Cited By
View all- Du YMarchal LPallez GRobert Y(2022)Optimal Checkpointing Strategies for Iterative ApplicationsIEEE Transactions on Parallel and Distributed Systems10.1109/TPDS.2021.309944033:3(507-522)Online publication date: 1-Mar-2022
Profiles of Upcoming HPC Applications and Their Impact on Reservation Strategies
Pages 1178 - 1190
Abstract
With the expected convergence between HPC, BigData and AI, new applications with different profiles are coming to HPC infrastructures. We aim at better understanding the features and needs of these applications in order to be able to run them efficiently on HPC platforms. The approach followed is bottom-up: we study thoroughly an emerging application, <italic>Spatially Localized Atlas Network Tiles</italic> (SLANT, originating from the neuroscience community) to understand its behavior. Based on these observations, we derive a generic, yet simple, application model (namely, a linear sequence of stochastic jobs). We expect this model to be representative for a large set of upcoming applications from emerging fields that start to require the computational power of HPC clusters without fitting the typical behavior of large-scale traditional applications. In a second step, we show how one can use this generic model in a scheduling framework. Specifically we consider the problem of making reservations (both time and memory) for an execution on an HPC platform based on the application expected resource requirements. We derive solutions using the model provided by the first step of this work. We experimentally show the robustness of the model, even with very few data points or using another application, to generate the model, and provide performance gains with regards to standard and more recent approaches used in the neuroscience community.
References
[1]
D. Andresen, W. Hsu, H. Yang, and A. Okanlawon, “Machine learning for predictive analytics of compute cluster jobs,” CoRR, vol. abs/1806.01116, 2018. [Online]. Available: http://arxiv.org/abs/1806.01116
[2]
J. Ansel, K. Arya, and G. Cooperman, “DMTCP: Transparent checkpointing for cluster computations and the desktop,” in Proc. IEEE Int. Symp. Parallel Distrib. Process., 2009, pp. 1–12.
[3]
G. Aupy, A. Gainaru, V. Honoré, P. Raghavan, Y. Robert, and H. Sun, “Reservation strategies for stochastic jobs,” in Proc. 33rd IEEE Int. Parallel Distrib. Process. Symp., 2019, pp. 166–175.
[4]
L. Bautista-Gomez, S. Tsuboi, D. Komatitsch, F. Cappello, N. Maruyama, and S. Matsuoka, “FTI: High performance fault tolerance interface for hybrid systems,” in Proc. Int. Conf. High Perform. Comput. Netw. Storage Anal., 2011, pp. 1–12.
[5]
J. Breitbart, S. Pickartz, S. Lankes, J. Weidendorfer, and A. Monti, “Dynamic co-scheduling driven by main memory bandwidth utilization,” in Proc. IEEE Int. Conf. Cluster Comput., 2017, pp. 400–409.
[6]
J. Bruno, P. Downey, and G. N. Frederickson, “Sequencing tasks with exponential service times to minimize the expected flow time or makespan,” J. ACM, vol. 28, no. 1, pp. 100–113, 1981.
[7]
Y. Chen, “Checkpoint and restore of micro-service in docker containers,” in Proc. 3rd Int. Conf. Mechatronics Ind. Informat., 2015, pp. 915–918.
[8]
J. T. Daly, “A higher order estimate of the optimum checkpoint interval for restart dumps,” Future Gener. Comput. Syst., vol. 22, no. 3, pp. 303–312, 2006.
[9]
J. Dean and S. Ghemawat, “MapReduce: Simplified data processing on large clusters,” Commun. ACM, vol. 51, no. 1, pp. 107–113, Jan. 2008.
[10]
A. Gainaru, G. Aupy, A. Benoit, F. Cappello, Y. Robert, and M. Snir, “Scheduling the I/O of HPC applications under congestion,” in Proc. IEEE Int. Parallel Distrib. Process. Symp., 2015, pp. 1013–1022.
[11]
A. Gainaru, et al., “Reservation and checkpointing strategies for stochastic jobs,” in Proc. 34th IEEE Int. Parallel Distrib. Process. Symp., 2020, pp. 853–863.
[12]
A. Gainaru, G. Pallez, H. Sun, and P. Raghavan, “Speculative scheduling for stochastic HPC applications,” in Proc. 48th Int. Conf. Parallel Process., 2019, Art. no.
[13]
A. Gainaru, H. Sun, G. Aupy, Y. Huo, B. A. Landman, and P. Raghavan, “On-the-fly scheduling versus reservation-based scheduling for unpredictable workflows,” Int. J. High Perform. Comput. Appl., vol. 33, pp. 1140–1158, 2019.
[14]
R. Garg, A. Mohan, M. Sullivan, and G. Cooperman, “CRUM: Checkpoint-restart support for CUDA's unified memory,” in Proc. IEEE Int. Conf. Cluster Comput., 2018, pp. 302–313.
[15]
E. Gaussier, J. Lelong, V. Reis, and D. Trystram, “Online tuning of EASY-backfilling using queue reordering policies,” IEEE Trans. Parallel Distrib. Syst., vol. 29, no. 10, pp. 2304–2316, Oct. 2018.
