E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. Ergastulum: an approach to solving the workload and device configuration problem. Technical report HPL--SSP--2001.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....can be derived from various business and insurance costs (related to risk management) but the proper way to set performance targets is a different problem. Workload characterization [4, 8] can provide a first estimate of performance requirements and can even be used to guide storage system design [2, 3], but further tuning by the administrator will be necessary. Providing an easy to use interface for this refinement is a necessity. We believe that using a system of complaints from the administrator can provide just such a mechanism. Complaint based tuning will use complaints about the ....

Eric Anderson, Mahesh Kallahalla, Susan Spence, Ram Swaminathan, and Qian Wang. Ergastulum: an approach to solving the workload and device configuration problem. Technical report HPL--SSP--2001.

....allocation decisions. Thus, wide striping may be easier to use from the perspective of storage applications, while narrow striped systems may make potentially better storage decisions and yield better performance. Although placement of objects in large storage systems have been extensively studied [3, 4, 5, 21], surprisingly, no systematic study of these tradeoffs of wide and narrow striping exists in the literature. This is the focus of the present work. Our work seeks to address the following questions: How much (and what) information about the workload is necessary for a policy managed system to ....

....a system can either isolate stores using narrow striping, or group stores with similar QoS requirements, partition the system based on storage requirements of each group, and wide stripe each group within the partition. 4 Related Work The design of policy managed storage systems was pioneered by [3, 5, 4, 18], where techniques for automatically determining storage system configuration were studied. This work determines: 1) the number and types of storage systems that are necessary to support a given workload, 2) the RAID levels for the various objects, and (3) the placement of the objects on the ....

E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. Ergastulum: An Approach to Solving the Workload and Device Configuration Problem. Technical Report HPL-SSP-2001-05, HP Laboratories SSP, May 2001.

.... and adaptive resource management for data intensive network services has been studied in [9] The design of highly dependable ( selfhealing ) Internet services has been studied [14] From the perspective of storage systems, techniques for designing self managing storage have been studied in [2, 4]. The design of such systems involves several sub tasks and issues such self configuration [2, 4] capacity planning [8] automatic RAID level selection [5] initial storage system configuration [3] SAN fabric design [21] and on line data migration [12] have been studied. These efforts are ....

.... The design of highly dependable ( selfhealing ) Internet services has been studied [14] From the perspective of storage systems, techniques for designing self managing storage have been studied in [2, 4] The design of such systems involves several sub tasks and issues such self configuration [2, 4] , capacity planning [8] automatic RAID level selection [5] initial storage system configuration [3] SAN fabric design [21] and on line data migration [12] have been studied. These efforts are complementary to our work which focuses on automatic storage bandwidth allocation to applications with ....

Eric Anderson, Mahesh Kallahalla, Susan Spence, Ram Swaminathan, and Qian Wang. Ergastulum: an approach to solving the workload and device configuration problem. HP Laboratories SSP technical memo HPLSSP -2001-05, May 2001.

....and insurance costs (related to risk management) but the proper way to set performance targets is a different problem. Workload characterization can provide a first estimate of performance requirements (when replacing an existing system) and can even be used to guide storage system design [2, 3], but further tuning by the administrator will be necessary. Providing an easy to use interface for this refinement is a necessity. We believe that using a system of complaints from the administrator can provide just such an interface. Complaint based tuning will use complaints about the ....

E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. Ergastulum: an approach to solving the workload and device configuration problem. Technical report HPL--SSP--2001.

....the system, and adapt its implementation to unexpected changes such as increased failure rates for a component, get the storage system to do it. Figure 1 illustrates this automatic goal directed design process. We have successfully applied this methodology to performance related storage goals [1, 2, 3]. We believe the time is right to extend it to dependability. In particular, we concentrate on the reliability (whether the system discards, loses or corrupts data) and performability (whether the data can be accessed at a particular performance level at a given time) aspects of dependability ....

....storage design system The storage design system is a suite of tools that chooses (or, initially, assists with choosing) data protection techniques that meet the specified goals. We view the problem as an optimization one, by analogy to our earlier work on storage system design for performance [1, 3]. We envision a solution including the following tools: user interrogation tool: a graphical user interface (GUI) that asks users what they want, and then maps their intuitively specified goals into a quantitative specification. This tool removes the need for users to supply reams of ....

[Article contains additional citation context not shown here]

E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. "Ergastulum: an approach to solving the workload and device configuration problem," HP Laboratories SSP technical memo HPL-SSP-2001-05, May 2002.

No context found.

Eric Anderson, Mahesh Kallahalla, Susan Spence, Ram Swaminathan, and Qian Wang. Ergastulum: An approach to solving the workload and device configuration problem. Technical note, HPL-SSP2001 -5, HP Labs, July 2001.

