Abstract
In this contribution we compare the performance of the Input/Output load (I/O) of a High-Performance Computing (HPC) application on two different File Systems: CephFS and Lustre; our goal is to assess whether CephFS could be considered a valid choice for intense HPC applications. We perform our analysis using a real HPC workload, namely RegCM, a climate simulation application, and IOR, a synthetic benchmark application, to simulate several I/O patterns using different I/O parallel libraries (MPI-IO, HDF5, PnetCDF). We compare writing performance for the two different I/O approaches that RegCM implements: the so-called spokesperson or serial, and a truly parallel one. The small difference registered between the serial I/O approach and the parallel one motivates us to explore in detail how the software stack interacts with the underlying File Systems. For this reason, we use IOR and MPI-IO hints related to Collective Buffering and Data Sieving to analyze several I/O patterns on the two different File Systems.
Finally we investigate Lazy I/O, a unique feature of CephFS, which disables file coherency locks introduced by the File System; this allows Ceph to buffer writes and to fully exploit its parallel and distributed architecture. Two clusters were set up for these benchmarks, one at CNR-IOM and a second one at Pawsey Supercomputing Centre; we performed similar tests on both installations, and we recorded a four-times I/O performance improvement with Lazy I/O enabled.
Preliminary results collected so far are quite promising and further actions and new possible I/O optimizations are presented and discussed.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
- 2.
- 3.
Refer to “How do I use hints?” in the FAQ of the project: https://ior.readthedocs.io/en/latest/userDoc/faq.html.
References
Giorgi F., Coppola E., Solmon F., Mariotti L.: RegCM4: model description and preliminary tests over multiple CORDEX domains. Clim. Res. 52, 7–29 (2011). https://doi.org/10.3354/cr01018. The RegCM versions used in this study are available in the release page of the project on GitHub: https://github.com/ictp-esp/RegCM/releases
Li, J., et al.: Parallel netCDF: a scientific high-performance I/O interface. In: The Proceedings of ACM/IEEE Conference on Supercomputing, pp. 39, November 2003. Project web-site: https://parallel-netcdf.github.io. Accessed 22 July 2019
Wauteleta, P., Kestener, P.: Parallel IO performance and scalability study on the PRACE curie supercomputer, Partnership For Advanced Computing in Europe (PRACE), Technical report, September 2009
Thakur, R., Ross, R., Lusk, E., Gropp, W., Latham, R.: Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation. http://citeseerx.ist.psu.edu/viewdoc/download?rep=rep1&type=pdf&doi=10.1.1.218.9852
Wautelet, P.: Parallel I/O experiments on massively parallel computers. In: ScicomP 17 Conference, Paris, May 2011. http://www.idris.fr/docs/docu/IDRIS/IDRISioscicomp2011.pdf
IOR benchmark project on GitHub. https://github.com/hpc/ior. Accessed 22 July 2019
CephFS documentation about POSIX compliance. http://docs.Ceph.com/docs/mimic/Cephfs/posix/. Accessed 22 July 2019
CephFS documentation about Lazy I/O feature. http://docs.Ceph.com/docs/master/Cephfs/lazyio/. Accessed 22 July 2019
Carns, P., et al.: Understanding and improving computational science storage access through continuous characterization. In: Proceedings of 27th IEEE Conference on Mass Storage Systems and Technologies (MSST 2011) (2011). http://www.mcs.anl.gov/uploads/cels/papers/P1859.pdf. Project web-site: https://www.mcs.anl.gov/research/projects/darshan/
Acknowledgments
We thank Luca Cervigni, Pawsey Supercomputing Centre, for the opportunity to collaborate and extract Lazy I/O timings from their Ceph test cluster.
We thank Pablo Llopis, CERN, for all the precious HPC support provided.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Chiusole, A., Cozzini, S., van der Ster, D., Lamanna, M., Giuliani, G. (2019). An I/O Analysis of HPC Workloads on CephFS and Lustre. In: Weiland, M., Juckeland, G., Alam, S., Jagode, H. (eds) High Performance Computing. ISC High Performance 2019. Lecture Notes in Computer Science(), vol 11887. Springer, Cham. https://doi.org/10.1007/978-3-030-34356-9_24
Download citation
DOI: https://doi.org/10.1007/978-3-030-34356-9_24
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-34355-2
Online ISBN: 978-3-030-34356-9
eBook Packages: Computer ScienceComputer Science (R0)