Performance Of Data Center Networks Research Articles

A Novel Conditional Connectivity and Hamiltonian Connectivity of BCube with Various Faulty Elements

BCube is one of the main data center networks because it has many attractive features. In practical applications, the failure of components or physical connections is inevitable. In data center networks in particular, switch failures are unavoidable. Fault-tolerance capability is one main aspect to measure the performance of data center networks. Connectivity, fault tolerance Hamiltonian connectivity, and fault tolerance Hamiltonicity are important parameters that assess the fault tolerance of networks. In general, the distribution of fault elements is scattered, and it is necessary to consider the distribution of fault elements in different dimensions. We research the fault tolerance of BCube when considering faulty switches and faulty links/edges that distribute in different dimensions. We also investigate the connectivity, fault tolerance Hamiltonian connectivity, and Hamiltonicity. This study better evaluates the fault-tolerant performance of data center networks.

Topology configuration scheme for accelerating coflows in a hyper-FleX-LION

To address the challenges that data center networks (DCNs) are currently facing, optical circuit switching in DCNs has been introduced and a number of architectures to improve the performance of DCNs in terms of port capacity, energy efficiency, and data transfer latency have been proposed. This led to a few all-optical interconnects (AOIs), among which Hyper-FleX-LION possesses the reconfigurability that can efficiently support various traffic patterns. In this work, we study the topology management of AOIs in Hyper-FleX-LION to efficiently schedule coflows in them so the average coflow completion time (CCT) can be minimized. We first propose time-efficient algorithms to address the provisioning of a single coflow and then extend the algorithms for multi-coflow scenarios. The simulation results demonstrate that AOIs in Hyper-FleX-LION can accelerate coflows better than the traditional AOIs based on optical cross-connects and achieve an acceleration of up to 5.6 × under the assumed situations. Our simulations also verify that, for coflow scheduling in Hyper-FleX-LION, our proposal can reduce the average CCT and significantly outperform benchmarks.

Survey on Traffic Management in Data Center Network: From Link Layer to Application Layer

Due to the explosive growth of all kinds of Internet services, data centers have become an irreplaceable and vital infrastructure to support this soaring trend. Compared with traditional networks, data center networks (DCNs) have unique features, such as high bandwidth, low latency, many-to-one communication mode, shallow buffered switches, and multi-root topology. These new characteristics pose a lot of challenges to previous network technics (e.g., Ethernet, Equal Cost Multi-path (ECMP), TCP), making them hard to adapt to DCNs and leading to severe performance degradation. In order to solve these challenges, DCNs have attracted a lot of attention from the industry and academia in recent years, and many new mechanisms in different layers are proposed to improve the transmission performance of data center networks. In the meantime, many surveys have emerged currently to introduce the current research of data center networks. However, previous surveys of DCNs mainly focus on only one specific network layer, making them difficult for readers to know about advanced researches on a holistic level. To help readers comprehend the current research progress of data center networks quickly, we employ a multi-layered top down taxonomy to classify the literature and propose several probable dimensions for future research in this area.

BL‐Hybrid: A graph‐theoretic approach to improving software‐defined networking‐based data center network performance

AbstractExisting data centers can handle a massive amount of data transmitted in a short time with minimum errors. Data center networks (DCNs) use cost‐effective, multipath topologies to distribute flows through alternative paths between core layers and hosting servers. Software‐defined networking (SDN) separates the control plane from the data plane to ease network management. To perform active load balancing among different paths, we address the challenges of load balancing in multipath DCNs. However, most DCNs' load balancing systems rely on a static approach that distributes flows among different paths sequentially, regardless of the path status. Consequently, some paths become congested, degrading the overall performance of the DCN in terms of throughput and delay. Increasing the number of requests worsens the congestion problem, causing some connections to become overwhelmed, thereby decreasing DCN efficiency. In this paper, we propose and evaluate the performance of a novel dynamic load‐balancing system named BL‐Hybrid, which uses the graph‐theoretical centrality of betweenness as a base metric for data forwarding. We perform a comprehensive comparative analysis of our proposed method relative to the round‐robin and least‐congested algorithms in different traffic scenarios, using the most popular SDN‐based DCNs, namely, fat tree, BCube, and DCell. Our evaluation results show that BL‐Hybrid outperforms the round‐robin and least‐congested schemes. It maximizes the average throughput by 4% and 11% and the average data transferred by 7% and 13%, respectively. Also, it reduces the average round‐trip time latency by 14% and 41%, respectively.

