R. Thakur, E. Lusk, and W. Gropp. Users guide for ROMIO: A high-performance, portable MPI-IO implementation. Technical Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, October 1997.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....full control over data movement and caching. MPI IO, the I O part of the MPI 2 standard [12] is to provide an interface specifically intended for portable and high performance parallel I O. MPI IO defines a comprehensive interface with many features for both portability and performance. ROMIO [16] is a well known implementation of MPI IO with high performance and portability. The current version runs on different types of machines, including parallel machines, SMP and networks of workstations (NOW) Supported file systems are IBM PIOFS, Intel PFS, HP HFS, SGI XFS, NEC SFS, Clemson PVFS, ....

....of MPI IO with high performance and portability. The current version runs on different types of machines, including parallel machines, SMP and networks of workstations (NOW) Supported file systems are IBM PIOFS, Intel PFS, HP HFS, SGI XFS, NEC SFS, Clemson PVFS, NFS and any Unix file system (UFS) [16]. Even though MPI IO implementations are available for different file systems, so far, there is no implementation to take advantage of the emerging DAFS standard. In this paper, we take on such a challenge. We design, develop, and evaluate an implementation of MPI IO (ROMIO) on DAFS over cLAN ....

[Article contains additional citation context not shown here]

R. Thakur, W. Gropp, and E. Lusk. Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation (Revised Version). ANL/MCS-TM-234, Sept. 2000.

....have been successful in improving performance. Many parallel applications require access to small, potentially non contiguous, data chunks. Collective I O merges multiple I O accesses (both contiguous and noncontiguous) generated by multiple processes, into a single I O function call. MPI I O [26, 28], which is included as part of the MPI 2 standard release, provides for collective communication so that multiple MPI processes can access a single MPI I O file. We utilize MPI collective I O (we will refer to collective I O as MPI I O in this paper) in the work presented here. But when multiple ....

R. Thakur, W. Gropp, and E. Lusk. Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation. Mathematics and Computer Science Division, Argonne National Laboratory, Oct. 1997. ANL/MCS-TM-234.

....Figure 1: Overview of the MCIO. significantly increased. This optimization has many variants [19, 13, 22] although any CIO technique can be used for MCIO, the one used in this study is two phase I O as implemented in ROMIO, a portable implementation of MPI IO from Argonne National Laboratories [26]. ROMIO has been incorporated into several MPI libraries, including the MPI implementations of several vendors (e.g. HP, SGI, NEC) and MPICH and LAM, two widelyused, freely available, portable MPI implementations. Numerous techniques have been proposed in literature to optimize I O accesses. The ....

Thakur, R., Lusk, E., Gropp, W. Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation. Technical Memorandum ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, IL, Revised July 1998. APPENDIX A An Example LP Model Ee give the LP model for the access in

....Figure 1: Overview of the MCIO. significantly increased. This optimization has many variants [19, 13, 22] although any CIO technique can be used for MCIO, the one used in this study is two phase I O as implemented in ROMIO, a portable implementation of MPI IO from Argonne National Laboratories [26]. ROMIO has been incorporated into several MPI libraries, including the MPI implementations of several vendors (e.g. HP, SGI, NEC) and MPICH and LAM, two widelyused, freely available, portable MPI implementations. Numerous techniques have been proposed in literature to optimize I O accesses. The ....

Thakur, R., Lusk, E., Gropp, W. Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation. Technical Memorandum ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, IL, Revised July 1998. APPENDIX A An Example LP Model Ee give the LP model for the access in

.... each processor to do I O on behalf of other processors if doing so improves the overall performance [9] This optimization has many variants [28, 26, 38] the one used in this study is two phase I O as implemented in ROMIO, a portable implementation of MPI IO from Argonne National Laboratories [41]. ROMIO has been incorporated into several MPI libraries, including the MPI implementations of several vendors (e.g. HP, SGI, NEC) and MPICH and LAM, two widely used, freely available, portable MPI implementations. In this implementation, I O is performed in two phases: an I O phase and a ....

Rajeev Thakur, Ewing Lusk, and William Gropp. Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation. Technical Memorandum ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, Revised July 1998. 57

....[28] Galley [18] or the MPI IO initiative, which de ned a parallel le interface for the Message Passing Interface (MPI) standard [17, 7] The MPI I O standard has been widely accepted as a programmers interface to parallel I O. A portable implementation of this standard is the ROMIO library [30]. Runtime libraries aim for being tools for the application programmer. Therefore the executing application can hardly react dynamically to changing system situations (e.g. number of available disks or processors) or problem characteristics (e.g. data reorganization) because the data access ....

