This paper presents BCube, a new network architecture specifically designed for shipping-container based, modular data centers. At the core of BCube is its server-centric structure, where servers with multiple ports connect to layers COTS (commodity off-the-shelf) mini-switches. Servers act as not only end hosts, but also relay nodes each other. supports various bandwidth-intensive applications by speeding-up one-to-one, one-to-several, and one-to-all traffic patterns, providing high...

A fundamental challenge in data center networking is how to efficiently interconnect an exponentially increasing number of servers. This paper presents DCell, a novel network structure that has many desirable features for networking. DCell recursively defined structure, which high-level constructed from low-level DCells and at the same level are fully connected with one another. scales doubly as node degree increases. fault tolerant since it does not have single point failure its distributed...

IEEE 802.11 medium access control (MAC) is proposed to support asynchronous and time bounded delivery of radio data packets in infrastructure ad hoc networks. The basis the WLAN MAC protocol a distributed coordination function (DCF), which carrier sense multiple with collision avoidance (CSMA/CA) binary slotted exponential back-off scheme. Since has its own characteristics that are different from other wireless protocols, performance reliable transport over needs further study. This paper...

This paper presents BCube, a new network architecture specifically designed for shipping-container based, modular data centers. At the core of BCube is its server-centric structure, where servers with multiple ports connect to layers COTS (commodity off-the-shelf) mini-switches. Servers act as not only end hosts, but also relay nodes each other. supports various bandwidth-intensive applications by speeding-up one-to-one, one-to-several, and one-to-all traffic patterns, providing high...

Over the past one and half years, we have been using RDMA over commodity Ethernet (RoCEv2) to support some of Microsoft's highly-reliable, latency-sensitive services. This paper describes challenges encountered during process solutions devised address them. In order scale RoCEv2 beyond VLAN, designed a DSCP-based priority flow-control (PFC) mechanism ensure large-scale deployment. We addressed safety brought by PFC-induced deadlock (yes, it happened!), transport livelock, NIC PFC pause frame...

The goal of data center networking is to interconnect a large number server machines with low equipment cost, high and balanced network capacity, robustness link/server faults. It well understood that, the current practice where servers are connected by tree hierarchy switches cannot meet these requirements (Fares et al., 2008 Guo 2008). In this paper, we explore new server-interconnection structure. We observe that commodity used in today's centers usually come two built-in Ethernet ports,...

Rate adaptation is critical to the system performance of wireless networks. Typically, rate considered as a MAC layer mechanism in IEEE 802.11. Most previous work relies only on frame losses infer channel quality, but performs poorly if are mainly caused by interference. Recently SNR- based schemes have been proposed, most them not studied real environment. In this paper, we first conduct systematic measurement-based study confirm that general SNR good prediction tool for and identify two...

Shipping-container-based data centers have been introduced as building blocks for constructing mega-data centers. However, it is a challenge on how to interconnect those containers together with reasonable cost and cabling complexity, due the fact that center can hundreds or even thousands of aggregate bandwidth among easily reach tera-bit per second. As new inner-container server-centric network architecture, BCube [9] interconnects servers inside container provides high support typical...

It has been a challenging problem to support VoIP-type delay sensitive applications in an 802.11 wireless LAN, because the standard handoff procedure implemented many current products occurs deem unacceptable VoIP users. To reduce this delay, we have developed fast scheme called Proactive Scan. employs two new techniques. The first is decouple time-consuming channel scan from actual handoff, and eliminate by doing early interleaving it with ongoing traffic non-intrusive way. second technique...

Clos-based networks including Fat-tree and VL2 are being built in data centers, but existing per-flow based routing causes low network utilization long latency tail. In this paper, by studying the structural properties of VL2, we propose a per-packet round-robin algorithm called Digit-Reversal Bouncing (DRB). DRB achieves perfect packet interleaving. Our analysis simulations show that, compared with random-based load-balancing algorithms, results smaller bounded queues even when traffic load...

