Search

Offloading of tasks to the cloud is one of the approaches to improve the performance of distributed applications. When monetary constraints are present, selection of the tasks to be offloaded becomes important in order to ensure efficient use of the available cloud resources. This becomes a challenge for large scale distributed applications as the decisions on offloading have to be made locally at the nodes without an exact global view of the system. In our earlier work, we modeled this challenge as a new class of formal problems termed cloudy knapsack problem and derived some theoretical bounds on the solution space for worst case task sequences. In many real world applications, the task sequences have inherent patterns which can be exploited to improve offloading. In this work, we propose a cloud offloading algorithm that exploits these patterns through offline and online learning. Experimental evaluation using realistic datasets for a cloud-assisted peer-to-peer search case study reveals that the proposed solution performs close to a hypothetical omniscient offloading algorithm having a complete view of the system. The proposed cloud-assisted peer-to-peer search engine provides a cost-effective approach to address scalability bottleneck in peer-to-peer search engines.

We adapt a polynomial-time approximation algorithm for computing the canonical basis of implications to approximately compute frequent implications, also known as exact association rules. To this end, we define a suitable notion of approximation that takes into account the frequency of attribute subsets and show that our algorithm achieves a desired approximation with high probability. We experimentally evaluate the proposed algorithm on several artificial and real-world data sets.

Developing an event-driven applications for network management systems(NMS) requires data-wrangling operations to be performed in a data pipeline with low latency and high throughput. Data stream processing engines(SPEs) enables the pipeline to feature extract complex data in near real time. We present a low latency and high throughput data pipeline that performs operations such as data cleansing, filtering, aggregations and join of streams on pipeline stages for event-driven machine learning applications. When the data stream has imbalance in arrival of records, a standalone system handle it ineffectively that reduces the performance of data pipeline. In this work, the system powered by SPEs handle the imbalance by scaling out resources and degree of parallelism such that performance is not degraded.

As the volume of data generated is growing at an unprecedented rate, it becomes important to analyze this data in real-time. To handle the huge volume of data streaming at a high velocity, we not only require powerful machines but also means to distribute the computation involved on the multiple machines. There are several open-source distributed stream processing frameworks such as Apache {Storm, Flink, Spark} and Confluent Kafka for building real-time machine learning applications. Prior works benchmarked some of these platforms using low-level operations like filters, joins, windowed computations etc. Our work includes benchmarking these popular frameworks for their applicability to classical machine learning models: Online K-Means, Online Linear Regression and Online Logistic Regression. We study the following quantitative metrics of evaluation: throughput, latency, CPU, and memory usage. The experiments were conducted in both standalone and clusters setups to determine the scalability of the models. This study will help system designers choose the right model and the right framework, given a specific configuration on streaming data.

Today, batch processing frameworks like Hadoop MapReduce are difficult to scale to multiple clouds due to latencies involved in inter-cloud data transfer and synchronization overheads during shuffle-phase. This inhibits the MapReduce framework from guaranteeing performance at variable load surges without over-provisioning in the internal cloud (IC). We propose BStream, a cloud bursting framework for MapReduce that couples stream-processing in the external cloud (EC) with Hadoop in the internal cloud (IC). Stream processing in EC enables pipelined uploading, processing and downloading of data to minimize network latencies. We use this framework to meet job deadlines. BStream uses an analytical model to minimize the usage of EC. We propose different checkpointing strategies that overlap output transfer with input transfer/processing and simultaneously reduce the computation involved in merging the results from EC and IC. Checkpointing further reduces job completion time. We experimentally compare BStream with other related works and illustrate performance benefits due to stream processing and checkpointing strategies in EC. Lastly, we characterize the operational regime of BStream. � 2013 IEEE.

