Java platform for hIgh PErformance and Real-time large scale data management (JUNIPER)

Czech title:Java platform for hIgh PErformance and Real-time large scale data management (JUNIPER)
Reseach leader:Smrž Pavel
Team leaders:Rychlý Marek, Škoda Petr
Team members:Dytrych Jaroslav, Fučík Otto, Jeřábek Jan (), Kouřil Jan, Musil Petr, Otrusina Lubomír, Zachariáš Michal
Agency:Information and Communication Technologies (ICT) 7th Framework programme - Seventh Research Framework Programme
Keywords:Performance guarantees, realtime, Big Data, streaming data, stored data, parallelisation, Java
The efficient and real-time exploitation of large streaming data sources and stored data poses many questions regarding the underlying platform, including: 1) Performance - how can the potential performance of the platform be exploited effectively by arbitrary applications; 2) Guarantees - how can the platform support guarantees regarding processing streaming data sources and accessing stored data; and 3) Scalability - how can scalable platforms and applications be built. The fundamental challenge addressed by the project is to enable application development using an industrial strength programming language that enables the necessary performance and performance guarantees required for real-time exploitation of large streaming data sources and stored data. The project's vision is to create a Java Platform that can support a range of high-performance Intelligent Information Management application domains that seek real-time processing of streaming data, or real-time access to stored data. This will be achieved by developing Java and UML modelling technologies to provide: 1) Architectural Patterns - using predefined libraries and annotation technology to extend Java with new directives for exploiting streaming I/O and parallelism on high performance platforms; 2) Virtual Machine Extensions - using class libraries to extend the JVM for scalable platforms; 3) Java Acceleration - performance optimisation is achieved using Java JIT to Hardware (FPGA), especially to enable real-time processing of fast streaming data; 4) Performance Guarantees - will be provided for common application real-time requirements; and 5) Modelling - of persistence and real-time within UML / MARTE to enable effective development, code generation and capture of real-time system properties. The project will use financial and web streaming case studies from industrial partners to provide industrial data and data volumes, and to evaluate the developed technologies. 318763


2015Scheduling Advisor for Performance Tuning of Juniper Applications, software, 2015
Authors: Rychlý Marek
2014Apache Storm topology for processing tweets, software, 2014
Authors: Kouřil Jan
 Heterogeneity-Aware Scheduler for Stream Processing Frameworks, software, 2014
Authors: Škoda Petr


2015RYCHLÝ Marek, ŠKODA Petr and SMRŽ Pavel. Heterogeneity-Aware Scheduler for Stream Processing Frameworks. International Journal of Big Data Intelligence. Olney: Inderscience Publishers, 2015, vol. 2, no. 2, pp. 70-80. ISSN 2053-1397.
2014RYCHLÝ Marek, ŠKODA Petr and SMRŽ Pavel. Scheduling Decisions in Stream Processing on Heterogeneous Clusters. In: 2014 Eighth International Conference on Complex, Intelligent and Software Intensive Systems. Birmingham: IEEE Computer Society, 2014, pp. 614-619. ISBN 978-1-4799-4325-8.

Your IPv4 address:
Switch to IPv6 connection

DNSSEC [dnssec]