SNIA EUROPE I CLOUD
Cloud tape
MARCUS SCHNEIDER SNIA Europe Board member, Fujitsu
It sounds just like a contradiction. The most hyped topic of these days, cloud, the technology that will make everything cheap and easy on the one hand - and the technology of our “grandfathers in IT”, tape, the awkward one, on the other. Do we really have to find a cloud aspect on every single existing technology?
Well, these days almost all presentations about tape start with mentioning that Google didn’t lose data because they used tape and Amazon did lose data because they did not use tape. And here we are in the middle of the debate.
Let’s have a look at the simple truth in life: Cloud infrastructure typically embraces huge amounts of data. Very often, not always, you will find valuable information in it that must be protected. So we need to back it up.
As cloud is a super modern idea, it uses super modern infrastructure. Data is copied a couple of times and might be distributed around the globe. Fancy. But to be a back up, we need to do that with a time stamp, we need a “something” to know, what data has to be copied when and where and we need a mechanism to recover that data (to meet RPO objectives). And as in “real” IT ☺ a physical separation between the operational cloud infrastructure and the backup infrastructure creates an additional level of security.
So we are going to create a lot of data, almost off line, from the cloud storage we have today. Do we want to install the same expensive cloud storage again? And here we see a marvelous space for tape, in its usual role, as the last line of defense against data loss, with a very small price tag.
Sounds reasonable but there are some new challenges. Unlike in most of today’s data centers, we see a huge piece of cloud storage. It might be NAS, it might be homogenous, it might be many petabytes. How do we move all that data to tape? This is a most exciting challenge for the storage industry, to find the right approach, to find the right data movers that can transport such enormous amounts of data fast to a back up, to find the right back up logic to slice back ups up and to be able to distribute chunks of data in a meaningful way and, finally, to find the right
4 SEPTEMBER 2011 |
WWW.SNSEUROPE.COM
virtualization layer to make it easy for cloud storage to talk to tape silos. Once these challenges are addressed, we see a new breed emerging; let’s call it “Cloud Tape”.
Cloud Tape would have a second, tremendous use case, and that is archiving. Cloud might be different from traditional IT in many ways, in one way it is not: A tremendous amount of data is completely useless. Just think about all the information you and your friends dump to facebook. It does not last long; in days it becomes old and uninteresting and is hardly ever accessed again. Why does it have to stay on spinning disk for eternity? When you have Cloud Tape, a tape infrastructure that is seamlessly integrated with the online cloud storage, you just dump such dead data out to tape and safe an enormous amount of money.
As big data emerges, e.g. data that is automatically generated by machines, we see an almost unimaginable amount of data coming up that is getting obsolete even faster than today’s data. If the data is that obsolete, can we erase it? Wait a minute; we might want to run some business analytics on it later on so we would want to keep it, just in case. Well, put it on Cloud Tape and it might rest in peace, but at least not at the full storage price.
If you are a close friend of the spinning disc and have read that this is just another trial to revive a dead technology and everything, including space travel and deep sea diving, can be done with disks. I wouldn’t argue, I am not a diver. But we have to face some physics: flash and disk have exploited the potential density of information on these media to quite an extend. Tape offers the most potential to significantly increase density in the future. Facing the tremendous data growth we love to talk about, more density will not be a luxury it will be a basic need.
This is another reason why we need to create Cloud Tape rather sooner than later.
In any case, I am very curious what will happen the next time when data is lost in cloud and whether it was properly protected. I am interested in hearing your thoughts.
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44