[16]
A. Goel and P. Indyk, “Stochastic load balancing and related problems,” in Proc. 40th Annu. Symp. Found. Comput. Sci., 1999, pp. 579–586.
[17]
P. H. Hargrove and J. C. Duell, “Berkeley lab checkpoint/restart (BLCR) for Linux clusters,” J. Phys., vol. 46, pp. 494–499, 2006.
[18]
J. Haxby, et al., “A common, high-dimensional model of the representational space in human ventral temporal cortex,” Neuron, vol. 72, pp. 404–16, Oct. 2011.
[19]
B. Hindman, et al., “Mesos: A platform for fine-grained resource sharing in the data center,” in Proc. 8th USENIX Conf. Netw. Syst. Des. Implementation, 2011, pp. 295–308.
[20]
Y. Huo, A. Carass, S. M. Resnick, D. L. Pham, J. L. Prince, and B. A. Landman, “Combining multi-atlas segmentation with brain surface estimation,” in Proc. SPIE Med. Imag. Image Process., 2016, Art. no.
[21]
Y. Huo, et al., “Consistent cortical reconstruction and multi-atlas brain segmentation,” NeuroImage, vol. 138, pp. 197–210, 2016.
[22]
Y. Huo, et al., “Spatially localized atlas network tiles enables 3D whole brain segmentation from limited data,” in Proc. Int. Conf. Med. Image Comput. Comput. Assisted Intervention, 2018, pp. 698–705.
[23]
Y. Huo, et al., “3D whole brain segmentation using spatially localized atlas network tiles,” NeuroImage, vol. 194, pp. 105–119, 2019.
[24]
T. Hérault and Y. Robert, Eds., Fault-Tolerance Techniques for High-Performance Computing. Berlin, Germany: Springer Verlag, 2015.
[25]
M. Isard, M. Budiu, Y. Yu, A. Birrell, and D. Fetterly, “Dryad: Distributed data-parallel programs from sequential building blocks,” in Proc. 2nd ACM SIGOPS/EuroSys Eur. Conf. Comput. Syst., 2007, pp. 59–72.
[26]
J. Kleinberg, Y. Rabani, and E. Tardos, “Allocating bandwidth for bursty connections,” in Proc. Annu. ACM Symp. Theory Comput., 1997, pp. 664–673.
[27]
R. Kumar and S. Vadhiyar, “Identifying quick starters: Towards an integrated framework for efficient predictions of queue waiting times of batch parallel jobs,” in Proc. Workshop Job Scheduling Strategies Parallel Process., 2013, pp. 196–215.
[28]
P. J. LaMontagne, et al., “OASIS-3: Longitudinal neuroimaging, clinical, and cognitive dataset for normal aging and alzheimer disease,”, Cold Spring Harbor Laboratory Press, 2019. [Online]. Available: https://doi.org/10.1101/2019.12.13.19014902
[29]
B. Landman, “Medical-image Analysis and Statistical Interpretation (MASI) Lab.” [Online]. Available: https://my.vanderbilt.edu/masi/
[30]
S. Li, T. Ben-Nun, S. D. Girolamo, D. Alistarh, and T. Hoefler, “Taming unbalanced training workloads in deep learning with partial collective operations,” in Proc. 25th ACM SIGPLAN Symp. Princ. Practice Parallel Program., 2020, pp. 45–61.
[31]
D. A. Lifka, “The ANL/IBM SP scheduling system,” in Proc. Workshop Job Scheduling Strategies Parallel Process., 1995, pp. 295–303.
[32]
A. Matsunaga and J. A. B. Fortes, “On the use of machine learning to predict the time and resources consumed by applications,” in Proc. 10th IEEE/ACM Int. Conf. Cluster Cloud Grid Comput., 2010, pp. 495–504.
[33]
A. Merzky, M. Santcroos, M. Turilli, and S. Jha, “Radical-pilot: Scalable execution of heterogeneous and dynamic workloads on supercomputers,” CoRR, vol. abs/1512.08194, 2015. [Online]. Available: http://arxiv.org/abs/1512.08194
[34]
A. Mirkin, A. Kuznetsov, and K. Kolyshkin, “Containers checkpointing and live migration,” in Proc. Ottawa Linux Symp., 2008, pp. 85–90.
[35]
R. H. Möhring, A. S. Schulz, and M. Uetz, “Approximation in stochastic scheduling: The power of LP-based priority policies,” J. ACM, vol. 46, no. 6, pp. 924–942, 1999.
[36]
A. W. Mu'alem and D. G. Feitelson, “Utilization, predictability, workloads, and user runtime estimates in scheduling the IBM SP2 with backfilling,” IEEE Trans. Parallel Distrib. Syst., vol. 12, no. 6, pp. 529–543, Jun. 2001.
