LTO tape versus Optical Disk Archiving (ODA) for long term data protection and retention
Eighty percent of your data is rarely or never accessed after 90 days of creation. However, the business value of much of this data means it still has to be retained and readily available over extended lengths of time.
Today, we’re taking a look at LTO tape technology in comparison with Sony’s Optical Disk Archive solution for long term archiving.
One of the compelling benefits of tape technology is its durability. Tape is a proven, secure and reliable medium for the long term preservation of data. LTO Ultrium is rated for at least 30 years archival warranty life and a number of suppliers, including HPE, have published data from accelerated ageing tests to demonstrate the validity of these claims.
In absolute terms, ODA is rated for even longer retention periods - up to fifty years according to Sony. So does that mean ODA is a better choice than tape for preserving data for decades to come? Not necessarily.
Economies of scale and cost of ownership
It’s true that using tape, data would need to be migrated between tape generations at some point in the future, more frequently than perhaps would be required with ODA.
But each time this happens, fewer tapes will be needed. LTO-10 is slated to be 48 TB native per cartridge and is probably less than a decade away. So not only is this tape management ‘overhead’ not as great as people might think, LTO will also accommodate the vast amounts of digital data that will be created moving forward.
Meanwhile, what does the roadmap for ODA look like?
According to Sony ODA will reach 6 TB in its third generation, by which time, LTO will be well on the way to 120 TB per media unit (compressed).
LTO Ultrium Roadmap
Data is increasing at 40% per year and 90% of it is unstructured (IDC). So a 1 Petabyte archive today will be around 5.5 PB in five years time.
Today, that 1 PB will need about 300 x 3.3 TB disks versus 170 LTO-7 tapes. A 3.3 TB ODA disk costs around $200, whereas an LTO-7 tape (6.25 TB) is around $150 per cartridge. In cost per TB terms, LTO-7 is 60% cheaper.
In 5 years, if we assume LTO-9 is available, a customer would need almost 900 x 6 TB ODA disks versus just 215 x 25 TB LTO-9’s!
If these differentials are broadly the same for future generations, the difference in media expenditure in the future is going to be enormous! Tape’s Total Cost of Ownership ratio will be many times lower than that for ODA.
Using today’s pricing, 900 disks at $200 each would be $180,000! 215 tapes at $150 would be just $32,250. That’s almost $150,000 lower media costs in absolute terms and 80% cheaper in cost per terabyte values on the media alone.
Open standards and multi-vendor support
One of the primary considerations has to be how easy it would be to migrate to a new supplier if something happened to an incumbent provider in the future.
Today you can buy LTO Ultrium technology from HPE as well as IBM, Quantum, Dell, Oracle, Spectra,
Overland Storage, Qualstar, Tandberg and other system integrators. One can buy media from pretty much all of those companies too.
ODA, on the other hand, is proprietary technology (like AIT) and dependent on Sony continuing to fund its R&D program. What will customers do in the future if Sony decides to suspend or cease ODA development, as it once did with AIT tape? There is no other alternative whereas for LTO there are many competing offers in the market place. Furthermore, the convergence of Enterprise and LTO Ultrium tape technologies arguably secures ongoing tape manufacturing for the foreseeable future. I don’t believe that one could say the same is true for ODA, at least not with anything like the same degree of confidence.
Generally speaking, competition means lower prices. LTO is a competitive, open standard market. ODA isn’t. Once the customer buys ODA, they are locked into a proprietary solution and dependent on a single vendor with respect to pricing, upgrade terms and roadmap delivery.
Although archive data may not be accessed frequently, it still needs to be stored in the first place and large scale data archiving needs careful structuring to avoid network overhead and disruption. The transfer rate of 250 MB/sec for ODA is quite slow versus LTO-7 today and very slow compared to LTO in the future. Tape excels at streaming huge quantities of data.
The next publicly communicated iteration of ODA is proposed to increase performance to around 375 MB/sec when it reaches 3Gbps. In comparison, LTO-9 is projected to have 708 MB/sec and LTO-10 is proposed to have 1,100 MB/sec, nearly 3X the speed and thus a 3X advantage in terms of reducing archival windows.
Furthermore, the LTO Ultrium format supports AES-256 hardware-based encryption, which means data can archived and secured ‘on the fly’ without the need for either encryption software or a secondary (more cost) appliance.
Tape seems to be more durable in lower recommended temperatures and humidities, ODA gives your customer equivalent ‘margin’ at higher temperatures and humidities.
In reality, ANY storage media, optical or tape, will ideally be kept in a cool, stable environment so there does not seem to be a lot to choose between the ‘long term recommended’ values for ODA above versus LTO tape.
In the end, it’s hard to see past the fact that tape technology has the support of a much greater number of major IT vendors and, on paper, a colossal advantage in terms of TCO.
Once you factor in the costs of buying sufficient quantities of media to cope with spiraling increases in the amount of data to be archived, the benefit of LTO Ultrium tape over Optical Disk Archive seems quite compelling.
Contact your BackupWorks.com Account Rep today and
find out if LTO Tape is right for your data storage environment. - 866 801 2944