Scaling parallel I/O performance through I/O delegate and caching system

Nisar, Arifa; Liao, Wei-keng; Choudhary, Alok

doi:10.1109/sc.2008.5214358

Cited by 66 publications

(50 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…I/O forwarding has also been used in Prost et al (2001), Oldfield et al (2006), Nisar et al (2008), Fu et al (2010), Docan et al (2010) and May (2001) to reduce the I/O impact on computing. The IBM Blue Gene series of supercomputers (Yu et al, 2006) uses independent I/O nodes in their system to handle I/O requests, which are generated in computer nodes and forwarded to I/O nodes.…”

Section: Related Workmentioning

confidence: 99%

A fast input/output library for high-resolution climate models

Wang

Yang

et al. 2014

Geosci. Model Dev.

View full text Add to dashboard Cite

Abstract. We describe the design and implementation of climate fast input/output (CFIO), a fast input/output (I/O) library for high-resolution climate models. CFIO provides a simple method for modelers to overlap the I/O phase with the computing phase automatically, so as to shorten the running time of numerical simulations. To minimize the code modifications required for porting, CFIO provides similar interfaces and features to parallel Network Common Data Form (PnetCDF), which is one of the most widely used I/O libraries in climate models. We deployed CFIO in three highresolution climate models, including two ocean models (POP and LICOM) and one sea ice model (CICE). The experimental results show that CFIO improves the performance of climate models significantly versus the original serial I/O approach. When running with CFIO at 0.1 • resolution with about 1000 CPU cores, we managed to reduce the running time by factors of 7.9, 4.6 and 2.0 for POP, CICE, and LI-COM, respectively. We also compared the performance of CFIO against two existing libraries, PnetCDF and parallel I/O (PIO), in different scenarios. For scenarios with both data output and computations, CFIO decreases the I/O overhead compared to PnetCDF and PIO.

show abstract

Section: Related Workmentioning

confidence: 99%

A fast input/output library for high-resolution climate models

Wang

Yang

et al. 2014

Geosci. Model Dev.

View full text Add to dashboard Cite

show abstract

“…To manage the synchronous time impact of writing, systems like DataTap [2] shift to using asynchronous IO, but face scheduling issues to avoid impacting the existing communication phases of the host application. This approach and other data staging efforts [25] require additional resources and break the direct connection between the end of an IO phase in the host application and the actual writing of data to disk. The communication interference issues have been successfully addressed [1] through several scheduling techniques affording the introduction of services into this output stream or in a staging area.…”

Section: Related Workmentioning

confidence: 99%

Advanced I/O for large-scale scientific applications.

Klasky

Schwan

Oldfield³

et al. 2010

View full text Add to dashboard Cite

As scientific simulations scale to use petascale machines and beyond, the data volumes generated pose a dual problem. First, with increasing machine sizes, the careful tuning of IO routines becomes more and more important to keep the time spent in IO acceptable. It is not uncommon, for instance, to have 20% of an application's runtime spent performing IO in a 'tuned' system. Careful management of the IO routines can move that to 5% or even less in some cases. Second, the data volumes are so large, on the order of 10s to 100s of TB, that trying to discover the scientifically valid contributions requires assistance at runtime to both organize and annotate the data. Waiting for offline processing is not feasible due both to the impact on the IO system and the time required. To reduce this load and improve the ability of scientists to use the large amounts of data being produced, new techniques for data management are required. First, there is a need for techniques for efficient movement of data from the 3 compute space to storage. These techniques should understand the underlying system infrastructure and adapt to changing system conditions. Technologies include aggregation networks, data staging nodes for a closer parity to the IO subsystem, and autonomic IO routines that can detect system bottlenecks and choose different approaches, such as splitting the output into multiple targets, staggering output processes. Such methods must be end-to-end, meaning that even with properly managed asynchronous techniques, it is still essential to properly manage the later synchronous interaction with the storage system to maintain acceptable performance. Second, for the data being generated, annotations and other metadata must be incorporated to help the scientist understand output data for the simulation run as a whole, to select data and data features without concern for what files or other storage technologies were employed. All of these features should be attained while maintaining a simple deployment for the science code and eliminating the need for allocation of additional computational resources.4 Acknowledgment

show abstract

“…To achieve desired I/O semantics, file locking mechanism is used to guarantee the access permissions of individual I/O requests. Two important consistency requirements from POSIX standard known to restrict parallel I/O performance from scaling are atomicity and cache coherence [4], [5]. When multiple processes concurrently access a shared file, file locks may cause serialization of the I/O operations which adversely affects the I/O performance.…”

mentioning

confidence: 99%

“…This paper presents an I/O delegation system that aims to minimize file lock conflicts and improve the MPI independent I/O performance. The I/O delegation work was initiated in [5] which provided an intermediate software layer between the application processes and parallel file systems to enable several I/O optimizations. I/O delegation system employs a set of additional compute processes to carry out the I/O requests for the application processes.…”

mentioning

confidence: 99%

Delegation-Based I/O Mechanism for High Performance Computing Systems

Nisar

Liao

Choudhary

2012

IEEE Trans. Parallel Distrib. Syst.

Self Cite

View full text Add to dashboard Cite

Abstract-Massively parallel applications often require periodic data checkpointing for program restart and post-run data analysis. Although high performance computing systems provide massive parallelism and computing power to fulfill the crucial requirements of the scientific applications, the I/O tasks of highend applications do not scale. Strict data consistency semantics adopted from traditional file systems are inadequate for homogeneous parallel computing platforms. For high performance parallel applications independent I/O is critical, particularly if checkpointing data is dynamically created or irregularly partitioned. In particular, parallel programs generating a large number of unrelated I/O accesses on large scale systems often face serious I/O serializations introduced by lock contention and conflicts at file system layer. As these applications may not be able to utilize the I/O optimizations requiring process synchronization, they pose a great challenge for parallel I/O architecture and software designs. We propose an I/O mechanism to bridge the gap between scientific applications and parallel storage systems. A static file domain partitioning method is developed to align the I/O requests and produce a client-server mapping that minimizes the file lock acquisition costs and eliminates the lock contention. Our performance evaluations of production application I/O kernels demonstrate scalable performance and achieve high I/O bandwidths.

show abstract

Scaling parallel I/O performance through I/O delegate and caching system

Cited by 66 publications

References 15 publications

A fast input/output library for high-resolution climate models

A fast input/output library for high-resolution climate models

Advanced I/O for large-scale scientific applications.

Delegation-Based I/O Mechanism for High Performance Computing Systems

Contact Info

Product

Resources

About