Kotz, D., Toh, S., and Radhakrishnan, S. A Detailed Simulation Model of the HP 97560 Disk Drive. Tech. Rep. PCS-TR91-220, Dept. of Computer Science, Dartmouth College, July 1994.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....prone to starvation, the effects can be lessened if older requests are given higher priority or if the disk head is sometimes forcibly moved to a new region of the disk. A number of recent papers attempt to model disks and disk activity. Ruemmler and Wilkes [19] and Kotz, Toh, and Radhakrishnan [13] developed detailed models of Hewlett Packard disks. In a separate paper Ruemmler and Wilkes [18] describe disk activity in various unix systems. The traces they obtained were later used by Worthington, Ganger, and Patt [23] to evaluate existing disk scheduling algorithms. Methods for obtaining ....

D. Kotz, S. B. Toh, and S. Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical Report PCS--TR94--220, Dartmouth College, 1994.

....and to utilize it to aggressively tune performance. While detailed models of disk behavior are highly desirable for aggressivescheduling algorithms [14, 13, 17, 1] and comprehensive system modeling [9] obtaining accurate data from state of the art disk drives for detailed disk models (e.g. [10, 3, 5]) is at best tedious. Although methodologies for accurate disk models are well understood [10] and general techniques for on line extraction of parameters from SCSI disk drives have been proposed [18, 1] there is no available system for easily retrieving parameters from a given disk drive. This ....

....more disk parameters, its performance more closely approximates the performance of the real disk drive. However, they do not discuss how they obtained parameters from real disk drives that were used in their models. The observations made by Ruemmler Wilkes have been validated by Kotz et al. [5] who built a detailed simulator of the HP 97560 disk drive. Ganger et al. have made available a highly con gurable disk simulator called DiskSim [3] The simulator models several independent parts of the disk drive system: device drivers, busses, controllers, adapters, and disk drives. DiskSim ....

D. Kotz, S. B. Toh, and S. Radhakishnan. A detailed simulation model of the HP 97560 disk drive. Technical Report PCS-TR94-220, Dept. of Computer Science, Darthmouth College, July 1994. 19

....disk simulation and modelling in this thesis. Optimizing disk layout is not performed. We assume that the disks are either in nitely fast or have a simple xed latency, allowing us to focus on bottlenecks in the core of the le system. Disk modelling and simulation has been studied by Kotz et al. [33]. On an invalidation based snoopy bus protocol. 2.3 Operating Systems Operating systems typically integrate le systems by providing a virtual le system (VFS) interface. The operating system uses this interface to create les, read and write les, obtain le attributes, create directories, ....

David Kotz, Song Bac Toh, and Sriram Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical report PCS-TR94-220, Department of Computer Science, Dartmouth College, July 1994.

....the real world arrival time and the request details, and it returns the computed service time. After the appropriate real time delay, the timing loop tells the storage interface component to report completion. The emulator based evaulation of eager writing [25] used a disk simulator by Kotz et al. [15] in this way. Although it is straightforward, this first approach often does not properly handle concurrent requests. For example, a new request arrival may affect the service time of outstanding requests due to bus contention, request overlapping, or request scheduling. A more general approach ....

David Kotz, Song Bac Toh, and Sriram Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical report PCS TR94 220. Department of Computer Science, Dartmouth College, July 1994.

....model of the HP97560 disk drive. We modeled the components of the I O subsystem of interest: the disk head, the disk controller, the data bus, and the read and write caches. We based our models on techniques described by Ruemmler and Wilkes [7] and an implementation by Kotz, Toh, and Radhakrishnan [6]. We then used the Snake 5 and Snake 6 file system traces collected by Ruemmler and Wilkes [8] to drive a detailed simulation of the disk subsystem. To validate our approach of grouping segments into hot and cold regions, we looked at the effect of manually varying the size of the hot region of ....

David Kotz, Song Bac Toh, and Sriram Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical Report PCS--TR94--220, Department of Computer Science, Dartmouth College, 1994.

....the real world arrival time and the request details, and it returns the computed service time. After the appropriate real time delay, the timing loop tells the storage interface component to report completion. Wang et al. s emulator based evaluation of eager writing used Kotz s disk simulator [15] in this way. Although it is straightforward, this rst approach often does not properly handle concurrent requests. For example, a new request arrival may a ect the service time of outstanding requests due to bus contention, request overlapping, or request scheduling. A more general approach is ....

D. Kotz, S. B. Toh, and S. Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical report PCS{TR94-220. Department of Computer Science, Dartmouth College, July 1994.

....standard DTV bitrate) We also assume that fast scan streams are played at 5 fps (DR f ) so that the viewer 2 Please refer to appendix A for a derivation of the equation. 4 can comprehend and react in time to the content. For computing the seek overhead we follow closely the model developed in [11, 14], which has been proven to be asymptotically close to the real disks. Note that although a different set of disk parameters can lead to different absolute performance values, the relative performance between schemes remain the same. 4.1 Memory Requirement To measure system performance and cost, ....

