The Office of the National Archives and Records Administration (NARA) has responsibility for the preservation of permanently accessioned electronic records (ER). NARA’s 10-year recopy plan requires the migration of their ER holdings to denser magnetic media incorporating enhanced storage capacity and improved data transfer rates. The corresponding study determined the accelerated environmental performance of these denser magnetic media, specifically the IBM 3590 and both Quantum’s Digital Linear Tape (DLT) and Super DLT products. The study compared the life expectancy profiles of the corresponding tape media and was the basis for evaluating a denser magnetic medium for NARA.

1. Aging (VSM Studies)

Neither IBM nor Quantum manufactures their own media for these tape drive products. The results included the data from 2 suppliers of IBM 3590 tape, 2 suppliers of DLT Tape and a single supplier of SDLT tape. The magnetic parameters of Ir and Is were evaluated for a minimal change. The premise being that permanent changes in magnetic characteristics are detrimental to performance and the most favorable product/media is one where little or no change takes place. In all three temperatures reported (40, 50 & 80C) and associated humidities, the Digital Linear Tape media exhibited the least change. Similar results were reported for the Life Expectancy (LE) based on Log T _0.1 data.

2. Life Expectancies

The LE projection was derived for the different evaluated media on the basis of data collected for the ERROR study. Typical LE’s are reported based upon the media performance of the three tape types exposed to the three different environmental conditions. The effectiveness of the different Storage Device’s ECC efficiency is observed via the differences in LE’s reported. These results were consistent with the error reporting and trend analysis also used in the study.

3. Binder Hydrolysis

The binder hydrolysis extracts from the evaluated tape media subjected to GC-MS analyses revealed minor changes in the compositions of their extracts over time. Samples subjected to an extreme condition of 100 ^oC and 100% RH revealed the same set of molecules seen at the beginning of the study. Tapes subjected to these most severe aging conditions however did show significant signs of physical breakdown after approximately 5 days at 100 ^oC and 100% RH. Specifically, the tapes became stiff and brittle with small particle flakes falling off of the substrate backing. The Digital Linear Tape tapes were amongst the slowest to show visual signs of degradation, although these too degraded eventually.

 

4. Physical Properties

The changing values of Resistivity with environmental exposure were substantial but within specification limit for these tape products. Under all circumstances, it is preferable for no Resistivity change to occur. Tapes were ranked with respect to the degree of resistive change on exposure to the test environments.

Changes in tape friction could be detrimental to performance and the most favorable product/ media is one where little or no change takes place. The changes in friction coefficient resulting from time exposures are reported for the three environmental conditions. Tapes were ranked with respect to the least change in friction coefficient.

The ERROR study reported tape and drive error performance during both the initial write and all subsequent reads. For the purpose of this study the data being written and read was random and designed to induce tape streaming for all products tested.

The tape error evaluation was based upon noting tape errors at receipt and error changes resulting from environmental exposure. Parameters of interest during write/read were correctable(s) error, retries, non-correctable(s) error and the error-rate (B*ER).

The efficiency of the ECC for each Storage Device is demonstrated in its effect on the B*ER. In both the write and read cases, the B*ER for all the test media was used for the media rating.

In addition to the error quality during first write and read, error changes were also reported as a function of environmental exposure. The resulting data was consistent with the above findings.

The frequency response of each tape sample was compared to its original after each interval of Temperature/RH exposure. The data was summarized throughout all test environments and the data was used to rate tape stability.

Copies of the full Report and Appendix are also available for purchase in either a printed bound volume or CD. Requests for pricing can be obtained from Arkival Technology at NARAstudy@ARKIVAL.com. Please specify Hard copy, CD or both and the number of copies requested. Non-profit discounts are available.