PFO: Priority-based flow scheduling for online social network datacenter

We study the problem of flow scheduling in data center networks. Using speed scaling, our aim is to find an online scheduling algorithm that minimizes the total energy consumption of the network by determining both the transmission order and rates of the arriving flows while providing a strict flow deadline guarantee. Observing the superlinear property of link power consumption, the key challenge is in constantly determining the minimum transmission rate for “delay-tolerable” flows without any priori knowledge. To leverage the flow arrival pattern, we propose a probability-based flow prediction model to capture the uncertainty of the network flows. Based on the prediction model, we propose a tunable online flow scheduling algorithm to solve the online flow scheduling problem effectively. By introducing a scaling factor on bandwidth allocation, this algorithm allows us to conduct arbitrary trade-offs between the conservative and aggressive behaviors in terms of energy conser- vation. The effectiveness of the proposed algorithm is validated through rigorous theoretical analysis and further confirmed by extensive numerical simulations.

Data Center Networks (DCNs) suffer from the congestions caused by unbalanced traffic distributions. Nevertheless, it is difficult to address such a challenge with the existent Ethernet-based DCN architecture. Wireless networking has been considered as a viable approach with a high potential to tackle the problem due to its flexibility. This paper presents Wireless Link Scheduling for Data Center Networks (WLS-DCN) to provide a feasible wireless DCN. More specifically, we propose a wireless DCN architecture as well as a scheduling mechanism. In our wireless scheduling, two different optimization objectives are considered, with one targeting the unbalanced traffic distribution and one maximizing the total network utility, under the constraints of limited wireless resources and co-channel interference. We propose two heuristics for the two problems and carry out extensive simulation study to validate the performance of our design. Our results demonstrate that WLSDCN can significantly improve the performance of DCN.

Numerous scheduling approaches have been proposed to improve user experiences in a data center network (DCN) by reducing flow completion time (FCT). Mimicking the shortest job first (SJF) has been proved to be the prominent way to improve FCT. To do so, some approaches require flow size or completion time information in advance, which is not possible in scenarios like HTTP chunk transfer or database query response. Some information-agnostic schemes require involving end-hosts for counting the number of bytes sent. We present Longer Stay Less Priority (LSLP), an information-agnostic flow scheduling scheme, like Multi-Level Feedback Queue (MLFQ) scheduler in operating systems, that aims to mimic SJF using P4 switches in a DCN. LSLP considers all the flows as short flows initially and assigns them to the highest priority queue, and flows get demoted to the lower priority queues over time. LSLP estimates the active time of a flow by leveraging the state-of-the-art P4 switch’s programmable nature. LSLP estimates the active time of a group of new flows that arrive during a time interval and assigns their packets to the highest priority. At the beginning of the next time interval, arriving packets of old flows are placed one priority lower except for those already in the lowest priority queue. Therefore, short flows can be completed in the few higher priority queues while long flows are demoted to lower priority queues. We have evaluated LSLP via a series of tests and shown that its performance is comparable to the existing scheduling schemes.

Traffic in today’s data center networks grows rapidly. Traditional routing and forwarding protocols in data center usually rely on static hashing, and may cause avoidable bottlenecks. We present a genetic algorithm to make better use of the resources. Our goal is to minimize the maximum single link utilization of the network. And results show that our algorithm performs better than a recent greedy algorithm.

Online Social Networks (OSN) have attracted a lot of end-users in the last few years. The success of Facebook, Twitter, Snapchat leads to a huge increase of data transiting in the network of telecom operators. To face to this increase, network operators have to find solutions. The Named Data Networking (NDN) paradigm, focusing on interest-based requests for content, is a good possible networking solution, having native caching features. Furthermore, it has been detected that a large number of people follow people who are in a close vicinity to them, except for very popular people. However, the current networking behaviour of the OSN applications do not take this into consideration. In this paper, we propose a locality-aware NDN-based delivery for OSNs in order to improve the quality for endusers (reduced delivery time) as well as to reduce the network load (and peering traffic) for network operators and the server load for the application provider. We modelled our solution for evaluating the networking impact and the tests we have performed proved the efficiency of our proposed solution, enabling to largely reduce the data delivery path as well as the peering traffic.

The advent and continued growth of large data centers has led to much interest in switch architectures that can economically meet the high capacities needed for interconnecting the thousands of servers in these data centers. Various multilayer architectures employing thousands of switches have been proposed in the literature. We make use of the observation that the traffic in a data center is a mixture of relatively static and rapidly fluctuating components, and develop a combined scheduler for both these components using a generalization of the load-balanced scheduler. The presence of the known static component introduces asymmetries in the ingress-egress capacities, which preclude the use of a load-balanced scheduler as is. We generalize the load-balanced scheduler and also incorporate an opportunistic scheduler that sends traffic on a direct path when feasible to enhance the overall switch throughput. Our evaluations show that this scheduler works very well despite avoiding the use of a central scheduler for making packet-by-packet scheduling decisions.