D. Kotz, S. B. Toh, and S. Radhakrishnan. A detailed simulation model of the hp 97560 disk drive. Dartmouth College Technical Report PCS-TR94-220, 1994.

....tracks are occupied and there is one free sector at the next rotational position in at least one of these tracks. Figure 1 validates the model of (2) with a simulation of the HP97560 and the Seagate ST19101 disks whose relevant parameters are detailed in Table 1. The well validated HP97560 model [17, 26] is approximately eight years old. The state of art Seagate disk [20, 28] is more complex but its simulator is only a coarse approximation. For example, the model simulates a single density zone while the actual disk has multiple zones. The simulated eager 3 0 0.5 1.0 1.5 2.0 2.5 3.0 0 20 ....

....and file systems shown in Figure 5. 4.1 The Regular Disk The regular disk module simulates a portion of the HP97560 disk or the Seagate ST19101 disk, whose relevant parameters are shown in Table 1. A ramdisk driver is used to store file data using 24 MB of kernel memory. The Dartmouth simulator [17] is ported into the kernel to ensure realistic timing behavior of the HP disk. We adjust the parameters of the Dartmouth model to coincide with those of the Seagate disk to simulate the faster disk. Although not as accurate, the Seagate model gives a clear indication of the impact of disk ....

Kotz, D., Toh, S., and Radhakrishnan, S. A Detailed Simulation Model of the HP 97560 Disk Drive. Tech. Rep. PCS-TR91-220, Dept. of Computer Science, Dartmouth College, July 1994.

....model of the HP97560 disk drive. We modeled the components of the I O subsystem of interest: the disk head, the disk controller, the data bus, and the read and write caches. We based our models on techniques described by Ruemmler and Wilkes [7] and an implementation by Kotz, Toh, and Radhakrishnan [6]. We then used the Snake 5 and Snake 6 file system traces collected by Ruemmler and Wilkes [8] to drive a detailed simulation of the disk subsystem. To validate our approach of grouping segments into hot and cold regions, we looked at the effect of manually varying the size of the hot region of ....

David Kotz, Song Bac Toh, and Sriram Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical Report PCS--TR94--220, Department of Computer Science, Dartmouth College, 1994.

....at around 1200 cylinders are believed to be due to drive thermal recalibration. 103 Source: fs.tex DRAFT of 11:06, June 28, 1996 drives are usually highly inaccurate unless they take into account these factors [Ruemmler94] A detailed model of one particular disk drive is described in [Kotz94] which manages to estimate the cost of most disk transactions to within 1 . The model is based around an event driven simulator, required intimate knowledge of the internals of the particular drive and takes over 12,000 lines of code. Techniques for on line extraction of the important parameters ....

David Kotz, Song Bac Toh, and Sriram Radhakrishnan. A Detailed Simulation Model of the HP 97560 Disk Drive. Technical Report PCS-TR94-220, Dartmouth College, July 1994. (p 104)

....we assume a very fast interconnect. Each processor has its own cache and prefetcher; we assume no overhead for memory copies incurred by cache hits. The I O nodes have no cache beyond the disk cache. The disk model we have used is Kotz s reimplementation of Ruemmler and Wilkes HP 97560 model [12, 15]. Table 3 lists the significant simulator parameters. Although this disk is not modern, advances in disk technology do not qualitatively affect the queuing effects we are addressing, and this particular model has been well tested and validated. There are several possible implementation strategies ....

KOTZ, D., TOH,S.B.,AND RADHAKRISHNAN, S. A Detailed Simulation Model of the HP 97560 Disk Drive. Tech. rep., Department of Computer Science, Dartmouth College, 1994.

....explained as follows. In the algorithm for the external Laplace equation solver, a submatix is forwardly and backwardly accessed in every other iteration repeatedly. Under a contiguous disk block allocation, disk blocks are accessed contiguously in both 6 The disk model is reimplemented by Kotz [4] and has been validated against disk traces provided by HP. 30000 40000 50000 60000 70000 80000 90000 100000 110000 3H 6H 9H 12H Running Time (ms) Memory Size (H = 1,024) Performance for 4 iterations (contigous allocation) FP version LD version 130000 132000 134000 136000 ....

D. Kotz, S. B. Toh, and S. Radhakrishnan. A detailed simulation model of the hp 97560 disk drive. Technical Report Technical Report PCS-TR94-220, Dept. of Computer Science, Dartmouth College, July 1994.

