ARTOS

Adaptive Resource Scheduling for Component-based Soft-Realtime Applications

von 2010 bis 2018

Personen


Principal Investigator

  • Prof. Franz J. Hauck

Researchers

  • Dr. Vladimir Nikolov
  • Eugen Frasch

Students

  • Philipp Heeg
  • Matthias Matousek
  • Patrick Mosby
  • Max Käufer
  • Roland Schust

Real-time systems usually require pre-determined statements about resource consumption. This contradicts modern usage scenarios in which applications (apps) from the network are installed on any system and used side by side as required. If applications access existing components, it is almost impossible to make assumptions about the required resources. Our goal is therefore a platform on which Java applications with soft real-time requirements, e.g. multimedia applications, can be dynamically installed and used while meeting their previously unknown CPU time requirements in the interests of the end user.

Applications have several operating modes, in each of which they provide different quality with different resource requirements. Based on the preferences of the end user, the platform automatically ensures a coordinated and ideal distribution of resources by prioritizing the applications. If the available resources are not sufficient, less important applications are switched to less-demanding operating modes or are even stopped.

The scientific challenges lie in the interface between application and platform with regard to possible operating modes, in the determination of the resources actually required for each mode, in the enforcement of resource limits, in the automatic configuration of optimal modes and in the dynamic adaptation of the system to changes in requirements or available resources.

 

Related Publications

2017

Nikolov, V., Bonfert, S., Frasch, E. and Hauck, F.J. 2017. Scheduling interactive HPC applications. Proc. of the 8th Int. Real-Time Scheduling Open Problems Seminar (RTSOPS). 15–16.
Nikolov, V., Wesner, S., Frasch, E. and Hauck, F.J. 2017. A hierarchical scheduling model for dynamic soft-realtime systems. Proc. of the 29th Euromicro Conference on Real-Time Systems (ECRTS) (Dubrovnik, Croatia, Jun. 2017).

2016

Nikolov, V. 2016. A hierarchical scheduling model for dynamic soft-realtime systems. Faculty of Engineering, Computer Science and Psychology, Ulm University. Dissertation.

2015

Nikolov, V., Hauck, F.J. and Schubert, L. 2015. Ein hierarchisches Scheduling-Modell für unbekannte Anwendungen mit schwankenden Ressourcenanforderungen. Echtzeit und Betriebssysteme (Boppard, Nov. 2015).
Nikolov, V., Hauck, F.J. and Wesner, S. 2015. Assembling a framework for unkown real-time applications with RTSJ. Proc. of the 13th Int. Workshop on Java Techn. for Real-time and Embedded Sys. (Paris, Oct. 2015).
Nikolov, V., Kempf, K., Hauck, F.J. and Rautenbach, D. 2015. Distributing the Complexity of Schedulability Tests. Proc. of the 21th IEEE Real-Time and Embedded Technology and Applications Symposium (2015).

2014

Nikolov, V., Kächele, S. and Hauck, F.J. 2014. CLOUDFARM: An Elastic Cloud Platform with Flexible and Adaptive Resource Management. In Proceedings of the IEEE/ACM 7th International Conference on Utility and Cloud Computing (UCC) (London, Dec. 2014).

2012

Nikolov, V., Matousek, M., Rautenbach, D., Draque Penso, L. and Hauck, F.J. 2012. ARTOS: System Model and Optimization Algorithm. Technical Report #VS-R08-2012. Institute of Distributed Systems, University of Ulm.