The future is bright for HPE StoreEver tape!

The future is bright for HPE StoreEver tape!

Earlier this year, HPE StoreEver announced a new line up of LTO-9 Ultrium tape solutions: standalone tape drives (internal and external models) as well as entry level and midrange tape automation systems.??You can find full product details on hpe.com/storage/storeever and more tape-related information at tapetember.com.

Over the last few weeks, I have been explaining the reasons for the ongoing value of LTO tape; what we at HPE StoreEver describe as “lasting innovation”. That means ‘lasting’ in terms of tape technology’s durability and undoubted market success, and ‘innovation’ reflecting all the ways that tape has reinvented itself as storage needs and data protection challenges evolve.

In this article, I want to recap some of the key points from the blog series while looking ahead to see what the future may hold for tape and its role in data protection.

Lasting innovation - with more to come!

In terms of lasting innovation, I think it’s clear that tape is an extremely innovative storage solution which is constantly reinventing itself as data protection needs evolve. Compared to hard disk storage, tape capacity growth is far more likely to match the 40-50% compound annual growth rate (CAGR) of data, fuelled by new technology like AI, 5G and billions of sensors in the Internet of Things. A one kilometre length of LTO-9 tape has a relatively low areal density, so there is ample scope within the format for future capacity points to keep up with high data growth rates.??The LTO roadmap presently extends to 480 TB at LTO-12 with prototypes of 550 TB in existence and 723 TB as the projected potential of today’s tape magnetic media recording technology.

Green your storage with StoreEver

Just as important as the underlying technology, however, is the future environmental impact of your storage and data protection choices. What makes tape quite different from other secondary storage systems - for example, disk arrays or object storage servers - is that it is offline by design. That means you can remove tapes from the host tape library and store them in a secure and remote location like a vault or warehouse.??At this point, disconnected and essentially inert, a large collection of tapes exerts only modest environmental demands. There is no need for the round-the-clock, industrial-class temperature and humidity controls that data centres deploy to prevent their intensive thermal workloads from overheating. Studies have suggested that an organisation with 10 PB of archive data, growing at 35% per annum, would produce 87% less carbon dioxide (specifically related to storing that data) if it stored the inactive data on tape storage instead of using disk storage.

But when it comes to preserving data in the future, even a 3 PB archive today will be around 87 PB in ten years time with a 41% compound growth rate. Assuming broad parity in street pricing (e.g. that a 100 TB Enterprise SATA disk in future will cost roughly the same as a 20 TB SATA disk today - about $600 street price), the acquisition of those 870 disks or 87 PB of capacity would cost over half a million dollars!?

Consider cost of ownership throughout the data lifecycle!

In comparison, (and again assuming that tapes in future are priced at broadly similar levels to today’s LTO-9 cartridges, around $200 per unit), the media cost of a??87 PB archive set would only be $23,200 if each cartridge had a capacity of 730 TB as projected by INSIC.??This is a saving of 182% on the cost of those 870 hard disks even before considerations like power, cooling and emissions.?

Of course, the cost of the disks themselves is just one factor to consider: total cost of ownership (TCO) analysis would also consider maintenance, floor space, hosting and egress (for the cloud) as well as power and energy costs (operation, cooling etc). Over ten years, analyst ESG has estimated that you could achieve savings of more than 80% over a ten year period using HPE StoreEver compared to all-disk or all-cloud solutions.

It’s worth emphasising that although the cloud, in particular, may have very low barriers to adoption, businesses clearly need to consider the longer term cost impact very carefully. Cloud service charges are a rental charge that needs to be paid, month after month, year after year, even before the additional levy for data retrieval.

The ultimate last line of defence against cyber attack

Another key benefit of LTO tape is allowing you to store your data offline and behind an air gap so that it cannot be corrupted by ransomware. Ransomware is in no shape or form a ‘new’ threat. What has perhaps changed is the level of public awareness following a spate of high profile ransomware incidents in North America and Europe during the last couple of years.?

Personally, I think we place too much trust in cloud-based technology and forget the physical realities that lie beneath its miraculous abstraction.??The cloud is not a cloud.??It’s a network made up of different technologies - compute, storage, infrastructure, software - all of which have a key role to play and none of which are infallible.??This is why I believe everyone needs some tape in addition to the other complementary technologies that you deploy to build a circular fortress that guards your data on all fronts.??Essentially, my opinion is that the ingenuity of cybercriminals will always find a way to breach the ingenuity of cyber defences. That being the case, it’s a zero sum game to try and argue that your own particular storage technology is better than someone else’s (as smaller, single platform vendors are prone to do). In a storm you wear layers of technical clothing because each has a specific role and benefit. The same is true of storage solutions to help you keep your data and infrastructure as safe as can be.??