....an IP formulation. To simplify the IP formulation, we make the following assumptions: a) all disks have the same capacity, b) all I Os are read only, c) only one stream is associated with each store, and d) all LUs are RAID1 0. We also use the simpler stream interaction formula as described in [5]. The simplified interaction formula is as follows: Conceptually, when two streams access data (stores) from the same disk, the load that they present to the 9 100 1000 10000 100000 1e 06 uniform zipf normal bimodal in logscale (s) Workload Ergastulum IP Figure 6: Comparison of ....

....Therefore, we can explore the overall search space by generating random orders for the stores, and performing just initial assignment. By executing this process many times, we can limit the probability that we missed a better design. The specific requirements on the design tool can be found in [5]; in essence all configurations have to be reachable, and adding stores to a device should not decrease the utilization. We observe that if we had an optimal configuration, we could order the stores by a prefix traversal of the final device tree, and the design tool would regenerate the same ....

[Article contains additional citation context not shown here]

E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. Ergastulum: an approach to solving the workload and device configuration problem. Technical Report HPL--SSP--2001.

....This section describes the components of the Hippodrome loop in more detail, and explains how the components interact to design a storage system iteratively. The Hippodrome components are the most recent versions of our group s ongoing research into storage system modeling and configuration [2, 3, 4, 5, 6, 10, 26, 30, 31, 32]. The goal of this paper is to show how the different components can work together to automate storage management. Therefore, this section summarizes the techniques used in Hippodrome. The details on each component may be found in the paper on that topic. Hippodrome uses four interdependent ....

....The model calculates the utilization of each LU by combining the estimated stream utilizations using the phasing algorithms found in [10] The algorithms ensure that the utilization of two streams is proportional to the fraction of time that they overlap. 3. 3 Solver component The solver [5] reads as input the workload description generated by the analysis component, and outputs the design of a system that meets the workload s performance requirements. The output specifies a number of disk arrays, the configuration of those arrays (e.g. number of disks, LU configurations, controller ....

[Article contains additional citation context not shown here]

E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. Ergastulum: an approach to solving the workload and device configuration problem. Technical Report HPL--SSP--

....millions of dollars and represents more than half the total system hardware cost. Perhaps even more important is the uncertainty that surrounds a manually designed system: how well) will it meet its performance and availability goals We believe that automatic methods for storage system design [1, 5, 7, 4] can overcome these limitations, because they can consider a wider range of workload interactions, and explore a great deal more of the search space than any manual method. To do so, these automatic methods need to be able to make RAID level selection decisions, so the question arises: what is the ....

....as input (1) a workload description and (2) information about the target disk array types and their configuration choices. The solver s output is a design for a storage system capable of supporting that workload. In the results reported in this paper, we use our thirdgeneration solver, Ergastulum [5] (prior solver generations were called Forum [7] and Minerva [1] Our solvers are constraint based optimization systems that use analytical and interpolation based performance models [3, 7, 18, 23] to determine whether performance constraints are being met by a tentative design. Although such ....

[Article contains additional citation context not shown here]

E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. Ergastulum: An approach to solving the workload and device configuration problem. Technical report HPL--SSP--

No context found.

E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. Ergastulum: an approach to solving the workload and device configuration problem. Technical report HPL--SSP--2001.

No context found.

Eric Anderson, Mahesh Kallahalla, Susan Spence, Ram Swaminathan, and Qian Wang. Ergastulum: an approach to solving the workload and device configuration problem. Technical report HPL--SSP-- 2001.

No context found.

E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. Ergastulum: an approach to solving the workload and device configuration problem. Technical report HPL--SSP--2001.

No context found.

Eric Anderson, Mahesh Kallahalla, Susan Spence, Ram Swaminathan, and Qian Wang. Ergastulum: an approach to solving the workload and device configuration problem. Technical report HPL--SSP-- 2001.

No context found.

E. Anderson et al. Ergastulum: An Approach to Solving the Workload and Device Configuration Problem. HP Laboratories SSP technical memo HPL-SSP-2001-05, May 2001.

No context found.

E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. Ergastulum: An approach to solving the workload and device configuration problem. Technical Report HPL-SSP-2001-05, HP Laboratories SSP, May 2001.

No context found.

E. Anderson, M. Kallahalla, S. Spence, and R. Swaminathan. Ergastulum: an approach to solving the workload and device configuration problem. Technical Report HPL-SSP-2001-05, HP Laboratories Storage Systems Program, 2001.

No context found.

E. Anderson, M. Kallahalla, S. Spence, R. Swaminathan, and Q. Wang. Ergastulum: an approach to solving the workload and device con guration problem. Technical Report HPL-SSP-2001-05, HP Laboratories, 2001.

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