Enterprise Tape Media Achieves Major Increase In Storage Efficiency – Virtual Tape Storage – Technology Information
Storage tape manufacturers and resellers were scratching their heads throughout most of 2000 wondering why another year of robust growth in enterprise storage was not accompanied by healthy sales of tape media. Although some declines were anticipated, the drop off in tape consumption was more severe than expected. 2000 brought a major decline in 3480/90 media sales, showed flat growth in 3590 media sales as that tape technology competed with the newer 9840 medium. Even 9840 tape growth was not as spectacular as past new tape technologies have been during their initial adoption phase.
Sales were not off due to any post-Y2K lull in the marketplace. Nor did tape media sales lag due to the popularity of storage solutions provided by high flyers like Sun or EMC. The major reason is the rapid ascendancy of Virtual Tape and the storage efficiencies that resulted from its implementation. The changes brought on by VTS are responsible for absorbing significant portions of previously sold tape capacity and for lessening recent demand for many tape media formats used in enterprise storage backup.
Virtual Tape Storage (VTS), high-speed Fibre Channel connections and centralized, shared storage have created revolutionary changes, dramatically affecting near term sales of tape cartridges, tape racks, and all tape related accessories. This, coupled with the 95 percent decline in the cost per gigabyte of tape storage, has caused major changes in the market. VTS combines the use of specialized disk storage as “virtual tape” with robotically controlled tape libraries. This integrated system provides very efficient use of “virtual” tape volumes and achieves major tape efficiency by stacking “real” tape volumes to utilize the entire unit capacity of tape cartridges in the library.
Between 1999 and 2001 the industry will ship 44 percent more enterprise tape-storage capacity. But, at the new levels of storage efficiency, these shipped tape cartridges will actually provide their users with three times more storage capacity.
Most 3480/90-tape libraries were full of tapes that were only partially (approximately 30 percent) loaded. Available tape management software was simply not capable of organizing data files so that tapes could be fully loaded. VTS software enables the stacking of files. 3590 tapes and 9840 tapes used in VTS systems are generally 85 percent loaded. Many tape libraries had numerous magnetic tape cartridges with only one or two small files on them. VTS consolidated those files, dramatically reducing the number of volumes, yet still providing for growth in capacity. In addition libraries have also gone from containing tape cartridges with less then one gigabyte capacities, utilized at only 30% of that capacity to containing 10-, 20-, or more-gigabyte tape cartridges loaded to 85-percent capacity.
If you combine that efficiency with the declining cost per gigabyte, it becomes clear why we could have dramatic increases in storage without increases–in fact declines–in tape media sales revenue. Any remaining spare capacities have to be consumed before any revenue growth comes back. That is not likely to happen before 2002. The likelihood of this lull in tape sales becomes even stronger when you consider how broad based the impact of VTS is. By the end of 2001, recent research indicates, 70 percent of all major enterprise computing sites will have installed VTS technology (Table 1).
This is not the end of tape. Tape will continue to play a major role because of the continuing growth and centralization of storage.
The average online storage requirements of a Global 2000 company in 1998 were 40 terabytes; at the current rate of growth, next year that number will be 300 terabytes. This magnitude of data storage increases will continue as the technology to convert all data into machine readable and robotically accessible bits and bytes grows. By 2005, estimates have 20% of all known data converted to computer-readable digital data. In all likelihood major portions of this data will be backed up on tape media, preserving the role of tape media in backup storage.
Centralization means that data storage is becoming a driver in systems designs. NAS and SAN are both designs that improve the ability for systems to locate, share and retrieve data. Some believe that robotically-accessible tape libraries will become the center of these storage networks. This increased centralization will also ultimately force companies to decide on one or two tape media platforms. That will eliminate the market for some tape media formats in existence today (Table 2).
Dominant Tape Media Formats
In enterprise storage, 3590 and 9840/9940 will be the formats of choice going forward. Recent data indicates that these two technologies currently share the market relatively evenly. The continuing financial and organization changes at StorageTek (9840) may ultimately shift more user decisions towards the IBM 3590 Magstar.
In the Server/Network segment of the tape storage market an exciting battle between DLT and LTO is just starting. DLT has owned this segment for some time now. The lucrative control of this market by DLT’s owner, Quantum, encouraged three competitors to cooperate and develop a strategy for this key tape market. Hewlett-Packard, Seagate and IBM have brought a “Linear Tape Open” platform to the market and encouraged all media and drive integrators to join. The collaborative assumption here is healthy “open” competition will translate into lower media and hardware prices to the end user. The LTO design is derived from the Magstar and has a long history of reliability and performance. The first generation products just released offer tape cartridges with 100gigabytes of storage. The only real competition is Super DLT. Both SDLT and LTO have suffered many delays since being announced. But as of January 2001 LTO looks slightly ahead in the race to market. LTO will not completely replace DLT, but the industry observers agree LTO will halt DLT growth and end its market dominance.
The other tape platform in this segment is Sony’s AIT. It is too soon to tell whether the emergence of LTO will prevent AIT from moving beyond its current niche. Some other tape platforms like 4mm/8mm, QIC and Travan are not growing and are likely to fade from the scene in the long run. The consensus is Server/Network storage will be an LTO and DLT world by 2003 (Tables 3 & 4).
This change in market conditions also affects the competitive lineup. In Enterprise Storage, the leading tape manufacturers and marketers are EMTEC and Imation. EMTEC, reorganized under new ownership after being shed by BASF in 1996, is still the largest magnetic-media coater in the industry and continues to have the broadest product lineup in the industry. Product offerings range from the audio video tapes for entertainment distribution and the consumer market, professional media products for broadcast and studio applications, a growing line of imaging products, as well as data media products. That versatility is proving a real strength as certain segments go through changes.
In the Server/Network tape media business the lineup of primary vendors besides EMTEC and Imation, includes FujiFilm, Maxell, and Sony.
Author’s note: EMTEC is expected to deliver LTO in 2001 and has just signed a global agreement to market the services of Front Porch Digital. Front Porch has unique analogue to digital conversion technology and has purchased the media service arm of Storage Tek. This combination of FrontPorch technology and service experience will enable EMTEC to offer the broadest media services capability tothe market by the second quarter of 2001.
Sig Tullmann is the president and CEO of EMTEC DataStoreMedia (Billerica, MA), the North American subsidiary of the Germany-based EMTEC Group, formerly BASF Magnetics.
COPYRIGHT 2001 West World Productions, Inc.
COPYRIGHT 2001 Gale Group