In the data center network, mixed flows which have contradictory service requirements are transmitted simultaneously. Switches usually aggregate similar flows to the same queue after flow classification and schedule them using fair queuing and its extension schemes capable of flow isolation. These schemes implement diverse bandwidth allocation by assigning different weights to queues. Existing solutions rely on rich statistics such as packet arrival rate and delay to realize dynamic bandwidth allocation. However, many statistics are difficult to accurately measure or even obtain in real switches due to resource limitations. Providing differentiated services for mixed flows under such restrictions is still a challenge. To solve this issue, this paper proposed a reinforcement learning-based scheduling optimization (RLSO) mechanism. First, mixed flows scheduling is modeled as the Markov decision process (MDP) and Q-learning is used to find the approximate optimal solution with a few statistics. Second, the solution space is compressed to reduce the complexity of the algorithm and adapt to the limited performance of switches. Finally, the performance of the proposed mechanism is evaluated on a hardware testbed with workloads that include coarsegrained and fine-grained flows. The results show that RLSO can effectively schedule mixed flows.

Cloud applications have reshaped the model of services and infrastructure of the Internet. Search engines, social networks, content delivery and retail and e-commerce sites belong to this group of applications. An important element in the architecture of data centers where these applications run is the communication infrastructure, commonly known as data center networks (DCNs). A critical challenge DCNs have to address is the processing of the traffic of cloud applications, which due to its properties is essentially different to the traffic of other Internet applications. In order to improve the responsiveness and throughput of applications, DCNs should be able to prioritize short flows (a few kB) over long flows (several MB). However, given the time and space variations the traffic presents, the information about flow sizes is not available in advance in order to plan the flow scheduling. There has been a wealth of solutions developed in this space, and prior work includes flow scheduling mechanisms optimizing for a specific workload but fall short when workloads are not known in advance, or comprise a collection of applications changing dynamically. In this paper, we present an adaptable mechanism called Adaptable Workload-Agnostic Flow Scheduling (AWAFS). It is an adaptable approach that can adjust in an agnostic way the scheduling configuration of DCN forwarding devices. This agnostic adjustment contributes to reduce the Flow Completion Time (FCT) of those short flows, representing around 85% of the traffic handled by cloud applications. AWAFS operates by observing the traffic and detecting statistical properties that provide a hint to adapt the scheduling parameters. Our evaluation results based on simulation show that AWAFS can reduce the average FCT of short flows between 16.9% and 45.2% when compared to the best existing agnostic non-adaptable solution, without inducing starvation on long flows. Indeed, it can provide improvements as high as 39% for long flows. Additionally, AWAFS can improve the FCT for short flows in scenarios with high heterogeneity in the traffic present in the network, with a reduction up to 5% for the average FCT and 15% for the tail FCT.

In data center networks (DCNs), the flow size is a crucial factor to schedule the traffic, as small flows are usually latency-sensitive (e.g., web search from users) while large flows are not (e.g., backup between two sites). Some existing flow scheduling schemes try to schedule flows based on the prior knowledge of flows, such as flow size and deadline. However, it may be unpractical to obtain the prior information in some scenarios. In this paper, we propose START to minimize the flow completion time (FCT) and guarantee the priority of small flows in DCNs. We control the priorities of packets by leveraging the multiple priority queues based on the amount of data a flow has sent. Switches schedule flows according to their priorities. In addition, to schedule the flows more intelligently, we adjust the threshold of each priority dynamically according to the network loads. A prototype of START has been implemented with ns2 simulations. START is compatible with legacy TCP/IP stacks and deployable with existing commodity switches. The results show that START significantly outperforms existing flow scheduling schemes DCTCP and L2DCT up to 40% and 45% for small flows in changing network status. In some scenarios, START can even achieve better performance than pFabric.

Resource competition and conflicts in datacenter networks (DCNs) are frequent and intense. They become inevitable when mixing elephant and mice flows on shared transmission paths, resulting in arbitration between throughput and latency and performance degradation. We propose a novel flow scheduling scheme, Freeway, that leverages on path diversity in the DCN topology to guarantee, simultaneously, mice flow completion within deadline and high network utilization. Freeway adaptively partitions the available paths into low latency and high throughput paths and provides different transmission services for each category. A M/G/1-based model is developed to theoretically obtain the highest value of average delay over the path that will guarantee for 99% of mice flows their completion time before the deadline. Based on this bound, Freeway proposes a dynamic path partitioning algorithm to adjust dynamically with varying traffic load the number of low latency and high throughput paths. While mice flows are transmitted over low latency paths using a simple equal cost multiple path (ECMP) scheduling, Freeway load balances elephant flows on different high-throughput paths. We evaluate Freeway in a series of simulation on a large scale topology and use real traces. Our evaluation results show that Freeway significantly reduces the mice flows completion time within deadlines, while achieving remarkable throughput compared with current schemes. It is remarkable that Freeway does not need any change of DCN switch fabrics or scheduling algorithms and can be deployed easily on any generic datacenter network with switches implementing VLANs and trunking.

