LTO Automation Realities.

Linear Tape-Open continues to gain momentum

When three of the biggest names in data storage--HP, IBM, and Seagate--said that they were forming a consortium to develop a new tape technology, it was an announcement that the industry did not take lightly. The result was an unprecedented list of vendors signing on to support this yet untried tape technology. With the tape market being a fairly mature and resilient market, why did so many vendors commit resources to support this new technology? And, more important, how safe is their investment?

Linear Tape-Open (LTO) technology was unveiled early in April 1998 by Hewlett-Packard, IBM Corp., and Seagate Technology. This new tape architecture would consist of two formats: Ultrium, implying ultra-high capacity, and Accelis, emphasizing access speed over capacity. The concept was clear from inception: develop an open tape standard supported by multiple vendors where compatibility would be ensured regardless of where the user purchased their drive or media.

IBM was the first to announce availability of product with its 3580 Ultrium tape drive in August 2000. HP and Seagate followed suit shortly thereafter, announcing shipment of their respective products in September 2000. By this time, nearly 30 companies had executed LTO license agreements. In addition to HP and IBM, this list reads like the "Who's who" of tape storage automation: ADIC, ATL, Benchmark, Breece Hill, Exabyte, Overland Data, Plasmon, Qualstar, and StorageTek.

According to IDC's 1999 Tape Drive Report, the midrange tape market is estimated to achieve a 15% CGR through 2003. Automation is a key contributor to this growth. The study notes that while 35% of tape drives are automated today, it is expected that 70% will be automated by 2003. With the explosion of data confronting organizations today, coupled with the migration to Storage Area Networks (SANs), higher capacity and faster drives are demanded in order to control this proliferation of information. Yet this is not a market without competition. Tape formats available for consideration include Sony's AIT, Quantum's DLTtape, and Exabyte's MammothTape technology. So what is it about LTO that makes these automation vendors allocate resources to support this new technology?

LTO's technology features include:

* Implementation of linear serpentine technology--for high track densities.

* The implementation of concurrent data channels--first generation enables eight data channels, future versions delivering up to 16.

* Enhanced servo and head technologies--critical for high track densities.

* Timing-based servo--monitors and controls the position of tape head to ensure accuracy and avoid overwrites.

* Intelligent data compression--maximizing both capacity and transfer rates.

* Designed for automation--soft load, low power consumption, and intelligent automation interface.

Specific to tape automation vendors, several of LTO's capabilities make it a very compelling product. LTO's new serial ACI (Auto Changer Interface) provides enhanced load/unload communications to the libraries' robotics. The soft load mechanism of LTO means that loading of the media is achieved without a reliance on motors or levers, making LTO a true auto-loading drive.

Cartridge memory, another feature, results in reduced media inventory time in an automated environment. Physical size is another attribute library vendors exploited. Given That LTO data cartridges are smaller than DLTtape cartridges, vendors can populate their LTO libraries or autoloaders with more cartridges and the massive amount of storage provided by each cartridge and the expected growth path than with DLTtape, given the same space requirements. The resulting storage density can prove to be a critical point of distinction for LTO in space limited data centers. The consortium's commitment to automation is further evident in the design of their cartridges. Not only are they contoured on the sides for easier gripping automation robotics, each incorporates gripping pads mounted on the sides of the cartridges to improve retention and alignment.

Does all of this ensure that LTO is an instant success? There remain several hurdles for the consortium to clear before the answer to this question is revealed. Will the promise of multiple vendor drive/media interchange be fulfilled? Will the independent soft ware vendors (ISVs) incorporate support for the myriad of standards the consortium has proposed--from cartridge memory to enhanced leader block data sets?

LTO's momentum is there. The industry support is present. Automation products are available. Ultimately, the verdict establishing its ultimate success or failure rests with the market. Deliberation is underway. Now we simply wait and see.

Tim Studebaker is the vice president of alftomation at Exabyte Corp. (Boulder, CO).