to‐many multicast communication and many‐to‐many concurrent communication. For instance, Guo et al propose an on‐line multicast scheduling algorithm, achieving bounded congestion and efficient bandwidth utilization. Shu et al optimize many‐to‐many concurrent traffic in RapidIO‐based fat‐trees communication.…”

Section: Background and Related Workmentioning

confidence: 99%

In‐network block repairing for erasure coding storage systems

Xia

Guo

Cheng

2019

Concurrency and Computation

View full text Add to dashboard Cite

Summary In the erasure coding storage system, it is necessary to extract multiple data blocks from other remaining storage nodes to a new node when a storage node fails, which repairs the failed data block satisfactorily. However, this would incur the incast problem at this new node. The existing solutions for the repair process in the incast problem mainly rely on path planning and resource allocation. Although these solutions improve the performance of repairing the failed data blocks, they still waste a large amount of storage and bandwidth resources unavoidably. In this paper, we propose the incast problem to be resolved economically via the in‐network aggregation. Specifically, we assume that the switches in data centers have certain data processing capabilities and can aggregate data flows efficiently. Thereafter, we propose a set of in‐network methods to repair a failed data block in the erasure coding storage systems, taking the fat‐tree data center as an example. Thus, the incast problem can be solved effectively during the data transmission process. Compared with the prior methods, our approach effectively avoids the overhead of extra path computing, as well as significantly reduces the link cost of repairing data blocks, while promising similar or faster repair speed.

show abstract

Section: Background and Related Workmentioning

confidence: 99%

In‐network block repairing for erasure coding storage systems

Xia

Guo

Cheng

2019

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…For instance, star couplers enable nonblocking unicast, multicast, and broadcast traffic delivery directly in the optical domain and can be added to a wavelength-routing architecture to support a rich set of traffic patterns. Multicast traffic involves the simultaneous dissemination of the same information copy to a group of recipients and constitutes a major portion of data center traffic (e.g., MapReduce) [14]- [21]. As well, advanced coordinated multipoint (CoMP) transmission techniques in radio access networks call for efficient optical multicasting from a central office (node) to a group of cooperating radio heads [22], [23].…”

Section: Introductionmentioning

confidence: 99%

Overcoming the Switching Bottlenecks in Wavelength-Routing, Multicast-Enabled Architectures

Keykhosravi

Rastegarfar

Peyghambarian

et al. 2019

J. Lightwave Technol.

View full text Add to dashboard Cite

Modular optical switch architectures combining wavelength routing based on arrayed waveguide grating (AWG) devices and multicasting based on star couplers hold promise for flexibly addressing the exponentially growing traffic demands in a cost-and powerefficient fashion. In a default switching scenario, an input port of the AWG is connected to an output port via a single wavelength. This can severely limit the capacity between broadcast domains, resulting in interdomain traffic switching bottlenecks. An unexplored solution to this issue is to exploit multiple AWG free spectral ranges (FSRs), i.e., to set up multiple parallel connections between each pair of broadcast domains. In this paper we i) study, for the first time, the influence of the FSR count on the throughput of a multistage switching architecture and ii) propose a generic and novel analytical framework to estimate the blocking probability. We assess the accuracy of our analytical results via Monte Carlo simulations. Our study points to significant improvements with a moderate increase in the number of FSRs. We show that an FSR count beyond four results in diminishing returns. Furthermore, to investigate the trade-offs between the network-and physical-layer effects, we conduct a cross-layer analysis, taking into account pulse amplitude modulation (PAM) and rate-adaptive forward error correction (FEC). We illustrate how the effective bit rate per port increases with an increase in the number of FSRs.Index Terms-Arrayed waveguide grating (AWG), blocking probability, coupler, free spectral range (FSR), multicast, physical layer, scheduling, switch architecture.

show abstract

“…Multicast scheduling has been studied in some types of networks. The bounded congestion multicast scheduling algorithm in Guo et al is a typical example, which can achieve bounded congestion and efficient bandwidth utilization under an arbitrary sequence of multicast flow requests. In Guo et al, the authors explored a server redundancy mechanism to reduce the multicast cost under the condition of nonblocking in fat‐tree DCNs.…”

Section: Introductionmentioning

confidence: 99%

Blocking cost‐driven multicast scheduling in fat‐tree data center networks

Guo

Yang

2017

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

Summary Multicast traffic scheduling in data center networks can not only improve network efficiency but also save network resources. However, the existing multicast scheduling algorithms cannot appropriately schedule flows to achieve traffic load balance so that the network may occur heavy blocking. This prevents the full utilization of high degree of link parallelism and causes unpredictable reduction of network performance. To address the problem, in this paper, we propose a blocking cost‐driven multicast scheduling algorithm by using optimization theory in fat‐tree data center networks. In particular, a model of multicast traffic subnetwork is established on the basis of the blocking probability of available paths at next time slot, which can reflect the blocking characteristics of multicast network and predict network state at next time slot. With the multicast blocking model, we derive the minimum blocking probability of multicast subnetworks, denoted as blocking cost. In addition, the algorithm can select the multicast subnetwork with minimum blocking cost to transfer multicast flows. Time complexity analysis and simulation results demonstrate the effectiveness and efficiency of our proposed multicast scheduling algorithm for different network traffic intensities.

show abstract

On-Line Multicast Scheduling with Bounded Congestion in Fat-Tree Data Center Networks

Cited by 60 publications

References 23 publications

In‐network block repairing for erasure coding storage systems

In‐network block repairing for erasure coding storage systems

Overcoming the Switching Bottlenecks in Wavelength-Routing, Multicast-Enabled Architectures

Blocking cost‐driven multicast scheduling in fat‐tree data center networks

Contact Info

Product

Resources

About