Achieving Efficient Routing in Reconfigurable DCNs

With the fast growth of cloud services and network scales, the heavy and highly dynamic traffic demands pose great challenges to the efficient traffic engineering in today's data center networks (DCNs) [21]. The DCN flows can be broadly classified into two main categories: delay-sensitive small flows (e.g., queries or realtime small messages) and throughput-sensitive large flows (e.g., the backup traffic). In general, more than 80% flows in data centers are small flows, while the majority of the traffic volume is contributed by the top 10% large flows [3, 7]. To handle the mixed traffic, today's data centers [1, 14] generally follow the tree-based topologies (e.g., fat-tree) and take the load-agnostic routing strategies based on random path selection (e.g., ECMP1) [14, 19]. Although it is applicable for routing small flows which are highly random, these strategies are likely to route several large flows through the same output link and lead to long-lived congestions [2, 8]. With the limited switch buffer occupied by large flows for a long time, small flows are reported to experience one order of magnitude larger delay, which compromises the performance of DCNs and makes the users suffer [3].

Load balance algorithm for an OPSquare datacenter network under real application traffic

To cope with the dynamic datacenter (DC) traffic matrix generated by multiple applications, load balancing algorithms can be deployed to improve the datacenter network (DCN) performance. OPSquare, as a promising optical DCN architecture, adopts fast optical switches to provide high bandwidth, low latency, large scalability, and cost and energy efficiency. Current load balancing algorithms are not optimal for improving the network performance of an OPSquare DCN. To optimize the network performance of OPSquare, a dedicated load balancing algorithm considering the peculiarity of OPSquare should be addressed. Moreover, network virtualization in today’s data DC creates new heterogeneous traffic patterns different from what so far have been reported in the literature. A synthetic virtualized DC traffic model is needed to properly evaluate the performance of the load balancing algorithms. In this paper, we provide a synthetic traffic model by capturing the traffic traces of the virtualized DC applications running in a DC. The captured traffic model is used in the simulator to quantitatively evaluate the network performance of an OPSquare DCN. Then we propose and evaluate the lowest path buffer (LPB) algorithm to optimize the OPSquare DCN performance. LPB considers the buffer occupation in the whole path (from the source to the destination) rather than the local buffer occupation. The network performance improvements of LPB are demonstrated under different central controller update periods, traffic destination distributions, and network sizes. Compared with the round-robin, DRILL, and LocalFlow load balancing algorithms, LPB could achieve 23.7%, 46%, and 32.1% less latency, respectively. Besides, LPB provides lower packet loss in comparison with round-robin, DRILL, and LocalFlow.

Demonstration of SDN-Enabled Hybrid Polling Algorithm for Packet Contention Resolution in Optical Data Center Network

We propose and experimentally demonstrate a Software-Defined Network (SDN) enabled Hybrid Polling Algorithm for packet Contention Resolution (HPACR) in optical Data Center Network (DCN) based on Fast Optical Switches (FOS). Fast Semiconductor Optical Amplifier (SOA)-based buffer-less FOS with nanoseconds reconfiguration time are employed for interconnecting Top-Of-the Racks (TORs) switches with distributed flow control. Under the realistic 10 Gb/s server traffic, the network performance of the DCN employing the proposed HPACR algorithm is experimentally assessed. Results show that the HPACR increases 13.6% throughput at the high load of 1.0 and decreases the 75% contention ratio for the load >0.5 with respect to the traditional packet contention resolution methods. Based on the experimental parameters, a simulation platform is built to numerically evaluate the HPACR performance of DCN with 64 TORs connecting 1280 servers. The results show that HPACR improves the system performance by increasing the throughput by 5% compared with existing algorithms, reducing the TOR-to-TOR latency by 18.5%, and decreasing the packet loss by 14.1% at the load of 1.

Pache: A Packet Management Scheme of Cache in Data Center Networks

The communication traffic among servers is a principal bottleneck for data center networks (DCNs), in which redundant traffic has a significant impact on the performance of DCNs. In this paper, we propose Pache , a distributed and efficient cache scheme to achieve traffic redundancy elimination in DCNs. In Pache, nodes cache packets to reduce bandwidth consumption and gain performance revenue. Each cache node manages the local cache and improves cache efficiency with counting bloom filter, hash table and Trie. Each cache node broadcasts the local cache information by a bloom filter to other servers, and servers have cache information tables (cache-table) to record them, which achieves a cache sharing mechanism. Moreover, Pache employs false positive information table (fp-table) to reduce the false positive rate caused by bloom filters. A server determines an original packet or the corresponding fingerprint packet to send by querying cache-table and fp-table. Cache node placement mechanism is also designed to select which node to cache a packet. For evaluating the performance of Pache, we execute extensive simulations and conduct a case study on Amazon EC2 platform from different aspects. The results show that Pache can eliminate about 40 percent redundant traffic on average, and only increase 10 percent runtime, and it is also a scalable and feasible cache scheme in DCNs.

