Follow Datanami:
November 2, 2011

Tape, It’s Not Just for Backup Anymore

Nicole Hemsoth

The term “exascale” has been the buzzword of choice in the high performance computing industry for a couple of years now, but as data sizes continue to expand, one can bet that large-scale organizations are starting to scramble for storage solutions to match their exabyte-range stores.

EMC, IBM and others have been pitching the value of tape for big data, but this has been seen as an archive and long term backup solution for most users. Furthermore, these vendors have used the term “big data” in a rather loose way, citing analyst figures that portend huge spikes in the amount of data organizations are gathering and storing long-term. The word “exascale” has yet to enter the picture for these other storage vendors (who, by the way, certainly don’t specialize in tape exclusively).

This week SpectraLogic announced a string of new products and featured aimed at bringing new big data business into the tape fold with their T-Series line. More specifically, they have set their sights on the exascale era, being among the first to pitch the role of tape in the coming exabyte era.

The company says that they have produced a “suite of advancements that meet the complete feature set, capacity and performance needs of the new market of exascale storage.” Such a solution would appeal to the expected camps in the HPC world, but SpectraLogic is also making an enterprise play by pointing to the needs an increasing number of businesses are facing in terms of their massive storage needs.

SpectraLogic claims that their T-Fininty Exascale Storage line represents the “world’s highest capacity storage system” which they describe as an enterprise tape library that can swallow more than 3.6 exabytes of data. The company has boosted its ability to manage even larger datasets with enhancements to the BlueScale 12 management system. They say that “a single T-Finity library will now expand to 40 frames for a capacity of up to 50,000 tape cartiridges” and that up to eight libraries can be strung together using the company’s “Skyway” technology to ramp up to 400,000 tape cartridges.

These are strong claims, even from a company that is one of the leaders in the tape storage camp. Still, the question remains, where does tape fit into the overall enterprise big data ecosystem when many business with real-time processing and major I/O needs view tape as an archive-only technology?

According to Addison Snell of Intersect360 Research, “tape continues to be an important part of scalable storage infrastructures because it offers more bytes per dollar, per watt, and per floor tile than disk. Also, tape lasts longer, so you don’t have to migrate data as often.”

Snell says that “recent technology developments are increasing tape’s strategic importance. Most notably, LTFS and the ability to export a file system over tape lets users see their archives with a file system interface.” He claims that this will give organizations the ability to manage active archives, in which the primary copies of data are on tape, not disk.

Snell feels that SpectraLogic’s series of announcements recently have all had the common theme of bringing large-scale enterprise data center features to tape libraries and “the tape is easier to manage and more reliable than before.” Furthermore, since it can scale to exabytes, this positions the company well in the era of big data.

In addition to their T-Finity and BlueScale 12 management software boost, the company says there are other elements critical to achieving exascale-capacity storage. These requirements are met by its announcement of a few other products this week, including their Bulk TeraPack Access Port (TAP) system, which they say will allow for faster and more efficient importing and exporting of tapes. They say customers can expect to import or export 500 TBs in just minutes for an extra cost on top of their enterprise tape libraries, including the new T-Finity as well as their T590.

Another key element in their exascale enhancements include the addition of their Media Lifecycle Management capabilities that extent to the IBM TS1140 tape drives, which are available with their T-Finity libraries. SpectraLogic says this allows customers to track and provide alerts before they dump into or out of damaged or full tap and performs general data integrity checks.

The company also described their TeraPack Optimized RAIT system which is archive software that they say improves “management, handling, import, export and storage, as well as up to 2x faster tape mounts of RAIT sets.” They claim this will stave off the risk of lost data in a manner similar to RAID, which is only available on disk-based systems. All of this is boosted, they say, by CarbideClean, which is part of the company’s best practices to extend the life of tap by using a carbide blade cleaning process before shipping to customers. This will not come at additional cost to customers as they have integrated this process into their production.

While SpectraLogic has a clear vision for the future of exascale storage on tape, the question that seems more pertinent here is how these technologies will be understood by customers who view tape almost exclusively as a backup technology—and nothing more. The answer lies, at least in part, in the company’s Active Archive approach.

Analyst George Crump from Storage Switzerland says that “conventional wisdom, or at least the wisdom of the disk manufacturers, is that the entire big data store needs to be online and available. The reality is that putting all data online is not a viable solution for most businesses from a cost perspective.”

He says that the answer to the question of where tape fits into big data is that organizations need to “leverage tape as a part of the access tier through the use of an Active Archive.” He says that this method allows the combination of high performance primary disk storage with secondary disk with the tape being used to create a single, fully integrated access point. This means that when analysis is needed, the data is heaped into the high performance tier with the secondary tier storing information that is critical but not reliant on high performance access. He also claims tape could then be used to store the other pertinent but non-mission critical data with the Active Archive software managing data movement across the tiers.

Do NOT follow this link or you will be banned from the site!
Share This