Dynamic Message Processing and Transactional Memory in the Actor Model
With the trend of ever growing data centers and scaling core counts, simple programming models for efficient distributed and concurrent programming are required. One of the successful principles for scalable computing is the actor model, which is based on message passing. Actors are objects that hold local state that can only be modified by the exchange of messages. To avoid typical concurrency hazards, each actor processes messages sequentially. However, this limits the scalability of the model. We have shown in former work that concurrent message processing can be implemented with the help of transactional memory, ensuring sequential processing, when required. This approach is advantageous in low contention phases, however, does not scale for high contention phases. In this paper we introduce a combination of dynamic resource allocation and non-transactional message processing to overcome this limitation. This allows for efficient resource utilization as these two mechanisms can be handled in parallel. We show that we can substantially reduce the execution time of high-contention workloads in a micro-benchmark as well as in a real-world application.
KeywordsActor Model Training Image List Actor Transactional Memory Simulation Grid
Unable to display preview. Download preview PDF.
- 1.Hewitt, C., Bishop, P., Steiger, R.: A universal modular actor formalism for artificial intelligence. In: International Joint Conference on Artificial Intelligence (IJCAI), pp. 235–245 (1973)Google Scholar
- 2.Karmani, R.K., Shali, A., Agha, G.: Actor frameworks for the jvm platform: A comparative analysis. In: International Conference on Principles and Practice of Programming in Java (PPPJ), pp. 11–20 (2009)Google Scholar
- 5.Heiss, H.U., Wagner, R.: Adaptive load control in transaction processing systems. Universität Karlsruhe (1991)Google Scholar
- 6.Herlihy, M., Luchangco, V., Moir, M., Scherer III, W.N.: Software transactional memory for dynamic-sized data structures. In: Annual Symposium on Principles of Distributed Computing (PODC), pp. 92–101. ACM (2003)Google Scholar
- 8.Bronson, N.G., Chafi, H., Olukotun, K.: CCSTM: A library-based STM for Scala. In: Annual Scala Workshop at Scala Days (2010)Google Scholar
- 10.Karmani, R.K., Agha, G.: Actors. In: Padua, D. (ed.) Encyclopedia of Parallel Computing, pp. 1–11. Springer (2011)Google Scholar
- 11.Haller, P.: On the integration of the actor model in mainstream technologies: A Scala perspective. In: Programming Systems, Languages and Applications based on Actors, Agents, and Decentralized Control Abstractions (AGERE), pp. 1–6 (2012)Google Scholar
- 13.Harris, T., Larus, J., Rajwar, R.: Transactional Memory. 2nd edn. Morgan and Claypool Publishers (2010)Google Scholar
- 14.Imam, S.M., Sarkar, V.: Integrating task parallelism with actors. In: International Conference on Object Oriented Programming Systems Languages and Applications (OOPSLA), pp. 753–772. ACM (2012)Google Scholar
- 15.Scholliers, C., Tanter, E., Meuter, W.D.: Parallel actor monitors. In: Brazilian Symposium on Programming Languages (SBLP) (2010)Google Scholar
- 16.ScalaSTM, http://nbronson.github.com/scala-stm/
- 19.Mariethoz, G., Renard, P., Straubhaar, J.: The Direct Sampling method to perform multiple-point geostatistical simulations. Water Resources Research, American Geophysical Union 46(11), 1–14 (2010)Google Scholar