SUMMARY It is difficult to express the parallelism present in complex computations by using existing higher level abstractions such as MapReduce and Dryad. These computations include applications from wide variety of domains, like Artificial Intelligence, Decision Tree Algorithms, Association Rule Mining, Recommender Systems, Graph Algorithms, Clustering Algorithms, Compute Intensive Scientific Workflows, Optimization Algorithms, and so forth. Their execution graphs introduce new challenges in terms of programmer expressibility and runtime performance such as iterative and recursive computations, shared communication model, and so forth. We propose an extension to MapReduce, called Generate-Map-Reduce (GMR), targeted towards modeling these applications. GMR introduces a new Generate abstraction into the MapReduce framework that captures recursive computations. The runtime also supports iterative jobs and a distributed communication model by using shared data structures. We illustrate recursive computations with GMR by modeling complex applications such as simulated annealing, A* search, and adaptive quadrature computation that require recursive spawning of new tasks to handle variable degree of parallelism. GMR runtime supports caching of common data across iterations in memory and local disks. We illustrate how this caching helps in achieving significant speedup for iterative computations by modeling k-means clustering. Copyright © 2013 John Wiley & Sons, Ltd.

Objectives: In the past decades, X-rays have been used widely for diagnosis in dentistry. However, it is well known that ionizing radiation causes damage (including single- and double-strand breaks) to deoxyribonucleic acid (DNA) and DNA-protein crosslinks, and induces cellular death. Therefore, outlining the cytogenetic effects induced by X-ray is necessary to identify the degree of cancer risk and minimize potential risks to patients and clinicians. Materials and methods: Cytogenetic biomonitonng studies focusing on oral mucosa cells in individuals exposed to dental X-ray were reviewed. Results: Dental X-ray can induce DNA damage and cytotoxicity in oral mucosa cells. Conclusion: These results will contribute to a better understanding of X-ray-induced effects upon the cellular system in individuals continually exposed to known genotoxic/cytotoxic agents.

Oral submucous fibrosis (OSMF) is one of the most poorly understood and unsatisfactorily treated disease. In this study, combined treatment modalities were tried based on the grading and staging of OSMF, and outcomes of treatment were noticed. In addition to this, visceral organ involvement evidenced by systemic fibrosis has been assessed to see, if OSMF is a part of systemic spectrum of disease.

Juvenile ossifying fibroma considered as a variant of ossifying fibroma by some authors is rare. Its aggressive behavior necessitates early detection and proper surgical management. It may mimic different pathologic entities of jaw at different stages. A case report of juvenile aggressive ossifying fibroma in a 12-year-old girl which was diagnosed, investigated and treated successfully is presented here.

Cloud-assisted approaches for both peer-to-peer systems and mobile apps require optimized use of elastic cloud resources. Due to budget constraints, a subset of the tasks has to be selected for offloading considering context parameters like device battery level and task variability. This leads to the challenging problem of context-sensitive task scheduling on elastic resources with a limited global view, which is not addressed by existing works. We identify a new class of formal problems called cloudy knapsack problems to effectively model the same. Abstracting out the problem formally can spur future independent works related to different variants of the problem and corresponding bounds and optimal algorithms. We illustrate the global view related issues through simulations, identify some theoretical bounds for a variant of cloudy knapsack problems and discuss several open problems.

Job completion in MapReduce framework depends upon the slowest running reduce task. Inordinate time gap among the completion points of reduce tasks delays a job significantly. Synchronization in reduce task completion not only completes a job faster but also increases resource utilization. Completion time of reduce tasks depends mainly upon the data load given to each of them. Partitioning of keys plays a major role in load balancing input data to the reduce tasks. Range partitioning using sampling technique has been proposed in the literature over hash based partitioning to provide better key distribution. But data sampling is also susceptible to data/partitioning skew since only a small portion of the actual data is scanned. It is very difficult to design a perfect load balancing partitioner without having prior information about the entire data. Chisel, allows data skew to occur during partitioning but handles it dynamically by creating multiple partial reducers and one global reducer for the skewed partition. But this technique suffers from the global reducer becoming a bottleneck when the amount of reducer output is huge. We propose an efficient range partitioning technique which eliminates the use of global reducer and yields better performance. We also discuss various heuristics to implement range partitioning. Chisel++ achieved 2 times improvement in terms of job completion time over Chisel in our experiments.