Coflow Scheduling in the Multi-Resource Environment

In data centers, a lot of cluster computing applications follow the coflow working pattern. That is, a collection of flows between two groups of machines is semantically related. On the other hand, network function virtualization sufficiently improves the performance of data center networks. It however complicates the network environment by introducing many multi-function middleboxes each with multiple resources. Coflows encounter extremely different processing delays under diverse network functions. Prior coflow scheduling schemes are insufficient to guarantee the coflow completion time in the multi-resource environment. In this paper, we propose, model, and analyze the coflow scheduling problem in the multi-resource environment. We present a dedicated method, data rate guarantee for coflow (DRGC), to guarantee the data rate requirements of coflows in this situation. DRGC prioritizes the coflow scheduling sequence, assigns precise data rates for coflows, and deploys a packet scheduling algorithm at middleboxes to guarantee their transmissions. In our experiments, DRGC efficiently guarantees the completion times of coflows and supports 15% more workload, compared with other scheduling schemes.

Performance-Aware Energy Saving for Data Center Networks

Today’s data center networks (DCNs) tend to have tens to hundreds of thousands of servers that provide massive and sophisticated services. The architectural design of DCNs are usually over-provisioned for peak workloads and fault tolerance. Statistically, DCNs remain highly under-utilized, with typical utilization of around 30%. Network over-provisioning and under-utilization can be exploited for energy-saving. Most research efforts on DCN energy saving focus on how to save maximum energy but have little or no consideration to the performance of the residual network. Thus, the DCN performance can become degraded and the network left vulnerable to sudden traffic surges. In this paper, we have studied the energy-saving problem in DCNs while preserving network performance. The problem was formulated as mixed integer linear problem (MILP) solvable by CPLEX in order to minimize the energy consumed by DCN; meanwhile, safety threshold constraints for links utilization are met. To overcome CPLEX high computational time, a heuristic algorithm to provide practical and efficient solution for the MILP is introduced. The heuristic algorithm uses switches grouping and links consolidation to switch the traffic to a small number of network devices and turn off unused switches and links. Valiant load balancing is used to distribute the loads over active links. Simulation experiments using synthetic and real packet traces were conducted to validate the heuristic in terms of energy consumption and network performance. The results show that the heuristic can save up to 45% of the network energy and improve the average imbalance scores for links and switches by more than 50% with minimal effect on network performance.

Flow-Aware Routing and Forwarding for SDN Scalability in Wireless Data Centers

Software-defined networking (SDN) paradigm and high-speed wireless technologies have been widely adopted by industries to improve the performance of data center networks. In SDN, a switch commonly uses ternary content addressable memory (TCAM) to store rules for fast flow table lookup. However, due to the limited TCAM entries, rule replacement will occur frequently when traffic load in data centers is heavy. This issue will be more serious when flow characteristics in data centers are diverse. Although wireless transmissions can benefit for reducing TCAM usage, we observe that routing and forwarding designs without considering the flow characteristics will result in rule replacement storm in switches and significantly harm the SDN scalability for wireless data centers. As a result, switches will issue massive flow setup requests which cannot be processed by an SDN controller in real time. This paper studies the SDN scalability problem with the objective of minimizing the number of flow setup requests under the hybrid architecture with wired and wireless links. We propose a routing and forwarding algorithm with the consideration of flow characteristics to tackle the problem. A series of simulation results agrees our observations and shows that our algorithm outperforms two well-known solutions for wired data centers and three approaches for wireless data centers. The results also provide useful insights and guidelines for designing SDN-based wireless data centers.

MNDM: MEMS-based N-dimensional modular hybrid data center network

The rapid development of virtualization technology and cloud computing technology for the data center has brought the explosive growth of business volume and scale. The data center network is facing new challenges, such as ultra-high bandwidth requirements and energy-efficiency. The introduction of optical interconnect technology improves the performance of data center networks by virtue of less cost of devices, complexity, power consumption. However, there are still three major problems when adopting optical switching, which are constraint of centralized control system, non-modularity and poor scalability. Aiming at these problems, this letter proposes a MEMS-based N-dimensional modular hybrid data center network (MNDM) and corresponding MEMS scheduling strategy. We evaluate the performance of MNDM via OPNET simulation and results demonstrate that the MNDM architecture is stable in saturation point of latency and saturation point of throughput when the network size changes or network traffic changes.