....a le scattered over several disks rather than residing on a single one. Thus, the I O can actually be done highly parallel. The application programmer need not care for the physical location of the le and can therefore treat a scattered le as one logical contiguous le. At the moment ROMIO[30] is the widest spread MPIIO implementation, which is part of the MPICH software package. Of less importance are 3 further MPI IO implementations, namely PMPIO[11] MPI IO PIOFS[9] and HPSS[12] ViMPIOS implements (similar to ROMIO) all routines de ned in the MPI 2 I O chapter except shared le ....

[Article contains additional citation context not shown here]

Rajeev Thakur, Ewing Lusk, and William Gropp. Users guide for ROMIO: A high-performance, portable MPIIO implementation. Technical Report ANL/MCS-TM234, Mathematics and Computer Science Division, Argonne National Laboratory, October 1997. 16

....node whose memory is used for fast backing storage and logically lies between the local physical memory and fast stable storage such as disks. ROMIO ROMIO is a high performance, portable implementation of MPI IO. A key feature component is an internal abstract I O device layer called ADIO [41]. Scalable I O Facility (SIOF) SIOF is a project to enable I O performance to scale with the computing performance of parallel computing systems and achieve terascale computing. Scotch Parallel Storage System (SPFS) Parallel storage systems are constructed as testbeds for the development of ....

Rajeev Thakur, Ewing Lusk, and William Gropp. Users guide for ROMIO: A high-performance, portable MPI-IO implementation. Technical Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, October 1997. 38

....is MPI IO as part of the MPI standard [4] which supports parallel file accesses analogously to the MPI message passing scheme. It allows noncontiguous file access patterns and supports collective disk operations. At the moment four different MPI IO implementations are available, namely ROMIO [7], PMPIO [2] HPSS [9] and MPI IO PIOFS an MPI IO implementation by IBM Watson Research Center specific for the IBM SP 2 architec # This work was carried out as part of the research project Language, Compiler, and Advanced Data Structure Support for Parallel I O Operations supported by the ....

Rajeev Thakur, Ewing Lusk, and William Gropp, Users guide for ROMIO: A high-performance, portable MPIIO implementation, Tech. Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, October 1997.

....or portable multipurpose systems that can be used on di erent hardware platforms (e.g. PASSION [12] Galley [8] PANDA [10] or ViPIOS [4] In addition to those systems MPI IO ( 7] has been proposed as a standard parallel le interface for MPI. With a reference implementation available (ROMIO [13]) this standard has been widely accepted by now. Despite of all the e orts to provide application programmers with means to perform parallel I O eciently, very few work has yet been done to automatically optimize the performance of I O systems. Considering how tedious and complex performance ....

Rajeev Thakur, Ewing Lusk, and William Gropp, Users guide for ROMIO: A high-performance, portable MPI-IO implementation, Tech. Report ANL/MCS-TM234, Mathematics and Computer Science Division, Argonne National Laboratory, October 1997.

....Argonne National Laboratory did not make a firm commitment to provide a full MPI 2 extension of their highly successful MPICH software, but some implementers were waiting for it to become available anyway. As first steps, ANL published an implementation of the parallel I O chapter, known as ROMIO [35], and a C binding as extensions to MPICH. As a consequence, the parallel I O chapter was the first part of MPI 2 available on a wide variety of parallel platforms, and it has since been integrated into several proprietary MPI implementations. Even two years after the publication of the MPI 2 ....

R. Thakur, E. Lusk and W. Gropp, Users' guide for ROMIO: a high performance, portable MPI-IO implementation, Technical Memorandum ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory (July 1998).

.... Passing Interface , is a well established standard that has recently been extended to a new standard called MPI2 [5] The most important of the extensions for our work is the addition of I O facilities, which have been given the name MPI IO [3] and are described in Chapter 10 of [5] ROMIO [8] is a high performance, portable implementation of MPI IO that is freely available at http: www.mcs.anl.gov home thakur romio. The key to ROMIO s portability is its implementation on top of ADIO [7] which stands for Abstract Device interface for parallel I O . In order to port ROMIO to a new ....

Rajeev Thakur, Ewing Lusk, and William Gropp. Users guide for ROMIO: A high-performance, portable MPI-IO implementation. Technical Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, October 1997.

....interface for parallel I O as part of the MPI 2 standard [9] This interface is commonly referred to as MPI IO. MPI IO is a rich interface with many features designed specifically for performance and portability. Multiple implementations of MPI IO, both portable and machine specific, are available [5, 7, 13, 15, 26]. To avoid the above mentioned problem of many distinct, small I O requests, MPI IO allows users to specify the entire noncontiguous access pattern and read or write all the data with a single function call. MPI IO also allows users to specify collectively the I O requests of a group of processes, ....

Rajeev Thakur, Ewing Lusk, and William Gropp. Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation. Technical Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, Revised December 1999.