Skew mitigation has been a major concern in distributed programming frameworks like MapReduce. It is becoming more prominent with the increasing complexity in user requirements and computation involved. We present Chisel, a self-regulating skew detection and mitigation policy for MapReduce applications. The novelty of the approach is that it involves no scanning or sampling of input data to detect skew and hence incurs low overhead, provides better resource utilization and maintains output order and file structure. It is also transparent to the users and can be used as a plugin whenever required. We use Hadoop to implement our skew handling policies. Chisel implements two skew handling policies for mitigating skew. It does late skew detection for map operators i.e at the last wave of map execution, where skewed maps are selected on the basis of remaining time to complete. More maps are created dynamically over remaining data per block. An early skew detection i.e before starting shuffle phase, is done for reduce operator. This prevents the expensive shuffle and sort phases from delaying skew detection and job completion time. Multiple reducers are created per skewed partition, each shuffling data from a subset of total maps and starts processing it when their portion of maps are over. They need not wait for the completion of all the maps. Therefore, the barrier between map and reduce phase no longer remains a constraint for effective resource utilization. Chisel additionally implements an online job profiler to determine the start point of reduce tasks and also modifies the capacity scheduler to distribute reduce tasks evenly in the cluster. Chisel significantly decreases the overall execution time of jobs and increases resource utilization. Improvement depends directly upon the availability of resources in the cluster and skewness in the job.

Optimizing ordered throughput not only improves the system efficiency but also makes cloud bursting transparent to the user. This is critical from the perspective of user fairness in customer-facing systems, correctness in stream processing systems, and so on. In this paper, we consider optimizing ordered throughput for near real-time, data-intensive, independent computations using cloud bursting. Intercloud computation of data-intensive applications is a challenge due to large data transfer requirements, low intercloud bandwidth, and best-effort traffic on the Internet. The system model we consider is comprised of two processing stages. The first stage uses cloud bursting opportunistically for parallel processing, while the second stage (sequential) expects the output of the first stage to be in the same order as the arrival sequence. We propose three scheduling heuristics as part of an autonomic cloud bursting approach that adapt to changing workload characteristics, variation in bandwidth, and available resources to optimize ordered throughput. We also characterize the operational regimes for cloud bursting as stabilization mode versus acceleration mode, depending on the workload characteristics like the size of data to be transferred for a given compute load. The operational regime characterization helps in deciding how many instances can be optimally utilized in the external cloud. � 1976-2012 IEEE.

The practice of computing across two or more data centers separated by the Internet is growing in popularity due to an explosion in scalable computing demands and pay-as-you-go schemes offered on the cloud. While cloud-bursting is addressing this process of scaling up and down across data centers (i.e. between private and public clouds), offering service level guarantees, is a challenge for inter-cloud computation, particularly for best-effort traffic and large files. The parallel workload we address is real-time and involves inter-cloud processing and analysis of images and documents. In our production printing domain, dedicated processing/network resources are cost-prohibitive. Further, the problem is exacerbated by data intensive computing - we encounter huge file sizes atypical of intercloud parallel processing. To address these problems we propose three flavors of autonomic cloud-bursting schedulers that offer probabilistic guarantees on service levels required by customers (such as speed-up and queue sequence preservation) by adapting to changing workload characteristics, variation in bandwidth and available resources. In particular, these opportunistic schedulers use a quadratic response surface model for processing time in concert with a time-of-day dependent bandwidth predictor to increase the throughput and utilization while simultaneously reducing out-of-sequence completions for a document processing workload.

Burning mouth disorder (BMD) is a burning or stinging sensation affecting the oral mucosa, lips and/or tongue, in the absence of clinically visible mucosal lesions. There is a strong female predilection, with the age of onset being approximately 50 years. Affected patients often present with multiple oral complaints, including burning, dryness and taste alterations. The causes of BMD are multifactorial and remain poorly understood. Recently, there has been a resurgence of interest in this disorder with the discovery that the pain of burning mouth syndrome (BMS) may be neuropathic in origin and originate both centrally and peripherally. The most common sites of burning are the anterior tongue, anterior hard palate and lower lip, but the distribution of oral sites affected does not appear to affect the natural history of the disorder or the response to treatment. BMS may persist for many years. This article provides updated information on BMS and presents a new model, based on taste dysfunction, for its pathogenesis.