An Efficient Exact Hypercube Model with Fully Dedicated Servers

The hypercube model is a useful descriptive tool to evaluate emergency services such as firefighters, police, and emergency medical services where geographically distributed vehicles and personnel serve users in emergencies. This study proposes an extension of the hypercube model to represent a dispatch policy in which advanced equipped servers serve solely life-threatening calls (called dedicated servers). The proposed approach is applied to two case studies of public medical emergency services in two different cities in Brazil and validated with discrete-event simulations. The computational experiments show the proposed model as more sensitive to respond to more life-threatening requests than other hypercube models in the literature, serving more of these requests under increased demand. In addition, to reduce the number of equilibrium equations and, consequently, the computational effort of the hypercube model, an aggregate model is shown based on the grouping of homogeneous servers located in the same station. The aggregation policy does not generate additional losses in the accuracy of the model, as shown through several experiments.

Towards a serverless CernVM-FS

The CernVM File System (CernVM-FS) provides a scalable and reliable software distribution and—to some extent—a data distribution service. It gives POSIX access to more than a billion binary files of experiment application software stacks and operating system containers to end user devices, grids, clouds, and supercomputers. Increasingly, CernVM-FSalso provides access to certain classes of data, such as detector conditions data, genomics reference sets, or gravitational wave detector experiment data. For most of the high- energy physics experiments, an underlying HTTP content distribution infrastructure is jointly provided by universities and research institutes around the world. In this contribution, we will present recent developments and future plans. For future developments, we put a focus on evolving the content distribution infrastructure and at lowering the barrier for publishing into CernVM-FS. Through so-called serverless computing, we envision cloud hosted CernVM-FS repositories without the need to operate dedicated servers or virtual machines. An S3 compatible service in conjunction with a content delivery network takes on data provisioning, replication, and caching. A chainof time-limited and resource-limited functions (so called “lambda function” or “function-as- a-service”) operate on the repository and stage the updates. As a result, any CernVM-FS client should be able to turn intoawriter, possession of suitable keys provided. For repository owners, we aim at providing cost transparency and seamless scalability from very small to very large CernVM-FS installations.

Detection of Erratic Behavior in Load Balanced Clusters of Servers Using a Machine Learning Based Method

With the explosion of the number of distributed applications, a new dynamic server environment emerged grouping servers into clusters, whose utilization depends on the current demand for the application. To provide reliable and smooth services it is crucial to detect and fix possible erratic behavior of individual servers in these clusters. Use of standard techniques for this purpose delivers suboptimal results. We have developed a method based on machine learning techniques which allows detecting outliers indicating a possible problematic situation. The method inspects the performance of the rest of the cluster and provides system operators with additional information which allows them to identify quickly the failing nodes. We applied this method to develop a Spark application using the CERN MONIT architecture and with this application, we analyzed monitoring data from multiple clusters of dedicated servers in the CERN data center. In this contribution, we present our results achieved with this new method and with the Spark application for analytics of CERN monitoring data.

Analysis of a Tollbooth Tandem Queue with Two-Class Customers and Two Heterogeneous Dedicated Servers

Do [European Journal of Operational Research, 247, 672–675] discussed a tollbooth tandem queue with two heterogeneous servers, where customers arrive the system according to a Poisson process and the service times at the two heterogeneous servers follow an exponential distribution. In this present paper, we aim to consider a tollbooth tandem queue with two-class customers and two heterogeneous dedicated servers, i.e., different types of customers should be served by their dedicated servers. Using matrix analytic method and spectral expansion method, we derive the steady state probabilities, which are then used for the computation of other performance measures and the Laplace–Stieltjes transform (LST) of the sojourn time of an arbitrary customer. Finally, some numerical examples are provided to illustrate the impact of several system parameters to the performance measures.

An effective system of electronic courses development and maintenance as a key problem of electronic information and educational environment development at university

This paper discusses one of the urgent problems of education quality management at the pedagogical university - the design and use of an effective system of electronic courses development and maintenance. This problem is analyzed taking into account the results of a pilot survey of pedagogical universities teachers, as well as the experience of deployment and modernization of electronic information and educational environment (EIEE) at various universities of the Samara Region. It is noted that in terms of classification of EIEE maturity levels, a significant number of regional pedagogical universities are at the second or the beginning of the third level, at which all EIEE opportunities from a quarter to a half of teachers and students are actively used. The paper presents the hypothesis substantiation: the transition to the next level of EIEE maturity can and should be carried out purposefully and systematically. At the same time the necessary condition for the growth of EIEE maturity is the qualification improvement of the majority of the university teaching staff, as well as the deployment of its own infrastructure (hardware and software), a possible version of which is also briefly described in the paper. An alternative is to rent several dedicated servers from a reliable service provider with data centers in the Russian Federation. The obtained materials can serve as a basis for the design of new approaches to the construction оf EIEE pedagogical universities in Russia.

Asymptotically Maximal Throughput in Tandem Systems with Flexible and Dedicated Servers

For a system of two tandem queues with a finite intermediate buffer, we study the asymptotically maximal throughput as the number of servers in each station grows to infinity. First, we study the system with only dedicated servers, and then we examine the system with both dedicated and flexible servers. We assume that travel times between the two stations are negligible and that each server can only work on one customer at a time. We model the system as a birth–death Markov process, derive a closed form solution for the stationary distribution, and quantify the maximal asymptotic normalized throughput as the number of servers grows to infinity. We show that flexibility is more favorable for small systems, and as the number of servers grows, the benefits of flexibility decrease. Furthermore, we prove that when the number of servers goes to infinity, there is no need of flexibility at all, as the maximum value of the throughput is obtained. However, flexibility still has a secondary beneficial effect — a little flexibility (on the order of the square root of the number of dedicated servers at each station) guarantees that all dedicated servers are busy and results in faster convergence to the maximum throughput.

Big Data Processing

This chapter is a description of MapReduce, which serves as a programming algorithm for distributed computing in a parallel manner on huge chunks of data that can easily execute on commodity servers thus reducing the costs for server maintenance and removal of requirement of having dedicated servers towards for running these processes. This chapter is all about the various approaches towards MapReduce programming model and how to use it in an efficient manner for scalable text-based analysis in various domains like machine learning, data analytics, and data science. Hence, it deals with various approaches of using MapReduce in these fields and how to apply various techniques of MapReduce in these fields effectively and fitting the MapReduce programming model into any text mining application.