....for parallel I O (commonly referred to as MPI IO) as part of the MPI 2 standard [19] MPI IO is a comprehensive API with many features intended specifically for I O parallelism, portability, and high performance. Implementations of MPI IO, both portable and machine specific, are already available [7, 13, 23, 24, 34]. In this paper, we discuss the issues involved in implementing MPI IO portably on multiple machines and file systems and also achieving high performance. We argue that if an implementation uses just the basic Unix I O functions in order to achieve portability, it will have limitations in both ....

....believe that the only way to implement MPI IO portably with complete functionality and high performance is to have a mechanism that can utilize the special features and functions of each file system. We describe such an architecture, called ADIO, which we use in our MPI IO implementation, ROMIO [34]. 2.1 Abstract Device Interface for I O ADIO [31] an abstract device interface for I O, is a mechanism specifically designed for implementing parallel I O APIs portably on multiple file systems. We developed ADIO before MPI IO became a standard, as a means to implement and experiment with ....

[Article contains additional citation context not shown here]

R. Thakur, E. Lusk, and W. Gropp. Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation. Technical Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, Revised July 1998.

....to a file, RId denotes the current identification number (runid) and each shadowed box (along with the name beside it) shows the SDM generated file for storing the corresponding data set. can be used from an MPI IO program the ROMIO implementation of MPI IO that we used supports direct I O [35]. We present performance results with both direct I O and regular (buffered) I O. We used two application templates, ASTRO3D and a three dimensional Euler solver, in our performance experiments. For ASTRO3D, we used a problem size of 256 Theta 256 Theta 256. We ran the program for one time step ....

Rajeev Thakur, Ewing Lusk, and William Gropp. Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation. Technical Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, Revised December 1999.

....overlap, the atomic mode should be used. Porting applications from Unix I O to MPI IO is easy, because MPI IO provides functions that are equivalent to those in Unix I O. For better performance, however, the special features of MPIIO must be used. Many implementations of MPI IO are in progress [30, 8, 13, 23, 22], and most vendors of parallel machines plan to provide MPI IO as part of their MPI 2 product. We ourselves are developing a portable MPI IO implementation called ROMIO. ROMIO 1.0.0 is freely available from http: www.mcs.anl.gov home thakur romio and works on most parallel computers and networks ....

R. Thakur, E. Lusk, and W. Gropp. Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation. Technical Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, October 1997.

....for parallel I O as part of the MPI 2 standard [18] This interface is commonly referred to as MPI IO. MPI IO is a rich interface with many features designed specifically for performance and portability. Multiple implementations of MPI IO, both portable and machine specific, are available [9, 14, 24, 26, 39]. To avoid the abovementioned problem of many distinct, small I O requests, MPI IO allows users to specify the entire noncontiguous access pattern and read or write all the data with a single I O function call. MPI IO also allows users to specify collectively the I O requests of a group of ....

....data layout in memory can be defined by using any MPI basic or derived datatype; therefore, any general, noncontiguous access pattern can be compactly represented. 3 ROMIO: A Portable Implementation of MPI IO We have developed a high performance, portable implementation of MPI IO, called ROMIO [39]. It is freely available from the Web site http: www.mcs.anl.gov romio. ROMIO is designed to run on multiple machines and file systems. The current version runs on the following machines: IBM SP; Intel Paragon; Cray T3E; HP Exemplar; SGI Origin2000; NEC SX 4; other symmetric multiprocessors from ....

R. Thakur, E. Lusk, and W. Gropp. Users Guide for ROMIO: A High-Performance, Portable MPI-IO Implementation. Technical Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, Revised July 1998.

....new API for parallel I O (which we call MPI IO) as part of the MPI 2 standard [17] MPI IO is a comprehensive API with many features intended specifically for I O parallelism, portability, and high performance. Implementations of MPI IO, both portable and machine specific, are beginning to appear [6, 12, 22, 23, 33]. In this paper, we discuss the issues involved in implementing MPI IO portably on multiple machines and file systems and also achieving high performance. We argue that if an implementation uses just the basic Unix I O functions in order to achieve portability, it will have limitations in both ....

....We believe that the only way to implement MPI IO portably with complete functionality and high performance is to have a mechanism that can utilize the special features and functions of each file system. We describe such an architecture, called ADIO, which we use in our MPI IO implementation, ROMIO [33]. 2.1 Abstract Device Interface for I O A key component of ROMIO that enables such a portable MPI IO implementation is an internal layer called ADIO [30] ADIO, an abstract device interface for I O, is a mechanism specifically designed for implementing parallel I O APIs portably on multiple ....

[Article contains additional citation context not shown here]

R. Thakur, E. Lusk, and W. Gropp. Users Guide for ROMIO: A High-Performance, Portable MPIIO Implementation. Technical Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, October 1997.

No context found.

R. Thakur, E. Lusk, and W. Gropp. Users guide for ROMIO: A high-performance, portable MPI-IO implementation. Technical Report ANL/MCS-TM-234, Mathematics and Computer Science Division, Argonne National Laboratory, October 1997.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC