ABSTRACT
This work considers an application executing for a fixed duration, namely the length of the reservation that it has been granted. The checkpoint duration is a stochastic random variable that obeys some well-known probability distribution law. The question is when to take a checkpoint towards the end of the execution, so that the expectation of the work done is maximized. We address two scenarios. In the first scenario, a checkpoint can be taken at any time; despite its simplicity, this natural problem has not been considered yet (to the best of our knowledge). We provide the optimal solution for a variety of probability distribution laws modeling checkpoint duration. The second scenario is more involved: the application is a linear workflow consisting of a chain of tasks with IID stochastic execution times, and a checkpoint can be taken only at the end of a task. First, we introduce a static strategy where we compute the optimal number of tasks before the application checkpoints at the beginning of the execution. Then, we design a dynamic strategy that decides whether to checkpoint or to continue executing at the end of each task. We instantiate this second scenario with several examples of probability distribution laws for task durations.
- Emmanuel Agullo, Luc Giraud, Abdou Guermouche, Jean Roman, and Mawussi Zounon. 2016. Numerical recovery strategies for parallel resilient Krylov linear solvers. Numerical Linear Algebra with Applications 23, 5 (2016), 888–905.Google ScholarCross Ref
- Chandra Chekuri, Waqar Hasan, and Rajeev Motwani. 1995. Scheduling problems in parallel query optimization. In PODS’1995, the 14th ACM SIGACT-SIGMOD-SIGART symposium on Principles of database systems (San Jose, California, United States). ACM Press, New York, NY, USA, 255–265. https://doi.org/10.1145/212433.212471Google ScholarDigital Library
- Alok Choudhary, Wei-keng Liao, Donald Weiner, Pramod Varshney, Richard Linderman, Mark Linderman, and Russell Brown. 2000. Design, Implementation and Evaluation of Parallel Pipelined STAP on Parallel Computers. IEEE Trans. Aerospace Electron. Systems 36, 2 (April 2000), 655 – 662. https://doi.org/10.1109/7.845238Google ScholarCross Ref
- J. T. Daly. 2006. A higher order estimate of the optimum checkpoint interval for restart dumps. Future Generation Comp. Syst. 22, 3 (2006), 303–312.Google ScholarDigital Library
- Ewa Deelman, James Blythe, Yolanda Gil, and Carl Kesselman. 2003. Workflow management in GriPhyN. In Grid Resource Management. Springer.Google Scholar
- Stanley C Eisenstat, Howard C Elman, and Martin H Schultz. 1983. Variational iterative methods for nonsymmetric systems of linear equations. SIAM J. Numer. Anal. 20, 2 (1983), 345–357.Google ScholarDigital Library
- Stanley P Frankel. 1950. Convergence rates of iterative treatments of partial differential equations. Math. Tables Aids Comput. 4, 30 (1950), 65–75.Google ScholarCross Ref
- Fernando Guirado, Ana Ripoll, Concepció Roig, Aura Hernandez, and Emilio Luque. 2006. Exploiting Throughput for Pipeline Execution in Streaming Image Processing Applications. In Euro-Par 2006, Parallel Processing(LNCS 4128). Springer, 1095–1105.Google ScholarDigital Library
- Fernando Guirado, Ana Ripoll, Concepcio Roig, and Emilio Luque. 2005. Optimizing Latency under Throughput Requirements for Streaming Applications on Cluster Execution. In Cluster Computing, 2005. IEEE International. 1–10. https://doi.org/10.1109/CLUSTR.2005.347051Google ScholarCross Ref
- Martin H Gutknecht. 1993. Variants of BICGSTAB for matrices with complex spectrum. SIAM journal on scientific computing 14, 5 (1993), 1020–1033.Google Scholar
- Timothy D. R. Hartley, Ahmed R. Fasih, Charles A. Berdanier, Fusun Ozguner, and Ümit V. Çatalyürek. 2009. Investigating the Use of GPU-Accelerated Nodes for SAR Image Formation. In Proceedings of the IEEE International Conference on Cluster Computing, Workshop on Parallel Programming on Accelerator Clusters (PPAC).Google ScholarCross Ref