In a connected environment, the attack surface is so much wider and multi-faceted. Criminals will never stop learning and exchanging knowledge about how to target businesses and their intellectual property. Hackers only have to get it right once whereas you must consider for every possible vulnerability across your entire infrastructure. Why use tape? If it’s that inexpensive and gives you extra peace of mind, why not use tape? Multiple government agencies recommend backing up your data, system images and configurations and keeping copies of mission critical information offline.??“Offline” for most practical purposes means on tape. And in my view, everyone needs it.

The myth of tape reliability?

Of course, when something has been around as long as tape, urban myths begin to develop around its performance and practical value in the field. A number of years ago, it was an easy Google search to uncover the apparently damning statistic ‘Gartner report that 71% of tape restores fail’. The only problem was that Gartner never said anything of the sort. It was fake news! In reality, even the largest and most demanding customers report outstanding reliability metrics from tape. The US National Energy Research Scientific Computing Center (NERSC) migrated over 40,000 tapes a decade ago and only encountered issues with 35 of them - a 0.00009 percent error rate. Today, CERN Laboratories in Switzerland store vast amounts of some of the most rarefied scientific data in the world (from the Large Hadron Collider project) using LTO Ultrium technology.

When we look at Bit Error Rate (BER) today, what we see is that not only has tape reliability been improving through the generations, but that LTO-9 tape reliability is orders of magnitude better than that of SATA hard disks. That’s essentially due to the laws of physics, which as all Trekkies know you cannot change! In order to increase the capacity of any magnetic media, tape or disk, you need to reduce the space occupied by each bit. This can only be done by making the magnetic grains or particles used to represent each bit smaller and increasing the areal density. Beyond a certain point - the superparamagnetic limit - it becomes impossible to make the particles smaller and have them retain their magnetic state (which is how data can be understood and read back).

LTO has superlative Error Correction technology which allows the tape drive to achieve the lowest possible error rate. With LTO, the areal density is low enough to allow data to be written in relatively large blocks, which means larger and more efficient ECCs. In HDD, because areal density is high and bit and block sizes are smaller, it’s significantly harder to get the same reliability.??As a consequence, the BER of LTO-9 is 10^19 versus 10^15 for an Enterprise SATA disk: this means an LTO tape drive is ten thousand times less likely to encounter an unrecoverable error than a HDD.??In real money, that means one uncorrectable error for every 10?petabytes?you write to LTO-9, versus one every 10?terabytes?for disk.

The reason why LTO has more areal density potential is simply because its recording surface is a kilometre long, whereas a hard disk platter is only three and a half inches in diameter. Even with their undoubted engineering expertise, HDD manufacturers can no longer easily increase the capacity of individual HDDs without deploying booster technologies like HAMR and MAMR to increase the underlying areal density.??

Everyone needs tape!

And that brings me neatly to my conclusion. My aim with this blog series hasn’t been to position tape as?The?Answer (to)?Practically?Everything. It isn’t and there are lots of storage and data protection challenges for which tape would be a strange and likely inadequate option. But equally, there are many situations where other technologies are thrust at customers as a?tape replacement?when they are nothing of the sort.??Tape can still offer organisations of every size and shape compelling benefits in terms of cost, scalability, security and sustainability - benefits that other storage solutions still cannot match or replicate.??In these areas, there is nothing on the horizon that looks likely to “out tape” tape.

For that reason, I am optimistic that by 2030 it is entirely feasible that a single LTO data cartridge could hold well over 700 TB. Prototypes of a 550 TB LTO cartridge based on Strontium Ferrite (SrFe) particle technology have already been demonstrated and the Information Storage Industry Consortium (INSIC) technology roadmap projects cartridges with 723 TB could be delivered by 2030.??HPE StoreEver LTO technology is well positioned to be a superlatively dense and low cost storage medium for preserving the vast quantities of data we will see in the Zettabyte Era.??

I hope you have found this series of articles interesting and useful. Thanks for reading and feel free to continue the conversation in the comments here on LinkedIn or by following me on Twitter @tapevine.

要查看或添加评论,请登录

Andrew Dodd的更多文章

社区洞察

其他会员也浏览了