With the rapid advancement of big data processing and artificial intelligence (AI), data center networks (DCNs) must deliver more efficient resource management and data transmission mechanisms. Unfortunately, due to the significant differences in bandwidth requirements, transmission patterns, and temporal characteristics across various traffic types in DCNs (such as short flows, long flows, and bursty flows), traditional buffer allocation strategies fail to adapt flexibly to these disparities. In this paper, we propose Enhanced Buffer Management (EBM), a novel buffer-sharing scheme designed for scenarios that require higher performance from DCNs. Unlike prior approaches, EBM employs a multi-level flow identification and adaptive threshold adjustment mechanism to enhance the flexibility and efficiency of buffer management under varying traffic conditions. Specifically, EBM first performs coarse-grained and fine-grained classification of traffic based on packet size, inter-arrival interval, and other flow characteristics. It then applies an improved threshold computation function to allocate buffer space differentially across traffic classes while maintaining allocation smoothness. Our evaluation results demonstrate that EBM significantly improves performance under realistic workloads. For instance, it reduces the 99th percentile Flow Completion Time (FCT) slowdown by 32.7% for short flows in the web-search workload and by 45.1% for incast flows in the hadoop workload, all without sacrificing overall throughput.

Data centers are integral part of cloud computing that support Web services, online social networking, data analysis, computation intensive applications and scientific computing. They require high performance components for their inter-process communication, storage and sub-communication systems. The performance bottleneck that used to be the processing power has now been shifted to communication speed within data centers. The performance of a data center, in terms of throughput and delay, is directly related to the performance of the underlying internal communication network. In this paper, we introduce an analytical model that can be used to evaluate the underlying network architecture in data centers. The model can further be used to develop simulation tools that extend the scope of performance evaluation beyond what it can be achieved by the theoretical model in terms of various network topologies, different traffic distributions, scalability, and load balancing. While the model is generic, we focus on its implementation for fat-tree networks that are widely used in data centers. The theoretical results are compared and validated with the simulation results for several network configurations. The results of this analysis provide a basis for data center network design and optimization.

It is our great pleasure to welcome you to the 2012 ACM CoNEXT Student Workshop in Nice, France. This year's workshop continues the workshop's tradition of being an engaging and interactive forum for graduate students to present their early research results and receive feedback from the networking research community. In addition to 32 student submissions, this year's workshop features keynote lectures by Paul Francis from the Max Planck Institute for Software Systems and Dina Papagiannaki from Telefonica Research. The call for papers attracted submissions from students from over 60 institutions spread across 15 universities across five continents. The submissions spanned all areas of networking research including routing protocols, data center networks, software-defined networking, wireless and cellular networks, network security, Internet measurement, online social networks, and future Internet architectures. Most submissions received three reviews from our program committee members. The final program consists of 32 accepted papers with authors drawn from over 35 universities and research labs. This year's workshop introduced two new features over previous workshops. First, we allowed students to indicate if their work forms the core of their Doctoral Dissertation. Second, selected accepted submissions will be given the opportunity to give an "elevator pitch" at the main conference to gain broader visibility and feedback beyond the workshop.

Many networking and security applications can benefit from exact detection of large flows over arbitrary windows (i.e. any possible time window). Existing large flow detectors that only check the average throughput over certain time period cannot detect bursty flows and are therefore easily fooled by attackers. However, no scalable approaches provide exact classification in one pass. To address this challenge, we consider a new model of exactness outside an ambiguity region, which is defined to be a range of bandwidths below a high-bandwidth threshold and above a low-bandwidth threshold. Given this new model, we propose a deterministic algorithm, EARDet, that detects all large flows (including bursty flows) and avoids false accusation against any small flows, regardless of the input traffic distribution. EARDet monitors flows over arbitrary time windows and is built on a frequent items finding algorithm based on average frequency. Despite its strong properties, EARDet has low storage overhead regardless of input traffic and is surprisingly scalable because it focuses on accurate classification of large flows and small flows only. Our evaluations confirm that existing approaches suffer from high error rates (e.g., misclassifying 1% of small flows as large flows) in the presence of large flows and bursty flows, whereas EARDet can accurately detect both at gigabit line rate using a small amount of memory that fits into on-chip SRAM.

Many data center applications perform rich and complex tasks (e.g., executing a search query or generating a user's news-feed). From a network perspective, these tasks typically comprise multiple flows, which traverse different parts of the network at potentially different times. Most network resource allocation schemes, however, treat all these flows in isolation -- rather than as part of a task -- and therefore only optimize flow-level metrics. In this paper, we show that task-aware network scheduling, which groups flows of a task and schedules them together, can reduce both the average as well as tail completion time for typical data center applications. To achieve these benefits in practice, we design and implement Baraat, a decentralized task-aware scheduling system. Baraat schedules tasks in a FIFO order but avoids head-of-line blocking by dynamically changing the level of multiplexing in the network. Through experiments with Memcached on a small testbed and large-scale simulations, we show that Baraat outperforms state-of-the-art decentralized schemes (e.g., pFabric) as well as centralized schedulers (e.g., Orchestra) for a wide range of workloads (e.g., search, analytics, etc).