As one of the fundamental infrastructures for cloud computing, data center networks (DCN) have recently been studied extensively. We currently use pure software-based systems, FPGA based platforms, e.g., NetFPGA, or OpenFlow switches, to implement and evaluate various DCN designs including topology design, control plane routing, congestion control. However, approaches suffer from high CPU overhead processing latency; platforms are difficult program incur cost; focuses on functions at...

Being a part of IEEE project 802, the 802.11 medium access control (MAC) is used to support asynchronous and time bounded delivery radio data packets. It proposed that distributed coordination function (DCF), which uses carrier sense multiple with collision avoidance (CSMA/CA) binary slotted exponential backoff, be basis WLAN MAC protocols. This paper proposes throughput enhancement mechanism for DCF by adjusting contention window (CW) resetting scheme. Moreover, an analytical model based on...

In this paper, we first identify several challenges in designing a joint channel assignment and routing (JCAR) protocol heterogeneous multiradio multichannel multihop wireless networks (M <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> WNs) using commercial hardware [e.g., IEEE 802.11 Network Interface Card (NIC)]. We then propose novel software solution, called Layer 2.5 JCAR, which resides between the MAC layer layer. JCAR jointly...

Recent advances in multi-radio multi-channel transmission technology have the potential of substantially improving system capacity multi-hop wireless networks. While previous work has primarily focused on link level protocol design, this paper we investigate achievable performance gain, by jointly optimizing routing and scheduling a multichannel network. We formulate optimization under deterministic model, seek to minimize overall activation time use satisfy given end-to-end traffic demands...

There have been some serious concerns about the TCP performance in data center networks, including long completion time of short flows competition with flows, and congestion due to incast. In this paper, we show that a properly tuned instant queue length based Explicit Congestion Notification (ECN) at intermediate switches can alleviate both problems. Compared previous work, our approach is appealing as it be supported on current commodity simple parameter setting does not need any...

Many data center network DCN applications require explicit routing path control over the underlying topologies. In this paper, we present XPath, a simple, practical and readily-deployable way to implement control, using existing commodity switches. At its core, XPath explicitly identifies an end-to-end with ID leverages two-step compression algorithm pre-install all desired paths into IP TCAM tables of Our evaluation implementation show that scales large DCNs is readily-deployable....

The goal of data-center networking is to interconnect a large number server machines with low equipment cost while providing high network capacity and bisection width. It well understood that the current practice where servers are connected by tree hierarchy switches cannot meet these requirements. In this paper, we explore new server-interconnection structure. We observe commodity used in today's data centers usually come two built-in Ethernet ports, one for connection other left backup...

Explicit Congestion Notification (ECN) is crucial for production datacenters, but current queue-length based ECN/RED implementation does not work with generic packet schedulers, leading to either degraded network performance or violated scheduling policies. In this paper, we first dive into issue and reveal that the invalidity of lies in difficulty measuring changing queue capacities under various schedulers traffic dynamics. Then present Time-based (TCN), a simple yet effective ECN...

The upcoming standard IEEE 802.11e will provide Quality of Service (QoS) support in Wireless Local Area Network (WLAN). contention based channel access method called Enhanced Distributed Channel Access (EDCA) is considered as the mandatory mode for Medium Control (MAC) 802.11e. This paper presents an accurate throughput analytical model EDCA saturated situation. suitable both basic and RTS/CTS mechanisms. It considers features such different Arbitration Inter-frame Space (AIFS) window...

Data center networks encode locality and topology information into their server switch addresses for performance routing purposes. For this reason, the traditional address configuration protocols such as DHCP require huge amount of manual input, leaving them error-prone.In paper, we present DAC, a generic automatic Address Configuration system. With an automatically generated blueprint which defines connections servers switches labeled by logical IDs, e.g., IP addresses, DAC first learns...