Automatic generation of behavioral hard disk drive access time models

Crume, Adam; Maltzahn, Carlos; Ward, Lee; Kroeger, Thomas M; Curry, Matthew

doi:10.1109/msst.2014.6855553

Cited by 1 publication

(3 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For future work a more elaborate approach for exploitation of the periodic storage performance could be used. Crume et al had success on access time prediction for single hard drives using Fourier analysis [4] or additional sinusoids as input for ANNs [3].…”

Section: Discussionmentioning

confidence: 99%

“…A Fourier analysis is used to determine the most important frequencies, which are then used as input information for the ANNs. In a following publication [3] they move away from Fourier analysis, but instead use additional sine waves as input for the ANNs. This seems to be a promising approach for predicting access times of single hard drives, where the rotational characteristics of the magnetic disk are an essential consideration.…”

Section: Prediction Of I/o Performance With Annsmentioning

confidence: 99%

“…The measurements were done on the supercomputer Mistral (phase 1 system) of the DKRZ (Deutsches Klimarechenzentrum) 3 . The phase 1 system of Mistral operates with the parallel distributed file system Lustre (Lustre 2.5, Seagate's edition) and consists of over 1500 compute nodes and 30 Petabyte storage.…”

Section: Test Systemmentioning

confidence: 99%

See 2 more Smart Citations

Predicting I/O Performance in HPC Using Artificial Neural Networks

Schmid

Kunkel

2016

JSFI

View full text Add to dashboard Cite

The prediction of file access times is an important part for the modeling of supercomputer's storage systems. These models can be used to develop analysis tools which support the users at integrating efficient I/O behavior. In this paper, we analyze and predict the access times of a Lustre file system from the client perspective. For this purpose, we measured file access times in various test series and develop different models for predicting access times. The evaluation shows that in models utilizing artificial neural networks the average prediciton error is about 30% smaller than in linear models. One phenomenon in the distribution of file access times is of particular interest: File accesses with identical parameters show several typical access times. The typical access times usually differ by orders of magnitude and can be explained with a different processing of the file accesses in the storage system-an alternative I/O path. We investigate a method to automatically determine the alternative I/O path and quantify the significance of knowledge about the internal processing. It is shown that the prediction error is improved significantly with this approach.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Prediction Of I/o Performance With Annsmentioning

confidence: 99%