[37]
J. Niño Mora, “Stochastic scheduling,” Encyclopedia of Optimization. Berlin, Germany: Springer, 2009, pp. 3818–3824.
[38]
T. Patki, D. K. Lowenthal, B. Rountree, M. Schulz, and B. R. De Supinski, “Exploring hardware overprovisioning in power-constrained, high performance computing,” in Proc. 27th Int. ACM Conf. Int. Conf. Supercomput., 2013, pp. 173–182.
[39]
T. Patki, J. J. Thiagarajan, A. Ayala, and T. Z. Islam, “Performance optimality or reproducibility: That is the question,” in Proc. Int. Conf. High Perform. Comput. Netw. Storage Anal., 2019, Art. no.
[40]
S. Pickartz, N. Eiling, S. Lankes, L. Razik, and A. Monti, “Migrating LinuX containers using CRIU,” in Proc. Int. Conf. High Perform. Comput., 2016, pp. 674–684.
[41]
B. Pourghassemi and A. Chandramowlishwaran, “cudaCR: An in-kernel application-level checkpoint/restart scheme for CUDA-enabled GPUs,” in Proc. IEEE Int. Conf. Cluster Comput., 2017, pp. 725–732.
[42]
M. Rodríguez, J. Moríñigo, and R. Mayo-García, “When you have a hammer, everything looks like a nail - Checkpoint/restart in Slurm,” SLURM User Group 2017.
[43]
J. Skovira, W. Chan, H. Zhou, and D. A. Lifka, “The EASY - LoadLeveler API project,” in Proc. Workshop Job Scheduling Strategies Parallel Process., 1996, pp. 41–47.
[44]
M. Tanash, B. Dunn, D. Andresen, W. Hsu, H. Yang, and A. Okanlawon, “Improving HPC system performance by predicting job resources via supervised mach. learning,” in Proc. Practice Experience Adv. Res. Comput. Rise Mach., 2019, Art. no.
[45]
C. T. Vaughan and S. D. Hammond, “Evaluating production load balancing functions for adaptive mesh schemes using mini-applications,” Sandia National Lab.(SNL-NM), Albuquerque, NM, USA, 2017.
[46]
V. K. Vavilapalli, et al., “Apache hadoop YARN: Yet another resource negotiator,” in Proc. 4th Annu. Symp. Cloud Comput., 2013, pp. 5:1–5:16.
[47]
K. Wolter, Ed., Stochastic Models for Fault Tolerance, Restart, Rejuvenation, and Checkpointing. Berlin, Germany: Springer Verlag, 2010.
[48]
L. T. Yang, X. Ma, and F. Mueller, “Cross-platform performance prediction of parallel applications using partial execution,” in Proc. ACM/IEEE Conf. Supercomput., 2005, p. 40.
[49]
J. W. Young, “A first order approximation to the optimum checkpoint interval,” Commun. ACM, vol. 17, no. 9, pp. 530–531, 1974.
[50]
S. Zrigui, R. de Camargo, D. Trystram, and A. Legrand, “Improving the performance of batch schedulers using online job runtime classification,” IEEE Int. Parallel Distrib. Process. Symp., [Online]. Available: https://hal.archives-ouvertes.fr/hal-03023222
Index Terms
- Profiles of Upcoming HPC Applications and Their Impact on Reservation Strategies
Index terms have been assigned to the content through auto-classification.
Recommendations
Preparing HPC Applications for Exascale: Challenges and Recommendations
NBIS '15: Proceedings of the 2015 18th International Conference on Network-Based Information SystemsWhile the HPC community is working towards the development of the first Exaflop computer (expected around 2020), after reaching the Petaflop milestone in 2008 still only few HPC applications are able to fully exploit the capabilities of Petaflop ...
Evaluation of HPC Applications on Cloud
OCS '11: Proceedings of the 2011 Sixth Open Cirrus SummitHPC applications are increasingly being used in academia and laboratories for scientific research and in industries for business and analytics. Cloud computing offers the benefits of virtualization, elasticity of resources and elimination of cluster ...
A grid advance reservation framework for co-allocation and co-reservation across heterogeneous local resource management systems
PPAM'07: Proceedings of the 7th international conference on Parallel processing and applied mathematicsCo-allocation and co-reservation is a key capability of Grid schedulers for supporting some complex Grid applications, e.g., workflow. The chief enabling technology of co-allocation and co-reservation is Advance Reservation (AR), which is typically ...
Comments
Information & Contributors
Information
Published In
1045-9219 © 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See https://www.ieee.org/publications/rights/index.html for more information.
Publisher
IEEE Press
Publication History
Published: 01 May 2021
Qualifiers
- Research-article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- View Citations1Total Citations
- 0Total Downloads
- Downloads (Last 12 months)0
- Downloads (Last 6 weeks)0
Reflects downloads up to 11 Aug 2024
Other Metrics
Citations
Cited By
View all- Du YMarchal LPallez GRobert Y(2022)Optimal Checkpointing Strategies for Iterative ApplicationsIEEE Transactions on Parallel and Distributed Systems10.1109/TPDS.2021.309944033:3(507-522)Online publication date: 1-Mar-2022
View Options
View options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in