....to the drive deciding it needed to perform thermal recalibration. Before recalibration, drives 84 were often observed to record a higher incidence of reads being missed on the rst pass. Even once the parameters are recovered, modelling the behaviour of a modern disk drive can be very complex. Kotz94] describes an event driven simulator that requires over 12,000 lines of code to estimate the cost of most disk operations to within 1 for a particular Hewlett Packard disk drive. Due to this complexity, few operating systems attempt to consider disk rotation while scheduling disk requests. To ....

David Kotz, Song Bac Toh, and Sriram Radhakrishnan. A Detailed Simulation Model of the HP 97560 Disk Drive. Technical Report PCS-TR94-220, Dartmouth College, July 1994. (p 85)

....model of the HP97560 disk drive. We modeled the components of the I O subsystem of interest: the disk head, the disk controller, the data bus, and the read and write caches. We based our models on techniques described by Ruemmler and Wilkes [7] and an implementation by Kotz, Toh, and Radhakrishnan [6]. We then used the Snake 5 and Snake 6 file system traces collected by Ruemmler and Wilkes [8] to drive a detailed simulation of the disk subsystem. To validate our approach of grouping segments into hot and cold regions, we looked at the effect of manually varying the size of the hot region of ....

David Kotz, Song Bac Toh, and Sriram Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical Report PCS--TR94--220, Department of Computer Science, Dartmouth College, 1994.

....and to utilize it to aggressively tune performance. While detailed models of disk behavior are highly desirable for aggressive scheduling algorithms [14, 13, 17, 1] and comprehensive system modeling [9] obtaining accurate data from state of the art disk drives for detailed disk models (e.g. [10, 3, 5]) is at best tedious. Although methodologies for accurate disk models are well understood [10] and general techniques for on line extraction of parameters from SCSI disk drives have been proposed [18, 1] there is no available system for easily retrieving parameters from a given disk drive. This ....

....more disk parameters, its performance more closely approximates the performance of the real disk drive. However, they do not discuss how they obtained parameters from real disk drives that were used in their models. The observations made by Ruemmler Wilkes have been validated by Kotz et al. [5] who built a detailed simulator of the HP 97560 disk drive. Ganger et al. have made available a highly con gurable disk simulator called DiskSim [3] The simulator models several independent parts of the disk drive system: device drivers, busses, controllers, adapters, and disk drives. DiskSim ....

D. Kotz, S. B. Toh, and S. Radhakishnan. A detailed simulation model of the HP 97560 disk drive. Technical Report PCS-TR94-220, Dept. of Computer Science, Darthmouth College, July 1994. 19

....the effects can be lessened if older requests are given higher priority or if the disk head is sometimes forcibly moved to a new region of the disk. A number of recent papers study disk scheduling from a more real world perspective. Ruemmler and Wilkes [59] and Kotz, Toh and Radhakrishnan [40] developed detailed models of Hewlett Packard disks. In a separate paper Ruemmler and Wilkes [58] describe disk activity in various unix 2 systems. The traces they obtained were later used by Worthington, Ganger and Patt [71] to evaluate many disk scheduling algorithms, including those described ....

D. Kotz, S. B. Toh, and S. Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical Report PCS--TR94--220, Dartmouth College, 1994.

No context found.

Kotz, D., Toh, S., and Radhakrishnan, S. A Detailed Simulation Model of the HP 97560 Disk Drive. Tech. Rep. PCS-TR91-220, Dept. of Computer Science, Dartmouth College, July 1994.

No context found.

Kotz, D., Toh, S. B., and Radhakrishnan, S. A detailed simulation model of the HP-97560 disk drive. Tech. Rep. PCS-TR94220, Department of Computer Science, Dartmouth College, July 1994.

No context found.

David Kotz, Song Bac Toh, and Sriram Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical report PCS-TR94-220, Department of Computer Science, Dartmouth College, July 1994.

No context found.

David Kotz, Song Bac Toh, and Sriram Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical report PCS--TR94--220. Department of Computer Science, Dartmouth College, July 1994.

No context found.

D. Kotz, S. B. Toh, and S. Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical Report TR94-220, Dartmouth College, 1994.

No context found.

David Kotz, Song Bac Toh, and Sriram Radhakrishnan, A detailed simulation model of the HP 97560 disk drive, Technique Report PCS-TR94-220, Department of Computer Science, Dartmouth College, July 1994.

No context found.

David Kotz, Song Bac Toh, and Sriram Radhakrishnan. A detailed simulation model of the HP 97560 disk drive. Technical Report TR94-220, Dartmouth College, 1994.

No context found.

D. Kotz, S.B. Toh, and S. Radhakrishnan, A Detailed Simulation Model of the hp 97560 Disk Drive, Technical Report PCS-TR94220, Darthmouth College, 1994.

No context found.

D. Kotz, S.B. Toh, and S. Radhakrishnan. A detailed simulation model of the hp 97560 disk drive. Dartmouth College Technical Report PCS-TR94220, 1994.

First 50 documents  Next 50

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