ACM Home Page
Please provide us with feedback. Feedback
Optimized pre-copy calibration of hard drives
Full text PdfPdf (267 KB)
Source ACM SIGMETRICS Performance Evaluation Review archive
Volume 33 ,  Issue 2  (September 2005) table of contents
Special issue on the workshop on MAthematical performance Modeling And Analysis (MAMA 2005)
Pages: 9 - 11  
Year of Publication: 2005
ISSN:0163-5999
Author
David M. Nicol  University of Illinois, Urbana-Champaign
Publisher
ACM  New York, NY, USA
Bibliometrics
Downloads (6 Weeks): 1,   Downloads (12 Months): 7,   Citation Count: 0
Additional Information:

abstract   index terms   collaborative colleagues  

Tools and Actions: Review this Article  
Save this Article to a Binder    Display Formats: BibTex  EndNote ACM Ref   
DOI Bookmark: Use this link to bookmark this Article: http://doi.acm.org/10.1145/1101892.1101896
What is a DOI?

ABSTRACT

In certain contexts a small window of time exists during which law enforcement has access to a hard-drive suspected of containing important information. Given legal authority to copy or seize this disk, a decision must be made whether to use that access time to make a copy of the disk (which may take more than an hour, depending on the size of the disk) and leave its owner unaware that it has been copied. The copying operation uses especially fast drivers that bypass normal error correction mechanisms. Therefore, for the copy to be successful it is necessary that the disk onto which the copy is placed yield exactly the same bits on subsequent reads as would the original disk. To gain confidence that the copy will be successful the copying software typically chooses some sectors at random, copies them, and determines whether their copies are identical to the original. We address the problem of quantifying the conditional probability that the disk will copy correctly given that some samples have copied correctly, as a function of the, number and placement of those samples. Our framework allows us then to choose the placement of those samples in such a way that this conditional probability is maximized.