ABSTRACT
Validating experimental results in the field of storage systems is a challenging task, mainly due to the many changes in software and hardware that computational environments go through. Determining if an experiment is reproducible entails two separate tasks: re-executing the experiment and validating the results. Existing reproducibility efforts have focused on the former, envisioning techniques and infrastructures that make it easier to re-execute an experiment. In this position paper, we focus on the latter by analyzing the validation workflow that an experiment re-executioner goes through. We notice that validating results is done on the basis of experiment design and high-level goals, rather than exact quantitative metrics. Based on this insight, we introduce a declarative format for specifying the high-level components of an experiment as well as describing generic, testable conditions that serve as the basis for validation. We present a use case in the area of distributed storage systems to illustrate the usefulness of this approach.
- R. D. Peng, "Reproducible research in computational science," Science, vol. 334, Dec. 2011, pp. 1226--1227.Google ScholarCross Ref
- J. Vitek and T. Kalibera, "Repeatability, reproducibility, and rigor in systems research," Proceedings of the ninth ACM international conference on embedded software, New York, NY, USA: ACM, 2011, pp. 33--38. Google ScholarDigital Library
- C. Collberg, T. Proebsting, and A.M. Warren, "Repeatability and benefaction in computer systems research," 2015.Google Scholar
- C. T. Brown, "How we make our papers replicable," 2014. Available at: http://ivory.idyll.org/blog/2014-our-paper-process.html.Google Scholar
- S. Krishnamurthi and J. Vitek, "The real software crisis: Repeatability as a core value," Commun. ACM, vol. 58, Feb. 2015, pp. 34--36. Google ScholarDigital Library
- B. Clark, T. Deshane, E. Dow, S. Evanchik, M. Finlayson, J. Herne, and J.N. Matthews, "Xen and the art of repeated research," Proceedings of the annual conference on USENIX annual technical conference, Berkeley, CA, USA: USENIX Association, 2004, pp. 47--47. Google ScholarDigital Library
- I. Jimenez, C. Maltzahn, J. Lofstead, A. Moody, K. Mohror, R.H. Arpaci-Dusseau, and A. Arpaci-Dusseau, "The role of container technology in reproducible computer systems research," 2015 IEEE international conference on cloud engineering (IC2E), Tempe, AZ: 2015.Google Scholar
- F. Chirigati, D. Shasha, and J. Freire, "ReproZip: Using provenance to support computational reproducibility," Proceedings of the 5th USENIX conference on theory and practice of provenance, Berkeley, CA, USA: USENIX Association, 2013, pp. 1--1. Google ScholarDigital Library
- I. Jimenez, C. Maltzahn, A. Moody, and K. Mohror, Redo: Reproducibility at scale, UC Santa Cruz, 2014.Google Scholar
- S. A. Weil, S. A. Brandt, E. L. Miller, D. D. E. Long, and C. Maltzahn, "Ceph: A scalable, high-performance distributed file system," Proceedings of the 7th symposium on operating systems design and implementation, Berkeley, CA, USA: USENIX Association, 2006, pp. 307--320. Google ScholarDigital Library
- R. Ricci and E. Eide, "Introducing CloudLab: Scientific infrastructure for advancing cloud architectures and applications,";login: vol. 39, Dec. 2014, pp. 36--38.Google Scholar
- K. Popper, The logic of scientific discovery, New Delhi: Routledge, 2002.Google Scholar
- J. P. Ignizio, "On the establishment of standards for comparing algorithm performance," Interfaces, vol. 2, Nov. 1971, pp. 8--11. Google ScholarDigital Library
- H. Crowder, R.S. Dembo, and J.M. Mulvey, "On reporting computational experiments with mathematical software," ACM Trans. Math. Softw., vol. 5, Jun. 1979, pp. 193--203. Google ScholarDigital Library
- C. Dietrich and D. Lohmann, "The dataref versuchung: Saving time through better internal repeatability," SIGOPS Oper. Syst. Rev., vol. 49, Jan. 2015, pp. 51--60. Google ScholarDigital Library
- D. G. Feitelson, "From repeatability to reproducibility and corroboration," SIGOPS Oper. Syst. Rev., vol. 49, Jan. 2015, pp. 3--11. Google ScholarDigital Library
