Performing Dynamically Injected Tasks on Processes Prone to Crashes and Restarts
To identify the tradeoffs between efficiency and fault-tolerance in dynamic cooperative computing, we initiate the study of a task performing problem under dynamic processes’ crashes/restarts and task injections. The system consists of n message-passing processes which, subject to dynamic crashes and restarts, cooperate in performing independent tasks that are continuously and dynamically injected to the system. The task specifications are not known a priori to the processes. This problem abstracts todays Internet-based computations, such as Grid computing and cloud services, where tasks are generated dynamically and different tasks may be known to different processes. We measure performance in terms of the number of pending tasks, and as such it can be directly compared with the optimum number obtained under the same crash-restart-injection pattern by the best off-line algorithm. We propose several deterministic algorithmic solutions to the considered problem under different information models and correctness criteria, and we argue that their performance is close to the best possible offline solutions.
KeywordsPerforming tasks Dynamic task injection Crashes and restarts Competitive analysis Distributed Algorithms
Unable to display preview. Download preview PDF.
- 1.Ajtai, M., Aspnes, J., Dwork, C., Waarts, O.: A theory of competitive analysis for distributed algorithms. In: Proc. of FOCS 1994, pp. 401–411 (1994)Google Scholar
- 2.Amazon Elastic Compute Cloud, http://aws.amazon.com/ec2
- 5.Awerbuch, B., Kutten, S., Peleg, D.: Competitive distributed job scheduling. In: Proc. of STOC 1992, pp. 571–580 (1992)Google Scholar
- 6.Bartal, Y., Fiat, A., Rabani, Y.: Competitive algorithms for distributed data management. In: Proc. of STOC 1992, pp. 39–50 (1992)Google Scholar
- 8.Chlebus, B.S., Kowalski, D.R., Shvartsman, A.A.: Collective asynchronous reading with polylogarithmic worst-case overhead. In: Proc. of STOC 2004, pp. 321–330 (2004)Google Scholar
- 11.Enabling Grids for E-sciencE (EGEE), http://www.eu-egee.org
- 12.Emek, Y., Halldorsson, M.M., Mansour, Y., Patt-Shamir, B., Radhakrishnan, J., Rawitz, D.: Online set packing and competitive scheduling of multi-part tasks. In: Proc. of PODC 2010, pp. 440–449 (2010)Google Scholar
- 13.Georgiou, C., Gilbert, S., Kowalski, D.R.: Meeting the deadline: on the complexity of fault-tolerant continuous gossip. In: Proc. of PODC 2010, pp. 247–256 (2010)Google Scholar
- 17.Hui, L., Huashan, Y., Xiaoming, L.: A Lightweight Execution Framework for Massive Independent Tasks. In: Proc. of MTAGS 2008 (2008)Google Scholar
- 20.Malewicz, G., Austern, M.H., Bik, A.J.C., Dehnert, J.C., Horn, I., Leiser, N., Czajkowski, G.: Pregel: A system for large-scale graph processing. In: Proc. of SIGMOD 2010, pp. 135–145 (2010)Google Scholar
- 24.Tech. Report of this work, http://www.cs.ucy.ac.cy/~chryssis/disc11-TR.pdf