Ruemmler and Wilkes. "Disk Shuffling". Hewlett-Packard Software and Systems Laboratories Technical Report, 1991. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
System-Oriented Evaluation of I/O Subsystem Performance - Ganger (1995) (7 citations) (Correct)
....of data. This concept can be applied in two ways to improve performance for reads and for writes, respectively. To improve read performance, one can occasionally re organize the data blocks to place popular blocks near the center of the disk and cluster blocks that tend to be accessed together. [Ruemmler91] uses an open subsystem model to evaluate the benefits of such an approach. Wolf89, Vongsathorn90] on the other hand, measure this approach by implementing it in real systems. To improve write performance, one can write data blocks to convenient locations (i.e. locations that are close to the ....
C. Ruemmler, J. Wilkes, "Disk Shuffling", Technical Report HPL-CSP-91-30, Hewlett-Packard Laboratories, October 3, 1991.
Adaptive Block Rearrangement Under UNIX - Akyurek, Salem (1994) (4 citations) (Correct)
....and Carson [Vongsath 90] showed that dynamically clustering frequently accessed data worked better than static placement. In that study, disk cylinders are dynamically rearranged using the organ pipe heuristic, according to observed data access frequencies. Recent work in the DataMesh project [Ruemmler 91] considered rearrangement of cylinders and blocks, with mixed results. Their conclusion that block shuffling generally outperforms cylinder shuffling corroborates one of our own. A similar approach is employed in the experimental iPcress file system[Staelin 91] which monitors access to files and ....
Ruemmler, C., J. Wilkes, "Disk Shuffling", HPL-91-156, Hewlett-Packard Laboratories, Palo Alto, CA, October, 1991.
Adaptive Block Rearrangement - Akyurek, Salem (1992) (6 citations) (Correct)
....seek times by as much as 40 45 . In these experiments, cylinder reference frequencies were measured daily. Based on the measured frequencies, all of the disk s cylinders were rearranged according to the organ pipe heuristic to service the next day s requests. Recent work in the DataMesh project [Ruemmler 91] considered rearrangement of cylinders and blocks, with mixed results. Their conclusion that block shuffling generally outperforms cylinder shuffling corroborates one of our own. A similar approach is employed in the experimental iPcress file system [Staelin 91] which monitors access to files ....
....studies driven by traces from a VMS system. We conclude with a summary of our results and possible extensions to our study. 2 Adaptive Rearrangement Overview The idea of block rearrangement is motivated by the fact that access to data stored on disks is highly skewed[Floyd 89, Staelin 91, Ruemmler 91] Of the thousands of data blocks stored on a disk, a small fraction of them absorbs most of the requests. If the hot (frequently accessed) blocks are spread over the surface of the disk, distant from each other, long seek delays may result. Hot blocks can be clustered to reduce seek times. ....
[Article contains additional citation context not shown here]
Ruemmler, C., J. Wilkes, "Disk Shuffling", HPL-91-156, Hewlett-Packard Laboratories, Palo Alto, CA, October, 1991.
A Case for Network-Attached Secure Disks - Gibson, al. (1996) (20 citations) (Correct)
....and functionality while exporting the same interface. Notable examples include Redundant Arrays of Inexpensive Disks (RAID) transparent failure recovery; real time geometry sensitive scheduling; buffer caching, readahead, and writebehind; compression; dynamic mapping and representation migration [Patterson88, Gibson92, Massiglia94, StorageTek94, Wilkes95, Ruemmler91, Varma95]. Currently, smart storage subsystems contain tens to hundreds of GBytes of storage, service thousands of accesses per second, and easily saturate double and quadruple speed SCSI buses. With this pressure on the performance of SCSI s physical interconnect, the industry is today (1996) experiencing ....
Ruemmler, C. and Wilkes, J., "Disk Shuffling", Hewlett-Packard Laboratories Concurrent Systems Project Tech Report HPL-CSP-91-30. page 19
File Server Scaling with Network-Attached Secure Disks - Gibson, Nagle, Amiri.. (1997) (72 citations) (Correct)
.... The level of indirection introduced by SCSI has also led to transparent improvements in storage performance such as RAID; transparent failure recovery; real time geometry sensitive scheduling; buffer caching; read ahead and write behind; compression; dynamic mapping; and representation migration [Patterson88, Gibson92, Massiglia94, StorageTek94, Wilkes95, Ruemmler91, Varma95]. However, in order to overcome the speed, addressability and connectivity limitations of current SCSI implementations [Sachs94, ANSI95] the industry is turning to high speed packetized interconnects such as Fibre Channel at up to 1 Gbps [Benner96] The disk drive industry anticipates the ....