- J. Freire, P. Bonnet, and D. Shasha, "Computational reproducibility: State-of-the-art, challenges, and database research opportunities," Proceedings of the 2012 ACM SIGMOD international conference on management of data, New York, NY, USA: ACM, 2012, pp. 593--596. Google ScholarDigital Library
- C. Neylon, J. Aerts, C.T. Brown, S.J. Coles, L. Hatton, D. Lemire, K.J. Millman, P. Murray-Rust, F. Perez, N. Saunders, N. Shah, A. Smith, G. Varoquaux, and E. Willighagen, "Changing computational research: The challenges ahead," Source Code for Biology and Medicine, vol. 7, Dec. 2012, pp. 1--2.Google ScholarCross Ref
- R. LeVeqije, I. Mitchell, and V. Stodden, "Reproducible research for scientific computing: Tools and strategies for changing the culture," Computing in Science Engineering, vol. 14, Jul. 2012, pp. 13--17. Google ScholarDigital Library
- V. Stodden, F. Leisch, and R.D. Peng, Implementing reproducible research, CRC Press, 2014.Google ScholarCross Ref
- D. L. Donoho, A. Maleki, I.U. Rahman, M. Shahram, and V. Stodden, "Reproducible research in computational harmonic analysis," Computing in Science & Engineering, vol. 11, Jan. 2009, pp. 8--18. Google ScholarDigital Library
- J. Achenbach, "The new scientific revolution: Reproducibility at last," The Washington Post, Jan. 2015.Google Scholar
- M. B. Yaffe, "Reproducibility in science," Science Signaling, vol. 8, Apr. 2015, pp. eg5--eg5.Google ScholarCross Ref
- Editorial, "Journals unite for reproducibility," Nature, vol. 515, Nov. 2014, pp. 7--7.Google ScholarCross Ref
- I. P. Gent, "The recomputation manifesto," arXiv:1304.3674 {cs}, Apr. 2013.Google Scholar
- S. Crouch, N. Hong, S. Hettrick, M. Jackson, A. Pawlik, S. Sufi, L. Carr, D. De Roure, C. Goble, and M. Parsons, "The software sustainability institute: Changing research software attitudes and practices," Computing in Science Engineering, vol. 15, Nov. 2013, pp. 74--80. Google ScholarDigital Library
- H. S. Gunawi, T. Do, P. Joshi, P. Alvaro, J.M. Hellerstein, A.C. Arpaci-Dusseau, R.H. Arpaci-Dusseau, K. Sen, and D. Borthakur, "FATE and DESTINI: A framework for cloud recovery testing," Proceedings of the 8th USENIX conference on networked systems design and implementation, Berkeley, CA, USA: USENIX Association, 2011, pp. 238--252. Google ScholarDigital Library
- K. Li, P. Joshi, A. Gupta, and M.K. Ganai, "ReproLite: A lightweight tool to quickly reproduce hard system bugs," Proceedings of the ACM symposium on cloud computing, New York, NY, USA: ACM, 2014, pp. 25:1--25:13. Google ScholarDigital Library
- C. Liu, B. T. Loo, and Y. Mao, "Declarative automated cloud resource orchestration," Proceedings of the 2Nd ACM symposium on cloud computing, New York, NY. Google ScholarDigital Library
- Tackling the reproducibility problem in storage systems research with declarative experiment specifications
Recommendations
Time-Independent Experiment Reproducibility: Turning the WalT Platform into a Time Machine
PE-WASUN '16: Proceedings of the 13th ACM Symposium on Performance Evaluation of Wireless Ad Hoc, Sensor, & Ubiquitous NetworksThis paper concerns the problem of testing wireless networks in a way that guarantees repeatability and reproducibility of experiments. To contribute to this research objective, we have developed WalT, a reproducible platform for running reproducible ...
Towards reproducibility in recommender-systems research
Numerous recommendation approaches are in use today. However, comparing their effectiveness is a challenging task because evaluation results are rarely reproducible. In this article, we examine the challenge of reproducibility in recommender-system ...
Research and Design of Experiment Management System Based on Web
ICCIS '12: Proceedings of the 2012 Fourth International Conference on Computational and Information SciencesTo improve experiment management efficiency in the universities and informatization management level of the laboratory, we attempt to develop the experiment management system based on the Web. Through the system, the laboratory manager can set the open ...
Comments