Cartridge Memory Stubbing for Augmented Tape Directory Content (17-Sep-2009)

Cartridge Memory Stubbing for Augmented Tape Directory Content

FIG. 1 shows tape cartridge 100 comprising cartridge body 101 and a door 106 from which tape exits the tape cartridge for reading and writing data in tape drive 404 (FIG. 4). FIG. 1 shows that tape cartridge enters the opening 406 (FIG. 4) of tape drive 404 along direction 107. Tape cartridge 100 has a wirelessly accessible cartridge memory 103 on printed circuit board (PCB) 105. The capacity of this cartridge memory 103 is highly limited, per Figure 2. Tape cartridge 100 could hold magnetic, optical or holographic tape. Cartridge 100 could equally be an optical disk cartridge holding one or more of the following disks: CD, DVD, HD-DVD, BD, magneto-optical, holographic and phase-change.

101

Figure 1. Isometric of Tape Cartridge with Cartridge Memory

Figure 2. Description of Cartridge Memory

This article proposes that the cartridge memory 103 holds a stub 402 which locates the actual directory contents 412 of the cartridge memory. This stub is architected as the concatenation of a WWPN (Worldwide Port Name) of a controller 410 or other storage (such as a solid-state drive) containing the directory and a specific address to that directory within the controller or other storage. The WWPN is a 16 digit hexadecimal number, which means that 8 bytes are needed just for the WWPN of the controller or other storage holding tape directories. Additional bytes are needed for the address A of the particular directory within that controller or other storage, and the concatenated pair would look like WWPN:A

1

100

103

105

107

106

(concatenation of a Worldwide Port Name and Address

A

)

. This address ma

y

alternatel

y

be the di

gest L

of a hash algorithm as shown in Table 300 of FIG. 3, and then the directory would be located by WWPN:L (concatenation of a Worldwide Port Name and Hash Digest L).

Table 300, FIG. 3, teaches a comparison of five major hashing algorithms, MD5, SHA-1, SHA-256, SHA-512, and Whirlpool, column 302. It is these algorithms which identify candidates for data deduplication. In column 302, MD5 stands for Message-Digest Algorithm 5 and SHA stands for Secure HASH Algorithm. Column 304 gives the length of the digest, L, which is a fixed number of bits of output of each hash algorithm. Column 306 gives the processor cycles per byte of data hashed, which is the metric of performance. Column 306 comes from "Performance Analysis and Parallel Implementation of Dedicated HASH Functions," Junko Nakajima, Mitsuru Matsui, (Mitsubishi Electric Corporation, Japan), Proceedings of EUROCRYPT 2002, pp. 165-180, 2002, http://www.iacr.org/archive/eurocrypt2002/23320159/hash.pdf . Column 308 is the normalization of column 306, to show that MD5 is a fast hashing algorithm, with the fewest cycles per...