Using Formal Grammars to Predict I/O Behaviors in HPC: The Omnisc'IO Approach

The increasing gap between the computation performance of post-petascale machines and the performance of their I/O subsystem has motivated many I/O optimizations including prefetching, caching, and scheduling. In order to further improve these techniques, modeling and predicting spatial and temporal I/O patterns of HPC applications as they run has become crucial. In this paper we present Omnisc'IO, an approach that builds a grammar-based model of the I/O behavior of HPC applications and uses it to predict when future I/O operations will occur, and where and how much data will be accessed. To infer grammars, Omnisc'IO is based on StarSequitur, a novel algorithm extending Nevill-Manning's Sequitur algorithm. Omnisc'IO is transparently integrated into the POSIX and MPI I/O stacks and does not require any modification in applications or higher-level I/O libraries. It works without any prior knowledge of the application and converges to accurate predictions of any N future I/O operations within a couple of iterations. Its implementation is efficient in both computation time and memory footprint.

[1]  Al Geist,et al.  Major Computer Science Challenges At Exascale , 2009, Int. J. High Perform. Comput. Appl..

[2]  Daniel A. Reed,et al.  Learning to Classify Parallel Input/Output Access Patterns , 2002, IEEE Trans. Parallel Distributed Syst..

[3]  Robert B. Ross,et al.  CALCioM: Mitigating I/O Interference in HPC Systems through Cross-Application Coordination , 2014, 2014 IEEE 28th International Parallel and Distributed Processing Symposium.

[4]  En-Hui Yang,et al.  Grammar-based codes: A new class of universal lossless source codes , 2000, IEEE Trans. Inf. Theory.

[5]  Daniel A. Reed,et al.  Automatic ARIMA time series modeling for adaptive I/O prefetching , 2004, IEEE Transactions on Parallel and Distributed Systems.

[6]  Darrell D. E. Long,et al.  The case for efficient file access pattern modeling , 1999, Proceedings of the Seventh Workshop on Hot Topics in Operating Systems.

[7]  Eric Barton,et al.  A Novel network request scheduler for a large scale storage system , 2009, Computer Science - Research and Development.

[8]  Y. Charlie Hu,et al.  Program-Counter-Based Pattern Classification in Buffer Caching , 2004, OSDI.

[9]  Fengguang Wu.  Sequential File Prefetching in Linux , 2010 .

[10]  Ian H. Witten,et al.  Identifying Hierarchical Structure in Sequences: A linear-time algorithm , 1997, J. Artif. Intell. Res..

[11]  Song Jiang,et al.  IOrchestrator: Improving the Performance of Multi-node I/O Systems via Inter-Server Coordination , 2010, 2010 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis.

[12]  Robert B. Ross,et al.  On the role of burst buffers in leadership-class storage systems , 2012, 012 IEEE 28th Symposium on Mass Storage Systems and Technologies (MSST).

[13]  Robert B. Ross,et al.  Omnisc'IO: A Grammar-Based Approach to Spatial and Temporal I/O Patterns Prediction , 2014, SC14: International Conference for High Performance Computing, Networking, Storage and Analysis.

[14]  Yang Liu,et al.  Automatic identification of application I/O signatures from noisy server-side traces , 2014, FAST.

[15]  Carlos Maltzahn,et al.  I/O acceleration with pattern detection , 2013, HPDC.

[16]  Robert B. Ross,et al.  PVFS: A Parallel File System for Linux Clusters , 2000, Annual Linux Showcase & Conference.

[17]  Changsheng Xie,et al.  A New Method of Pefetching I/O Requests , 2007, 2007 International Conference on Networking, Architecture, and Storage (NAS 2007).

[18]  Surendra Byna,et al.  Exploring Parallel I/O Concurrency with Speculative Prefetching , 2008, 2008 37th International Conference on Parallel Processing.

[19]  Samuel Lang,et al.  Server-side I/O coordination for parallel file systems , 2011, 2011 International Conference for High Performance Computing, Networking, Storage and Analysis (SC).

[20]  George H. Bryan,et al.  A Benchmark Simulation for Moist Nonhydrostatic Numerical Models , 2002 .

[21]  Andrew J. Hutton,et al.  Lustre: Building a File System for 1,000-node Clusters , 2003 .

[22]  Rajeev Thakur,et al.  KNOWAC: I/O Prefetch via Accumulated Knowledge , 2012, 2012 IEEE International Conference on Cluster Computing.

[23]  Jesús Carretero,et al.  Making the case for reforming the I/O software stack of extreme-scale systems , 2017, Adv. Eng. Softw..

[24]  Surendra Byna,et al.  Parallel I/O prefetching using MPI file caching and I/O signatures , 2008, 2008 SC - International Conference for High Performance Computing, Networking, Storage and Analysis.

[25]  Song Jiang,et al.  Opportunistic Data-driven Execution of Parallel Programs for Efficient I/O Services , 2012, 2012 IEEE 26th International Parallel and Distributed Processing Symposium.

[26]  Daniel A. Reed,et al.  Markov model prediction of I/O requests for scientific applications , 2002, ICS '02.

[27]  Steve Plimpton,et al.  Fast parallel algorithms for short-range molecular dynamics , 1993 .

[28]  Rajeev Thakur,et al.  On implementing MPI-IO portably and with high performance , 1999, IOPADS '99.

[29]  Craig G. Nevill-Manning,et al.  Inferring Sequential Structure , 1996 .

[30]  Adrien Lèbre,et al.  I/O Scheduling Service for Multi-Application Clusters , 2006, 2006 IEEE International Conference on Cluster Computing.