Efficient Large Flow Detection over Arbitrary Windows

Among the various strategies proposed for reducing or eliminating bias against small flows (in the presence of large flows), most require to identify and distinguish between small and large flows, besides having to track the ongoing sizes of all flows. Though these solutions do improve the response times of small flows (with negligible affect on the response times of large flows), they are not scalable with increasing traffic. In this context, we propose a new spike-detecting AQM that exploits TCP property in detecting large ‘spikes’, and hence large flows, from which packets are dropped, and importantly, only at times of congestion. We discuss two such AQM policies using spike-detection for improving the performance of small flows: one that drops packets deterministically, and other that drops packets randomly. We show, using simulations, by comparing a number of metrics, that these new policies, in particular the one that drops packets randomly, out-performs not only the traditional drop-tail buffer with FCFS server, but also the RED policy as well as a size-based scheduler (proposed specifically for improving the response time of small flows). The improvement in performance becomes more revealing in scenario where the router buffer is small (less than one-tenth of the bandwidth-delay-product).

Online social networks (OSN) have become a prevalent communication tool with a user scale of billions. Users send messages and share photos and videos to their friends every day. As a consequence, a large amount of flows are produced by file storage and backup in the back end. Most flows are deadline intensive, and the scale of flow account is large; these characteristics create new challenges for flow scheduling in data center networks. In this study, we conclude the flow distribution of OSN based on previous studies. OSN flow can be classified into a large amount of bursty flow and a limited amount of large flow. The bursty mice flows are mostly critical real-time requirements, which support instant messages and routing information; the large flows seldom require real-time transmission, because they mainly support file transferring and updating. A mixed flow causes bottlenecks in rack switches easily, thereby resulting in deadline loss for real-time applications and lower throughput. The study divides flow based on different metrics and provides priority-based flow scheduling for OSN datacenter (PFO). PFO allocates rate for different flows on the basis of flow size and deadline information. Our objective is to decrease the average completion time for bursty flows and ensure a high throughput. The analysis and evaluation shows that the proposed scheduling increases the goodput more than DCTCP and D2TCP, and decreases the fraction of missed deadlines.

The vast majority of application traffic in modern data center networks (DCNs) can be classified into two categories: throughput-sensitive large flows and latency-sensitive small flows. These two types of flows have the conflicting requirements on link buffer occupancy. Existing data transfer proposals either do not fully utilize the path diversity of DCNs to improve the throughput of large flows, or cannot achieve a controllable link buffer occupancy to meet the low latency requirement of small flows. Aiming to balance throughput with latency, we develop the eXplicit MultiPath (XMP) congestion control scheme for DCNs. XMP comprises two components: the BOS algorithm brings link queue buffers consumed by large flows under control, while the TraSh algorithm is responsible for shifting traffic from more congested paths to less congested ones, thus achieving high throughput. We implemented XMP and evaluated its performance on traffic shifting, fairness, goodput, buffer occupancy and link utilization by conducting comprehensive experiments and simulations. The results show that XMP outperforms existing schemes and achieves a reasonable tradeoff between throughput and latency.

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.

Single-machine scheduling with time windows and earliness/tardiness penalties

It has been well recognized that identifying very large flows (i.e., elephants) in a network traffic stream is important for a variety of network applications ranging from traffic engineering to anomaly detection. However, we found that many of these applications have an increasing need to monitor not only the few largest flows (say top 20), but also all of the medium-sized flows (say top 20,000). Unfortunately, existing techniques for identifying elephant flows at high link speeds are not suitable and cannot be trivially extended for identifying the medium-sized flows. In this work, we propose a hybrid SRAM/DRAM algorithm for monitoring all elephant and medium-sized flows with strong accuracy guarantees. We employ a synopsis data structure (sketch) in SRAM to filter out small flows and preferentially sample medium and large flows to a flow table in DRAM. Our key contribution is to show how to maximize the use of SRAM and DRAM available to us by using a SRAM/DRAM hybrid data structure that can achieve more than an order of magnitude higher SRAM efficiency than previous methods. We design a quantization scheme that allows our algorithm to read just enough from the sketch at SRAM speed, without sacrificing much estimation accuracy. We provide analytical guarantees on the accuracy of the estimation and validate these by means of trace-driven evaluation using real- world packet traces..

In current networks, large flows often co-exist with small flows and consume a significant proportion of the network bandwidth. This imposes a challenge on network resource management. Various hybrid switching architectures have been proposed in recent years aiming to address this challenge and improve efficiency and resource utilization. However, there is still a need for means for design and dimensioning of resources among multiple switching planes. In this work, we first argue that the resource allocation in hybrid switching systems is fact the interaction between three components: resource allocation, traffic partitioning and performance (or cost) constraints. We then introduce a framework called Blocking-LOss Curve (BLOC) to address the resource allocation problem in hybrid switching systems. BLOC can capture the behavior of hybrid switching systems under different resource allocation strategies. This framework also allows us to identify the feasible region for resource allocation, within which performance metrics like the packet loss and request blocking requirements can be satisfied. An optimal resource allocation strategy may be found based on the feasible region. We also discuss the application of BLOC in hybrid switching Data Center networks.

In this work, we present results of flow visualisation, pressure gradient measurements, and liquid holdup measurements for air-water flow without and with surfactants in vertical pipes with diameters of 34 mm, 50 mm, and 80 mm. The surfactants cause the formation of foam. This foam has a larger volume and a smaller density than the liquid. The larger volume results in a larger pressure gradient at large gas flow rates. At small gas flow rates, the lower density of the foam causes the transition between the regular annular flow regime and the irregular churn flow regime to shift to lower gas flow rates. As a result foam reduces the pressure gradient and the liquid holdup at small gas flow rates. Surfactants reduce the pressure gradient more effectively for thinner liquid films at the wall; therefore, they are more effective for small pipe diameters and small liquid flow rates.

