RT Journal Article
T1 A gearbox model for processing large volumes of data by using pipeline systems encapsulated into virtual containers
A1 Santiago Durán, Miguel
A1 González Compean, J.L.
A1 Brinkmann, André
A1 Reyes Anastacio, Hugo G.
A1 Carretero Pérez, Jesús
A1 Montella, Raffaele
A1 Toscano Pulido, Gregorio
AB Software pipelines enable organizations to chain applications for adding value to contents (e.g., confidentially, reliability, and integrity) before either sharing them with partners or sending them to the cloud. However, the pipeline components add overhead when processing large volumes of data, which can become critical in real-world scenarios. This paper presents a gearbox model for processing large volumes of data by using pipeline systems encapsulated into virtual containers. In this model, the gears represent applications, whereas gearboxes represent software pipelines. This model was implemented as a collaborative system that automatically performs Gear up (by using parallel patterns) and/or Gear down (by using in-memory storage) until all gears produce uniform data processing velocities. This model reduces delays and bottlenecks produced by the heterogeneous performance of applications included in software pipelines. The new container tool has been designed to encapsulate both the collaborative system and the software pipelines into a virtual container and deploy it on IT infrastructures. We conducted case studies to evaluate the performance of when processing medical images and PDF repositories. The incorporation of a capsule to a cloud storage service for pre-processing medical imagery was also studied. The experimental evaluation revealed the feasibility of applying the gearbox model to the deployment of software pipelines in real-world scenarios as it can significantly improve the end-user service experience when pre-processing large-scale data in comparison with state-of-the-art solutions such as Sacbe and Parsl.
PB Elsevier
SN 1872-7115
YR 2020
FD 2020-05-01
LK https://hdl.handle.net/10016/33801
UL https://hdl.handle.net/10016/33801
LA eng
NO This work has been partially supported by the “Spanish Ministerio de Economia y Competitividad ” under the project grant TIN2016-79637-P “Towards Unification of HPC and Big Data paradigms”.
DS e-Archivo
RD 18 jul. 2024