Ruemmler, C. and Wilkes, J., "Disk Shuffling", Hewlett-Packard Laboratories Concurrent Systems Project Tech Report HPL-CSP-91-30.
Active Storage for Large-Scale Data Mining and.. - Riedel, Gibson.. (1998) (67 citations) (Correct)
....6.3. Batching Scheduling Throughout this paper we have concentrated on the parallelism and filtering advantages of Active Disks. In storage systems research, however, the most common application specific optimizations are scheduling, batching and prefetching [Kotz94, Patterson95, Bitton88, Ruemmler91] and Active Disks can be expected to execute these types of remote functions as well. For database operations, disk scheduling has typically been done at much higher levels of query processing and data layout. However, interconnect scheduling may be useful to database operators such as parallel ....
Ruemmler, C. and Wilkes, J., "Disk Shuffling", HP Labs Technical Report HPL-CSP91 -30, 1991
Active Storage for Large-Scale Data Mining and Multimedia - Riedel, Gibson, Faloutsos (1998) (67 citations) (Correct)
....reducing the bottleneck. These two advantages are the focus of this paper because they promise orders of magnitude potential improvements. In storage systems research, however, the most common application specific optimizations are scheduling, batching and prefetching of disk operations [Bitton88, Ruemmler91]. Active Disks can be expected to execute these types of remote functions as well. In particular, we might expect Active Disks to participate as part of a disk directed I O model, where scatter gather accesses are optimized using local information at the disks [Kotz94] Or in prefetching systems ....
Ruemmler, C. and Wilkes, J., "Disk Shuffling" HP Labs Technical Report HPL-CSP-91-30, 1991
Embedded Inodes and Explicit Grouping: Exploiting Disk.. - Ganger, Kaashoek (1997) (31 citations) (Correct)
.... in same cylinder group as containing directory and data blocks in same cylinder group as owning inode) Similarly, several researchers have investigated the value of moving the most popular (i.e. most heavily used) data to the centermost disk cylinders in order to reduce disk seek distances [Vongsathorn90, Ruemmler91, Staelin91]. As described in Section 2, simply locating related objects near each other offers some performance gains, but such locality affects only the seek time and is thus limited in scope. Co locating related objects and reading writing them as a unit offers qualitatively larger improvements in ....
C. Ruemmler, J. Wilkes, "Disk Shuffling", Technical Report HPL-CSP-91-30, Hewlett-Packard Laboratories, October 3, 1991.
The HP AutoRAID hierarchical storage system - Wilkes, al. (1996) (125 citations) Self-citation (Wilkes) (Correct)
.... studies on I O access patterns, disk shuffling, and file system restructuring have shown that these conditions are often met in practice [Akyurek and Salem 1993; Deshpandee and Bunt 1988; Floyd and Schlatter Ellis 1989; Geist et al. 1994; Majumdar 1984; McDonald and Bunt 1989; McNutt 1994; Ruemmler and Wilkes 1991; Ruemmler and Wilkes 1993; Smith 1981] Such a storage hierarchy could be implemented in a number of different ways: Manually, by the system administrator. This is how large mainframes have been run for decades. Gelb 1989] discusses a slightly refined version of this basic idea. The ....
RUEMMLER, C., AND WILKES, J. 1991. Disk shuffling. Tech. Rep. HPL--91--156, Hewlett-Packard Laboratories, Palo Alto, Calif.
An Introduction to Disk Drive Modeling - Ruemmler, Wilkes (1994) (271 citations) Self-citation (Ruemmler Wilkes) (Correct)
No context found.
C. Ruemmler and J. Wilkes, "Disk shuffling," Tech. Report HPL-91-156, Hewlett-Packard Laboratories, Palo Alto, Calif., Oct. 1991.
Asynchronous Bulk File I/O in Titanium, a High-Performance SPMD.. - Bonachea (1999) (Correct)
No context found.
Ruemmler and Wilkes. "Disk Shuffling". Hewlett-Packard Software and Systems Laboratories Technical Report, 1991.
Active Disks - Remote Execution for Network-Attached Storage - Riedel (1999) (18 citations) (Correct)
No context found.
Ruemmler, C. and Wilkes, J., "Disk Shuffling", Technical Report HPL-CSP-91-30, HewlettPackard Labs, 1991
Using Data Clustering to Improve Cleaning Performance for.. - Chiang, Lee, Chang (Correct)
No context found.
C. Ruemmler and J. Wilkes, `Disk Shuffling', Technical Report HPL-91-156, Hewlett-Packard Laboratories, Palo Alto, CA, October 28, 1991.
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