The Siwalik hill of Nepal lies in between the two major thrusts; Main boundary Thrust and Main Frontal Thrust. Thrusting and continuous erosion of material from surface causes several landslides and debris flow problem in the Siwalik region. Debris flow commonly occurs after landslide during heavy rainfall in the steep slope. This paper described about the topographical and geological controls on debris flow occurring in gullies in the Siwaliks of the Malai River catchment of mid-Western Nepal. The length of gully, length of debris channel, area of debris channel, area of catchment of debris flow, area of gully without debris flow were found using Google earth pro, a free version online database. Similarly, a form factor (F), average gradient of stream, slope area ratio and a topographical factor (T) were calculated. Lithology and geological structure were studied in the field. The relation between each factor was identified.
 Gully having larger debris flow event had T value greater than 0.01 and that having small debris flow event had T value less than 0.01. Gully without debris flow had T value less than 0.001. Gully having debris flow had F more than 0.1 and that free from debris flow had F less than 0.1. Both topographic and form factors were found greater than 0.1 at the hanging wall of the Malai Thrust, where large size debris flows were encountered. The T and F values obtained from the Middle and Upper Siwaliks were greater than 0.1.The number of debris flow events and large debris flows were found high the in the Middle and the Upper Siwaliks, and hanging wall of the Malai Thrust.

DESPITE advances in experimental technology, measurement of capillary blood flow and tissue perfusion has remained an enigma. Most current methods involve an estimate of small vessel flow based upon a measurement of flow in a major artery leading to the organ in question. These methods necessitate the use of anesthesia and operative manipulation, both of which alter the dynamics of small vessel flow. Furthermore, techniques utilizing measurement of large vessel flow fail to differentiate nutritive (capillary) flow and nonnutritive flow through arteriovenous pathways. A method is needed which will measure capillary blood flow in the nonanesthetized animal without operative trauma. This report describes such a method. Potassium Chloride K 42 as a Determinant of Capillary Blood Flow Sapirstein and co-workers have shown that, from 9 to 64 seconds following a single intravenous injection of trace amounts of potassium chloride K 42, the amount of 42 K in a given

For more than a decade our group has been involved in the study of the epidemiology of peripheral vascular disease, using both traditional clinical assessments as well as highly accurate and reliable non-invasive measurements (Criqui 1985a, b, c, 1989). The non-invasive measures used are able to assess hemodynamics in the lower extremities from the largest (common femoral) to the smallest (arteries less than 2 mm in diameter) components of the peripheral arterial circulation. In evaluating circulation by these various non-invasive tests, we found that three groups of patients could be differentiated: a group with reduced flow in large vessels but normal small vessel (<2 mm diameter) flow; a group with normal large vessel flow but reduced small vessel flow; and a group with reduced flow in both large and small vessels. We labelled the first group large vessel peripheral arterial disease (LV-PAD), and the second group isolated small vessel (ISV-PAD). However, while we could characterise this third group as having LV-PAD, we were uncertain as to whether their reduced small vessel flow was due to the small vessels per se, or simply reduced flow due to proximal large vessel obstruction.

Design goals and reasons for the selection of a 60-cps square-wave electromagnetic blood flowmeter are presented. Detailed description of the flowmeter is given with emphasis on demodulator techniques. Complete performance data of step function response, stability, and performance with small and large flows are shown graphically. One application of the flowmeter and analog integrator is given showing response to a periodic hydraulic flow.

Emissions of toxic gases and particulates from photovoltaic module R&amp;D and manufacturing facilities may need treatment to meet current or anticipated regulations. In this study currently available pollution‐control options (e.g. filters, scrubbers, adsorption columns, combustion devices) are reviewed for their suitability to treat emissions (e.g. HCl, HF, Cl2, H2Se, SiH2, SiH2Cl2, AsH3, PH3, Cd and As compounds) from photovoltaic cell laboratories and production facilities. For small emission flow rates (e.g. &gt; 100 l min−−1) and small to moderate particulate loads (e.g. &lt;2 mg l−−1), on‐line filters can economically collect fumes and other particulates generated in the deposition chambers, whereas adsorption columns can very effectively treat toxic hydrates (e.g. AsH3, PH3, H2Se) at concentrations up to 1% (up to 10% for very small flow rates, e.g. &lt;11 min−−1). For large emission flow rates, (e.g. &gt; 100 1 min−−1), very low concentrations of hydrates can still be treated effectively with adsorption. Combined control of particulates and hydrates can be achieved by a bag‐house followed by a multistage scrubber; when the concentration of hydrates is relatively high, a thermal oxidation unit should precede the bag‐house scrubber for effective control. Finally, multicomponent mixtures can be treated either by wet scrubbing or combustion, depending on the mixture composition.

Rectangular sharp-crested weir used for flow measurement in flume is limited by the minimum head since head- discharge relation of clinging flow is different from free flow. Hydraulic characteristics of clinging flow under low head are studied via experiments associated with simulation in six rectangular sharp-crested weirs with different geometric parameters. The results show that head-discharge relation is stable under clinging flow; head relation between clinging and free flow is linearly correlated at the same discharge in a bistable regime, existing discharge formula of free flow can be applied to clinging flow by means of head transform; the pressure on downstream weir plate is negative, much less than atmosphere pressure, and flow regime can be judged based on the pressure. It is feasible to automatically measure large and small flow using the simplest weir without head limitation.

Results on flow from the ALICE Collaboration