ADITCA 2019 Abstracts

Full Papers Short Papers
Full Papers
Paper Nr: 2
Title:

A Holistic Approach to Proximity Marketing
Authors:

Dimitris Poulopoulos and Athina Kalampogia
Abstract: Proximity marketing, as a promotional technique, can benefit shopping centres and malls in terms of revenue, as well as customer loyalty, by analysing the customers’ data and using their profiles to better address their needs and target advertising and promotional campaigns. To this end, retailers exploit cellular technology to send marketing messages to users’ mobile devices, that are near a specific area of one of its stores. With more than six billion mobile phones in the hands of consumers today, every consumer with a smartphone is potentially susceptible to a proximity marketing campaign.
Download
Paper Nr: 3
Title:

Towards Intra-Datacentre High-Availability in CloudDBAppliance
Authors:

Luis Ferreira, Fábio Coelho, Ana N. Alonso and José Pereira
Abstract: In the context of the CloudDBAppliance (CDBA) project, fault tolerance and high-availability are provided in layers: within each appliance, within a data centre and between data centres. This paper presents the proposed replication architecture for providing fault tolerance and high availability within a data centre. This layer configuration, along with specific deployment constraints require a custom replication architecture. In particular, replication must be implemented at the middleware-level, to avoid constraining the backing operational database. This paper is focused on the design of the CDBA Replication Manager along with an evaluation, using micro-benchmarking, of components for the replication middleware. Results show the impact, on both throughput and latency, of the replication mechanisms in place.
Download
Paper Nr: 4
Title:

uCash: ATM Cash Management as a Critical and Data-intensive Application
Authors:

Terpsichori-Helen Velivassaki, Panagiotis Athanasoulis and Panagiotis Trakadas
Abstract: Distributed cloud databases wrapped with streaming analytics modules provide nowadays quick response to increasingly demanding real-time applications, relying on fast analytical and online processing of enormous amounts of data or very frequently updated. However, time-critical applications, dealing with sensitive data, typically run on mainframes, cannot fully benefit from existing solutions. Such applications can be found in Banking, Financial Services and Insurance (BFSI) industry, one notable being the ATM cash management. The paper presents uCash, an ATM cash management system, running on top cloud analytics appliances, which can be hosted insite. The proposed system allows data processing and Key Performance Indicators (KPIs) calculation and communication among diverse actors, resulting in highly efficient cash management over large ATM networks.
Download
Paper Nr: 5
Title:

A NUMA Aware SparkTM on Many-cores and Large Memory Servers
Authors:

François Waeselynck and Benoit Pelletier
Abstract: Within the scope of the CloudDBAppliance project, we investigate how Apache SparkTM can leverage a many cores and large memory platform, with a scale up approach as opposed to the commonly used scale out one: that is, the approach is to deploy a spark cluster to few large servers with many cores (up to several hundreds) and large memory (up to several tera-byte), rather than spreading it on many vanilla servers, and to stack several Spark executor processes per cluster node when running a job. It requires to cope with the non-uniform memory access within such servers, so we inculcate NUMA awareness to Spark, that provides a smart and application transparent placement of executor processes. We experiment it on a BullSequanaTM S series platform with the Intel HiBench suite benchmark and compare performance where NUMA awareness is off or on.
Download
Paper Nr: 6
Title:

The End of the Nightly Batch: How In-memory Computing and the Cloud Are Transforming Business Processes
Authors:

Antoine Chambille and Gaëlle Guimezanes
Abstract: This paper illustrates our conviction that data processing relying on nightly batches is a thing of the past. It shows how a complex analytics platform for huge amounts of data can be created from scratch in a matter of minutes on the cloud, thus negating the need for dedicated on-premise machines that would do the job slower at a higher cost. We also present the CloudDBAppliance project, where we aim at building a Cloud database platform that will be available as a service, and will integrate the operational database with the tools for further data analysis, eliminating the hassle of exporting the data for processing.
Download
Paper Nr: 7
Title:

Real Time Financial Risk Monitoring as a Data-intensive Application
Authors:

Petra Ristau and Lukas Krain
Abstract: This paper will examine the possibility of real-time risk calculations within the financial services industry. Due to regulatory standards, this paper will focus mainly on the calculations of value-at-risk (VaR) and expected shortfall (ES). Their computation currently requires simplified theory in order to be done within real-time. This demonstrates a real-world disadvantage to investment professionals since they need to comply with regulatory requirements when doing real-time decisions without knowing the accurate risk numbers at any one time. Within the CloudDBAppliance project, we designed an architecture that shall make real-time risk monitoring possible using cloud computing and a fast analytical processing platform.
Download
Paper Nr: 8
Title:

NUMA-aware Deployments for LeanXcale Database Appliance
Authors:

Ricardo Jiménez-Peris, Francisco Ballesteros, Pavlos Kranas, Diego Burgos and Patricio Martínez
Abstract: In this paper we discuss NUMA awareness for the LeanXcale database appliance being developed in cooperation with Bull-Atos in the Bull Sequana in the context of the CloudDBAppliance European project. The Bull Sequana is a large computer than in its maximum version can reach 896 cores and 140 TB of main memory. Scaling up in such a large computer with a deep NUMA hierarchy is very challenging. In this paper we discuss how LeanXcale database can be deployed in NUMA architectures such as the one of the Bull Sequana and what aspects have been taking into account to maximize efficiency and to introduce the necessary flexibility in the deployment infrastructure.
Download
Paper Nr: 9
Title:

Data Streaming for Appliances
Authors:

Marta Patiño and Ainhoa Azqueta
Abstract: Nowadays many applications require to analyse the continuous flow of data produced by different data sources before the data is stored. Data streaming engines emerged as a solution for processing data on the fly. At the same time, computer architectures have evolved to systems with several interconnected CPUs and Non Uniform Memory Access (NUMA), where the cost of accessing memory from a core depends on how CPUs are interconnected. This paper presents UPM-CEP, a data streaming engine designed to take advantage of NUMA architectures. The preliminary evaluation using Intel HiBench benchmark shows that NUMA aware deployment improves performance.
Download
Short Papers
Paper Nr: 1
Title:

Parallel Streaming Implementation of Online Time Series Correlation Discovery on Sliding Windows with Regression Capabilities
Authors:

Boyan Kolev, Reza Akbarinia, Ricardo Jimenez-Peris, Oleksandra Levchenko, Florent Masseglia, Marta Patino and Patrick Valduriez
Abstract: This paper addresses the problem of continuously finding highly correlated pairs of time series over the most recent time window and possibly use the discovered correlations to select features for training a regression model for prediction. The implementation builds upon the ParCorr parallel method for online correlation discovery and is designed to run continuously on top of the UPM-CEP data streaming engine through efficient streaming operators.
Download