R. Bennett, K. Bryant, A. Sussman, R. Das, and J. Saltz. Jovian: A framework for optimizing parallel I/O. In Proceedings of the Scalable Parallel Libraries Conference, pages 10--20, Mississippi State, MS, 1994. IEEE Computer Society Press.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....certain optimizations. Another advantage is that if any processor needs to access some data which was previously modified by some other processor, it can be done using just a read call without any additional synchronization. The idea of collective I O has also been used in other schemes such as in [1, 11, 10]. In the next section, we describe the Extended Two Phase Method for reading sections of out of core arrays. The method for writing sections is analogous and is discussed in Section 6. 5 Reading Sections of Out of Core Array Let us assume that each processor needs to read some regular section ....

....Write Extended Two Phase I (1:100:1, l 100p:100 100p:l) 1.839 3.250 II (l 100p:100 100p:l, 1:100:1) 2.678 2.501 III (200 200p:400 200p:l, 1:512:1) 11.64 8.715 IV (l 32p:16 32p:l, 1:4096:1) 98.96 10.25 V (200 200p:400 200p:l, l 200p:512 200p:l) 11.33 6. 461 7 Related Work The Jovian library [1] also provides support for accessing sections of out of core arrays. This library uses separate processes called coalescing processes to perform I O optimizations. All application processes send I O requests to predetermined coalescing processes. Each coalescing process is responsible for ....

R. Bennett, K. Bryant, A. Sussman, R. Das, and J. Saltz. Jovian: A Framework for Optimizing Parallel I/O. In Proceedings of the Scalable Parallel Libraries Conference, October 1994.

....processes to be accessed. A simple polling mechanism is built into the library and handles requests for pages from other application processes without the need for additional I O processes or kernel support. Disks local to each application process are used for data storage. The Jovian library [1] is something of a hybrid implementation, with a library of routines used by application processes to perform I O through a separate set of processes, known as coalescing processes, which exist for the duration of application execution in order to handle I O. These coalescing processes (CPs) ....

Robert Bennett, Kelvin Bryant, Alan Sussman, Raja Das, and Joel Saltz. Jovian: A framework for optimizing parallel I/O. In Proceedings of the Scalable Parallel Libraries Conference, pages 10--20, Mississippi State, MS, October 1994. IEEE Computer Society Press.

....that shows the best I O performance. 32 64 32 64 32 64 0.0 20.0 40.0 60.0 80.0 100.0 I O Bandwidth (MB Sec. Original) Level 1) Level 2 3) SDM Figure 7. I O bandwidth for RT 5. Related Work Several efforts have sought to optimize I O in parallel file systems and runtime libraries [3, 5, 6, 14, 16, 18, 22, 27, 31]. SRB (Storage Resource Broker) 2] provides an uniform interface to access various storage systems, such as file systems, Unitree, HPSS and database objects. However, it does not fully support the optimizations implemented in MPIIO. Shoshani et al. 28, 29] describe an architecture for op6 ....

R. Bennett, K. Bryant, A. Sussman, R. Das, and J. Saltz. Jovian: A Framework for Optimizing Parallel I/O. In Proceedings of the Scalable Parallel Libraries Conference, pages 10--20. IEEE Computer Society Press, Oct. 1994.

....in memory. For instance, 3D arrays distributed (CYCLIC(K) BLOCK, BLOCK) in memory of compute nodes can be written to disks at I O nodes in a row major (or column major) order. Array rearrangements between memory and disk are usually supported by the underlying parallel file system or I O library [1, 2, 5, 9]. If a fine grained data distribution is used in memory along one or more array dimensions, it can cause many small messages to be passed between compute nodes and I O nodes during I O operations; therefore, message passing performance of the underlying interconnect as well as disk subsystems can ....

R. Bennett, K. Bryant, A. Sussman, R. Das, and J. Saltz. Jovian: Framework for Optimizing Parallel I/O. In Proceedings of the Scalable Parallel Libraries Conference, pages 10--20, Oct. 1994.

....Commonly these methods are realized by runtime libraries, which are linked to the application programs. Thus, the application program performs the data accesses itself without the need for dedicated I O server programs. Examples for this group are the Two Phase method [3] the Jovian framework [1], and the Extended Two Phase method [27] 2.1.2 I O Level Methods The I O level methods try to reorganize the disk access requests of the application programs to achieve better performance. This is done by independent I O node servers, which collect the requests and perform the accesses. ....

....c) offset: 5 x 14 x 4 = 280 skip: 7 x 14 x 4 = 392 d) stride: 0 x 4 = 0 p5 p5 p5 p5 offset: 3 x 4 = 12 1 2 2 1 1 1 2 2 skip repl. by offset of irreg. block p5 p5 p5 p5 Figure 12: processor 5 basic block stride subtype sub count count repeat offset sub actual 0 0 0 1 [1] 24 8 [0] basic block stride subtype sub count count repeat offset sub actual 0 0 0 1 12 12 Dimension: 2 Access Desc skip 392 no blocks 1 basics Dimension: 1 Access Desc basics [0, 1] no blocks 2 basic block stride subtype sub count count repeat offset ....

[Article contains additional citation context not shown here]

Robert Bennett, Kelvin Bryant, Alan Sussman, Raja Das, and Joel Saltz. Jovian: A framework for optimizing parallel I/O. In Proceedings of the Scalable Parallel Libraries Conference, pages 10-20, Mississippi State, MS, October 1994. IEEE Computer Society Press.

No context found.

R. Bennett, K. Bryant, A. Sussman, R. Das, and J. Saltz. Jovian: A framework for optimizing parallel I/O. In Proceedings of the Scalable Parallel Libraries Conference, pages 10--20, Mississippi State, MS, 1994. IEEE Computer Society Press.

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