Neural Network Meets DCN

The emerging optical/wireless topology reconfiguration technologies have shown great potential in improving the performance of data center networks. However, it also poses a big challenge on how to find the best topology configurations to support the dynamic traffic demands. In this work, we present xWeaver, a traffic-driven deep learning solution to infer the high-performance network topology online. xWeaver supports a powerful network model that enables the topology optimization over different performance metrics and network architectures. With the design of properly-structured neural networks, it can automatically derive the critical traffic patterns from data traces and learn the underlying mapping between the traffic patterns and topology configurations specific to the target data center. After offline training, xWeaver generates the optimized (or near-optimal) topology configuration online, and can also smoothly update its model parameters for new traffic patterns. We build an optical-circuit-switch-based testbed to demonstrate the function and transmission efficiency of our proposed solution. We further perform extensive simulations to show the significant performance gain of xWeaver, in supporting higher network throughput and smaller flow completion time.

Neural Network Meets DCN

The emerging optical/wireless topology reconfiguration technologies have shown great potential in improving the performance of data center networks. However, it also poses a big challenge on how to find the best topology configurations to support the dynamic traffic demands. In this work, we present xWeaver, a traffic-driven deep learning solution to infer the high-performance network topology online. xWeaver supports a powerful network model that enables the topology optimization over different performance metrics and network architectures. With the design of properly-structured neural networks, it can automatically derive the critical traffic patterns from data traces and learn the underlying mapping between the traffic patterns and topology configurations specific to the target data center. After offline training, xWeaver generates the optimized (or near-optimal) topology configuration online, and can also smoothly update its model parameters for new traffic patterns. The experiment results show the significant performance gain of xWeaver in supporting smaller flow completion time.

Flexible Architecture and Autonomous Control Plane for Metro-Scale Geographically Distributed Data Centers

Enterprises and cloud providers are moving away from deployment of large-scale data centers and towards small- to mid-sized data centers because of their lower implementation and maintenance costs. An optical metro network is used to provide connectivity among these data centers. The optical network requires flexibility on bandwidth allocation and various levels of Quality of Service to support the new emerging applications and services including the ones enabled by 5G. As a result, next generation optical metro networks face complex control and management issues that needs to be resolved with automation. We present a converged inter/intra data center network architecture with an autonomous control plane for flexible bandwidth allocation. The architecture supports both single-rate and multi-rate data planes with two types of physical layer connections ( Background and Dynamic ) that provide connections with strict bandwidth and latency requirements. We demonstrate autonomous bandwidth steering between two data centers on our prototype. Leveraging a simulation platform, we show up to 5 $\times$ lower transmission times and 25% less spectrum usage compared with the single-rate conventional non-converged networks. This is a significant improvement in the data center network performance and energy efficiency.

BEEP: Balancing Energy, Redundancy, and Performance in Fat-Tree Data Center Networks

BEEP is an energy-efficient strategy for data center networks (DCNs) based on a fat-tree topology. This strategy combines multipaths and the global overview offered by software-defined networks (SDNs) to balance energy efficiency, equipment redundancy levels, and DCN performance. BEEP is implemented in an OpenFlow network with multipaths enabled by using Multipath TCP (MPTCP). Experimental results in different variations of the fat-tree topology and redundancy levels show gains in energy savings of 30-45 percent. Additionally, BEEP achieves better available bandwidth usage, because more alternative paths are available.

CCDN: Content-Centric Data Center Networks

Data center networks continually seek higher network performance to meet the ever increasing application demand. Recently, researchers are exploring the method to enhance the data center network performance by intelligent caching and increasing the access points for hot data chunks. Motivated by this, we come up with a simple yet useful caching mechanism for generic data centers, i.e., a server caches a data chunk after an application on it reads the chunk from the file system, and then uses the cached chunk to serve subsequent chunk requests from nearby servers. To turn the basic idea above into a practical system and address the challenges behind it, we design content-centric data center networks (CCDNs), which exploits an innovative combination of content-based forwarding and location [Internet Protocol (IP)]-based forwarding in switches, to correctly locate the target server for a data chunk on a fully distributed basis. Furthermore, CCDN enhances traditional content-based forwarding to determine the nearest target server, and enhances traditional location (IP)-based forwarding to make high utilization of the precious memory space in switches. Extensive simulations based on real-world workloads and experiments on a test bed built with NetFPGA prototypes show that, even with a small portion of the server’s storage as cache (e.g., 3%) and with a modest content forwarding information base size (e.g., 1000 entries) in switches, CCDN can improve the average throughput to get data chunks by 43% compared with a pure Hadoop File System (HDFS) system in a real data center.

TCP Flow Classification and Bandwidth Aggregation in Optically Interconnected Data Center Networks

Optical functionality is being used to realize new data center architectures that minimize electronic switching overheads, pushing the processing to the edge of the network. A challenge in optically interconnected data center networks is to identify the large, bandwidthhungry flows (i.e., elephants) and efficiently establish the optical circuits. Moreover, the amount of optical resources to be provisioned during the network planning phase is a critical design problem. Flow classification accuracy affects the efficiency of optical circuits. Optical channel bandwidth, on the other hand, directly relates to the additive- increase, multiplicative-decrease congestion control mechanism of the transmission control protocol and affects the effective bandwidth allocated to elephant flows. In this paper, we simultaneously investigate the impact of two important mechanisms on data center network performance: traffic flow classification accuracy and optical bandwidth aggregation (i.e., the consolidation of several low-capacity channels into a single high-capacity one by employing advanced modulation formats for short-reach communications). We develop a discrete-event simulator for a hybrid data center network, enabling the tuning of flow classification parameters. Our simulations indicate that data center performance is highly sensitive to the aggregation level.We could observe up to a 74.5% improvement in network throughput only due to consolidating the optical channel bandwidth. We further noticed that the role of flow classification becomes more pronounced with higher bandwidth per wavelength as well as with more hot-spot traffic. Compared to a random classification benchmark, adaptive flow classification could lead to throughput improvements as large as 54.7%.

Towards Petabit/s All-Optical Flat Data Center Networks Based on WDM Optical Cross-Connect Switches with Flow Control

Scaling the capacity while maintaining low latency and power consumption is a challenge for hierarchical data center networks (DCNs) based on electrical switches. In this work we present a novel all-optical flat DCN architecture OPSquare that potentially addresses the scaling issues by employing parallel intra-/inter-cluster switching networks, distributed fast WDM optical cross-connect (OXC) switches, and a novel top-of-rack (ToR) switch architecture. The fast (nanoseconds) WDM OXC switches allow flexible switching capability in both wavelength and time domains and statistical multiplexing. The OPSquare DCN performance targeting Petabit/s capacity has been thoroughly assessed. First the packet loss, latency, throughput, and scalability are numerically investigated under realistic data center traffic model. Results indicate that when scaling the DCN size up to 1024 ToR switches, a packet loss ratio below 10−6 and a server end-to-end latency lower than 2 µs can be guaranteed at load of 0.3 with limited 20 kB buffer. Then, the experimental evaluation of the DCN by employing 4 × 4 OXC prototypes shows multi-path dynamic switching with flow control operation. The case deploying 32 × 32 and 64 × 64 OXC switches connecting 1024 and 4096 ToRs are emulated and limited performance degradation has been observed. The potential of switching higher-order modulation and waveband signals further proves the suitability of OPSquare architecture for Petabit/s and low-latency DCN by using optical switches with moderate radix.

Optimized Virtual Machine Placement with Traffic-Aware Balancing in Data Center Networks

Virtualization has been an efficient method to fully utilize computing resources such as servers. The way of placing virtual machines (VMs) among a large pool of servers greatly affects the performance of data center networks (DCNs). As network resources have become a main bottleneck of the performance of DCNs, we concentrate on VM placement with Traffic-Aware Balancing to evenly utilize the links in DCNs. In this paper, we first proposed a Virtual Machine Placement Problem with Traffic-Aware Balancing (VMPPTB) and then proved it to be NP-hard and designed a Longest Processing Time Based Placement algorithm (LPTBP algorithm) to solve it. To take advantage of the communication locality, we proposed Locality-Aware Virtual Machine Placement Problem with Traffic-Aware Balancing (LVMPPTB), which is a multiobjective optimization problem of simultaneously minimizing the maximum number of VM partitions of requests and minimizing the maximum bandwidth occupancy on uplinks of Top of Rack (ToR) switches. We also proved it to be NP-hard and designed a heuristic algorithm (Least-Load First Based Placement algorithm, LLBP algorithm) to solve it. Through extensive simulations, the proposed heuristic algorithm is proven to significantly balance the bandwidth occupancy on uplinks of ToR switches, while keeping the number of VM partitions of each request small enough.