- Thomas Herault and Yves Robert (Eds.). 2015. Fault-Tolerance Techniques for High-Performance Computing. Springer Verlag.Google Scholar
- Jihie Kim, Yolanda Gil, and Marc Spraragen. 2004. A knowledge-based approach to interactive workflow composition. In 14th International Conference on Automatic Planning and Scheduling (ICAPS 04).Google Scholar
- Kathleen Knobe, James M. Rehg, Arun Chauhan, Rishiyur S. Nikhil, and Umakishore Ramachandran. 1999. Scheduling constrained dynamic applications on clusters. In Supercomputing’1999, the 1999 ACM/IEEE conference on Supercomputing (Portland, Oregon, United States). ACM, New York, NY, USA, 46. https://doi.org/10.1145/331532.331578Google ScholarDigital Library
- Julien Langou, Zizhong Chen, George Bosilca, and Jack Dongarra. 2008. Recovery patterns for iterative methods in a parallel unstable environment. SIAM Journal on Scientific Computing 30, 1 (2008), 102–116.Google ScholarDigital Library
- William L Oberkampf and Christopher J Roy. 2010. Verification and validation in scientific computing. Cambridge University Press.Google Scholar
- Anthony Rowe, Dimitrios Kalaitzopoulos, Michelle Osmond, Moustafa Ghanem, and Yike Guo. 2003. The discovery net system for high throughput bioinformatics. Bioinformatics 19, Suppl 1 (2003), i225–31.Google ScholarCross Ref
- Christopher J Roy and William L Oberkampf. 2011. A comprehensive framework for verification, validation, and uncertainty quantification in scientific computing. Computer methods in applied mechanics and engineering 200, 25-28 (2011), 2131–2144.Google Scholar
- Y. Saad. 2003. Iterative Methods for Sparse Linear Systems (2nd ed.). Society for Industrial and Applied Mathematics.Google ScholarDigital Library
- Youcef Saad and Martin H Schultz. 1986. GMRES: A generalized minimal residual algorithm for solving nonsymmetric linear systems. SIAM Journal on scientific and statistical computing 7, 3 (1986), 856–869.Google Scholar
- Olcay Sertel, Jun Kong, Hiroyuki Shimada, Ümit V. Çatalyürek, Joel H. Saltz, and Metin N. Gurcan. 2009. Computer-aided prognosis of neuroblastoma on whole-slide images: Classification of stromal development. Pattern Recognition 42, 6 (2009), 1093–1103.Google ScholarDigital Library
- Divyansh Sharma. 2022. Application checkpointing. https://hevodata.com/learn/application-checkpointing/.Google Scholar
- Wikipedia. 2023. Application checkpointing. https://en.wikipedia.org/wiki/Application_checkpointing.Google Scholar
- Wolfram Alpha. 2023. Mathematics. https://www.wolframalpha.com.Google Scholar
- David Young. 1954. Iterative methods for solving partial difference equations of elliptic type. Trans. Amer. Math. Soc. 76, 1 (1954), 92–111.Google ScholarCross Ref
- John W. Young. 1974. A first order approximation to the optimum checkpoint interval. Comm. of the ACM 17, 9 (1974), 530–531.Google ScholarDigital Library
Index Terms
- When to checkpoint at the end of a fixed-length reservation?
Recommendations
Checkpoint scheduling model for optimality
To minimize the expected execution time, a general checkpoint scheduling algorithm is proposed to determine the near optimal checkpointing time sequence. More precisely, based on a simple timing policy, an execution analytical model is introduced and ...
A Communication-Induced Checkpointing Algorithm Using Virtual Checkpoint on Distributed Systems
ICPADS '00: Proceedings of the Seventh International Conference on Parallel and Distributed SystemsCheckpointing is one of the fault-tolerant techniques to restore faults and to restart job fast. The algorithms for checkpointing on distributed systems have been under study for years. These algorithms can be classified into three classes: coordinated, ...
Exploiting Unblocking Checkpoint for Fault-Tolerance in Pregel-Like Systems
Web Information Systems Engineering – WISE 2021AbstractWith the explosive growth of graph size, a series of Pregel-like systems have emerged. Typically, these systems employ checkpointing and rollback mechanisms to achieve fault-tolerance in either blocking or unblocking manner. The blocking ...
Comments