With the growth of data volumes and variety of Internet applications, data centers (DCs) have become an efficient and promising infrastructure for supporting data storage, and providing the platform for the deployment of diversified network services and applications (e.g., video streaming, cloud computing). These applications and services often impose multifarious resource demands (storage, compute power, bandwidth, latency) on the underlying infrastructure. Existing data center architectures lack the flexibility to effectively support these applications, which results in poor support of QoS, deployability, manageability, and defence against security attacks. Data center network virtualization is a promising solution to address these problems. Virtualized data centers are envisioned to provide better management flexibility, lower cost, scalability, better resources utilization, and energy efficiency. In this paper, we present a survey of the current state-of-the-art in data center networks virtualization, and provide a detailed comparison of the surveyed proposals. We discuss the key research challenges for future research and point out some potential directions for tackling the problems related to data center design.

Data Center Network Virtualization: A Survey

Today's data centers face extreme challenges in providing low latency. However, fair sharing, a principle commonly adopted in current congestion control protocols, is far from optimal for satisfying latency requirements.We propose Preemptive Distributed Quick (PDQ) flow scheduling, a protocol designed to complete flows quickly and meet flow deadlines. PDQ enables flow preemption to approximate a range of scheduling disciplines. For example, PDQ can emulate a shortest job first algorithm to give priority to the short flows by pausing the contending flows. PDQ borrows ideas from centralized scheduling disciplines and implements them in a fully distributed manner, making it scalable to today's data centers. Further, we develop a multipath version of PDQ to exploit path diversity.Through extensive packet-level and flow-level simulation, we demonstrate that PDQ significantly outperforms TCP, RCP and D3 in data center environments. We further show that PDQ is stable, resilient to packet loss, and preserves nearly all its performance gains even given inaccurate flow information.

/pdf/finishing-flows-quickly-with-preemptive-scheduling-40ushwlf4x.pdf

Finishing flows quickly with preemptive scheduling

An ideal datacenter network should provide several properties, including low median and tail latency, high utilization (throughput), fair allocation of network resources between users or applications, deadline-aware scheduling, and congestion (loss) avoidance. Current datacenter networks inherit the principles that went into the design of the Internet, where packet transmission and path selection decisions are distributed among the endpoints and routers. Instead, we propose that each sender should delegate control---to a centralized arbiter---of when each packet should be transmitted and what path it should follow. This paper describes Fastpass, a datacenter network architecture built using this principle. Fastpass incorporates two fast algorithms: the first determines the time at which each packet should be transmitted, while the second determines the path to use for that packet. In addition, Fastpass uses an efficient protocol between the endpoints and the arbiter and an arbiter replication strategy for fault-tolerant failover. We deployed and evaluated Fastpass in a portion of Facebook's datacenter network. Our results show that Fastpass achieves high throughput comparable to current networks at a 240x reduction is queue lengths (4.35 Mbytes reducing to 18 Kbytes), achieves much fairer and consistent flow throughputs than the baseline TCP (5200x reduction in the standard deviation of per-flow throughput with five concurrent connections), scalability from 1 to 8 cores in the arbiter implementation with the ability to schedule 2.21 Terabits/s of traffic in software on eight cores, and a 2.5x reduction in the number of TCP retransmissions in a latency-sensitive service at Facebook.

/pdf/fastpass-a-centralized-zero-queue-datacenter-network-2qs0v5wk8q.pdf

Fastpass: a centralized "zero-queue" datacenter network

While today's data centers are multiplexed across many non-cooperating applications, they lack effective means to share their network. Relying on TCP's congestion control, as we show from experiments in production data centers, opens up the network to denial of service attacks and performance interference. We present Seawall, a network bandwidth allocation scheme that divides network capacity based on an administrator-specified policy. Seawall computes and enforces allocations by tunneling traffic through congestion controlled, point to multipoint, edge to edge tunnels. The resulting allocations remain stable regardless of the number of flows, protocols, or destinations in the application's traffic mix. Unlike alternate proposals, Seawall easily supports dynamic policy changes and scales to the number of applications and churn of today's data centers. Through evaluation of a prototype, we show that Seawall adds little overhead and achieves strong performance isolation.

/pdf/sharing-the-data-center-network-4jkf9sevrl.pdf

Sharing the data center network

Data center networks (DCNs) form the backbone infrastructure of many large-scale enterprise applications as well as emerging cloud computing providers. This paper describes the design, implementation and evaluation of OSA, a novel Optical Switching Architecture for DCNs. Leveraging runtime reconfigurable optical devices, OSA dynamically changes its topology and link capacities, thereby achieving unprecedented flexibility to adapt to dynamic traffic patterns. Extensive analytical simulations using both real and synthetic traffic patterns demonstrate that OSA can deliver high bisection bandwidth (60%-100% of the non-blocking architecture). Implementation and evaluation of a small-scale functional prototype further demonstrate the feasibility of OSA.

/pdf/osa-an-optical-switching-architecture-for-data-center-1hcikxf8tb.pdf

OSA: an optical switching architecture for data center networks with unprecedented flexibility

TCP incast congestion happens in high-bandwidth and low-latency networks, when multiple synchronized servers send data to a same receiver in parallel [15]. For many important data center applications such as MapReduce[5] and Search, this many-to-one traffic pattern is common. Hence TCP in-cast congestion may severely degrade their performances, e.g., by increasing response time. In this paper, we study TCP incast in detail by focusing on the relationship among TCP throughput, round trip time (RTT) and receive window. Different from the previous approach to mitigate the impact of incast congestion by a fine grained timeout value, our idea is to design an ICTCP (Incast congestion Control for TCP) scheme at the receiver side. In particular, our method adjusts TCP receive window proactively before packet drops occur. The implementation and experiments in our testbed demonstrate that we achieve almost zero timeout and high goodput for TCP incast.

/pdf/ictcp-incast-congestion-control-for-tcp-in-data-center-233rp50aa8.pdf

ICTCP: Incast Congestion Control for TCP in data center networks

Transport Control Protocol (TCP) incast congestion happens in high-bandwidth and low-latency networks when multiple synchronized servers send data to the same receiver in parallel. For many important data-center applications such as MapReduce and Search, this many-to-one traffic pattern is common. Hence TCP incast congestion may severely degrade their performances, e.g., by increasing response time. In this paper, we study TCP incast in detail by focusing on the relationships between TCP throughput, round-trip time (RTT), and receive window. Unlike previous approaches, which mitigate the impact of TCP incast congestion by using a fine-grained timeout value, our idea is to design an Incast congestion Control for TCP (ICTCP) scheme on the receiver side. In particular, our method adjusts the TCP receive window proactively before packet loss occurs. The implementation and experiments in our testbed demonstrate that we achieve almost zero timeouts and high goodput for TCP incast.

ICTCP: incast congestion control for TCP in data-center networks

Data center networks encode locality and topology information into their server and switch addresses for performance and routing purposes. For this reason, the traditional address configuration protocols such as DHCP require huge amount of manual input, leaving them error-prone.In this paper, we present DAC, a generic and automatic Data center Address Configuration system. With an automatically generated blueprint which defines the connections of servers and switches labeled by logical IDs, e.g., IP addresses, DAC first learns the physical topology labeled by device IDs, e.g., MAC addresses. Then at the core of DAC is its device-to-logical ID mapping and malfunction detection. DAC makes an innovation in abstracting the device-to-logical ID mapping to the graph isomorphism problem, and solves it with low time-complexity by leveraging the attributes of data center network topologies. Its malfunction detection scheme detects errors such as device and link failures and miswirings, including the most difficult case where miswirings do not cause any node degree change.We have evaluated DAC via simulation, implementation and experiments. Our simulation results show that DAC can accurately find all the hardest-to-detect malfunctions and can autoconfigure a large data center with 3.8 million devices in 46 seconds. In our implementation, we successfully autoconfigure a small 64-server BCube network within 300 milliseconds and show that DAC is a viable solution for data center autoconfiguration.

/pdf/generic-and-automatic-address-configuration-for-data-center-2tjtoibtvm.pdf

Generic and automatic address configuration for data center networks

Data center networks encode locality and topology information into their server and switch addresses for performance and routing purposes. For this reason, the traditional address configuration protocols such as DHCP require a huge amount of manual input, leaving them error-prone. In this paper, we present DAC, a generic and automatic Data center Address Configuration system. With an automatically generated blueprint that defines the connections of servers and switches labeled by logical IDs, e.g., IP addresses, DAC first learns the physical topology labeled by device IDs, e.g., MAC addresses. Then, at the core of DAC is its device-to-logical ID mapping and malfunction detection. DAC makes an innovation in abstracting the device-to-logical ID mapping to the graph isomorphism problem and solves it with low time complexity by leveraging the attributes of data center network topologies. Its malfunction detection scheme detects errors such as device and link failures and miswirings, including the most difficult case where miswirings do not cause any node degree change. We have evaluated DAC via simulation, implementation, and experiments. Our simulation results show that DAC can accurately find all the hardest-to-detect malfunctions and can autoconfigure a large data center with 3.8 million devices in 46 s. In our implementation, we successfully autoconfigure a small 64-server BCube network within 300 ms and show that DAC is a viable solution for data center autoconfiguration.

DAC: generic and automatic address configuration for data center networks

TCP incast congestion happens in high-bandwidth and low-latency networks, when multiple synchronized servers send data to a same receiver in parallel . For many important data center applications such as MapReduce and Search, this many-to-one traffic pattern is common. Hence TCP incast congestion may severely degrade their performances, e.g., by increasing response time. In this paper, we study TCP incast in detail by focusing on the relationship among TCP throughput, round trip time (RTT) and receive window. Different from the previous approach to mitigate the impact of incast congestion by a fine grained timeout value, our idea is to design an ICTCP (Incast congestion Control for TCP) scheme at the receiver side. In particular, our method adjusts TCP receive window proactively before packet drops occur. The implementation and experiments in our testbed demonstrate that we achieve almost zero timeout and high goodput for TCP incast.

/pdf/ictcp-incast-congestion-control-for-tcp-4f7wfrkb0d.pdf

Zhenqian Feng

Papers

ICTCP: Incast Congestion Control for TCP in data center networks

ICTCP: incast congestion control for TCP in data-center networks

Generic and automatic address configuration for data center networks

DAC: generic and automatic address configuration for data center networks

ICTCP: Incast Congestion Control for TCP