 from the Sands Convention Center, Las Vegas, Nevada. Extracting the signal from the noise. It's theCUBE, covering AWS re-invent 2015. Now your host, Stu Miniman. Welcome back to theCUBE, SiliconANGLE TV's live production of AWS re-invent 2013. I'm Stu Miniman with wikibon.com. Excited to have on the program, Mylon Thompson-Bukovec, who's the general manager of Amazon S3. Mylon, thank you so much for joining us. Stu, it's a pleasure to be here. Thanks for having me. All right, so Mylon, there were, as always, a lot of announcements here at the show. Really good excitement, things that people are kind of cheering for. And, you know, storage had a pretty prominent role in Andy Jassy's keynote and some of the announcements. Some of them, actually we've been before the show and at the show. So can you kind of walk us through it's the S3 infrequent access, snowball and the rest of the announcement suite? That's right. So it's a big show for us. There's a lot that we do in storage services and one of the things that we're pretty careful about doing as with all of the AWS services is listening pretty carefully to what customers ask for and doing whatever we can to deliver that as fast as we can. So here I have with me one of the new things that we launched. I'm very excited to introduce you, Stu, to a snowball. This is the Amazon AWS import-export snowball. This was one of the things that we launched a couple of days ago. And it really addresses a need that we heard from customers where they had hundreds of petabytes all locked up in older data centers. They didn't have the networking in place to make it easy to get the data out and into a cloud environment where it's so much cheaper to store. So those customers didn't want to incur that one-time cost to go amp up the network for a data center that they didn't really need to have it for anyway. And they asked us to help. So we built the snowball, which is a physical appliance. As you can tell, we built it to be very rugged, tamper-proof. And this device, which is about 50 pounds, holds 50 terabytes of storage. And so you would go up into our management console and you would essentially order snowballs. And the snowballs would arrive at your facility. You would go ahead and plug it in. We provide, as you can see, all the cables and all the cords for you. You would plug it into your network. You would turn it on and you would start the process of loading the storage that you have on your on-premise over your existing network into that snowball. So this is pretty exciting for customers. Makes it very real. You can order multiple snowballs at a time. There's a 10 gigabit connection network connection on this so you can daisy chain as many of these up. If you wanted to get, say, a petabyte of storage off, out of your data center, into a snowball in under a day, you would get maybe 20 of these snowballs. You would line them up and you'd have that petabyte of storage on your collection of snowballs and send it off. I mean, Maylon, bring up a really good point because one of the challenges is moving data is difficult. We have the laws of physics and, right, that very expensive if we wanted to have a really big pipe to be able to do that, so. Yeah. So I mean, Stu, one of the most exciting things about this is I feel like we've leveraged all the learning that we've had for Amazon retail to make the best shipping experience possible. So once you get done with the data transition or the data transfer, you just power down, you close up the device and we have an e-ink shipping label that automatically updates with the address to send back to the Amazon facilities where we'll go take all the data off that snowball, load it into either S3 or Glacier and then erase the snowball on your behalf. So what's really exciting about this is all you have to do is set it out with your UPS pickup and your UPS pickup now. We'll just take that snowball and just take it away to Amazon and there goes your petabyte of storage. Yeah, that's impressive. I mean, I think of us as consumers, everybody gets excited when the box with the little, you know, Amazon logo shows up. Now your data center people are going to be excited. When this box shows up, you got a Kindle sitting on front of it so that label with the e-ink technology is pretty cool and it tracks, you know, where your information is. Can you talk a little bit about just kind of the ruggedness and security of it because, you know, the physical, you know, we're worried about as well as the data itself that's sensitive information. That's right. We built this case to be very tough. It's very rugged. It's tamper proof. We actually have a hardware module inside the snowball that tracks if anybody tries to tamper with it. When the data is coming out of your data storage and going onto the snowball, it's encrypted with 256-bit encryption before it goes into the snowball. And all of the keys, the keys used for that encryption are never on the device itself. They're actually stored in Amazon's key management service in a very secure way and used with the encryption and decryption process. So this device that you see is both tough. I have to say our engineering team had a lot of fun with the testing of it because we got to drop it from very high places and see what happens. So this device we think is just going to unlock a lot of storage. And it's really the result of customers, a customer saying, I have a problem and can you help me? And, you know, the other two things that we launched this week. If I can, a couple of questions on the snowball. First of all, logistically, how long does it take to get it? How much does it cost to use this service? So the snowball itself takes a couple of days to arrive. By default, we've set it up for its two-day shipping. Just like everything else on Amazon, right? You can see how that makes sense. Now you can change that shipping at any time. The shipping is paid by you. So you can overnight, you can make it longer. The pricing of this is $200 per job. And then we won't charge you for the day that you receive the snowball. You get 10 days to go and transfer the storage onto the snowball and then ship it back to us. So we don't charge you the day it arrives and we don't charge you for the day that it leaves. You have 10 days to do it. If you need it for longer, it's only $15 a day to keep the snowball onsite a little bit longer. And one thing I did want to say is that the snowball today at launch, it supports the movement of data from on-premise into Amazon S3 or Glacier. It does not support the move of Glacier for data out of S3 and Glacier just yet, but that is going to be coming soon. We want you to be able to set up a job where we take the storage that you have on Amazon Web Services and be able to put it in a snowball and then ship it to wherever you would like. Yeah, and thank you for clarifying, because that's a point I've gotten a lot from the community is a lot of these services, it's easy or free to necessarily get it in, but if I want to get it out, there's a penalty on that. Is that, do you get feedback from your customers on that? Is there concern about making it easier to get my data out of Amazon? Yeah, the thing we hear the most from customers is it's basically surprised at how easy it is to get out. So we don't charge for deletes, for example. That's very core to us. If you have your data on Amazon S3 or Glacier, you should be able to delete it, you should be able to get it out. And in fact, the snowball makes it even easier to do that. All right, great. Thank you, snowball, really interesting. A lot of people, I'm sure, touching it, getting to know it, I lifted it, it's not that heavy. And yeah, some really, it's nice to see kind of some of the various pieces of Amazon working together, which sometimes we want to talk about how you work with the various services. But first, finish out kind of the announcements that you have, IA is a big one. Talk about how pricing of S3, IA and Glacier is working these days. I'm pretty excited about standard infrequent access. Again, just going back to how we think about storage. We've had standard storage out there for a while and customers love it, but what they told us is that there's a lot of storage that they have and they know that it's very infrequently accessed and they want a special solution for that. So what we did is we engineered a new storage class that we call infrequent access that is 55% cheaper than standard and really purpose built for data that you don't access very often. So in this storage class, we introduce a data retrieval charge. So if you don't retrieve that data very often, you just get all the benefits of that 55% storage discount and if you do it to retrieve it occasionally, it still ends up being cheaper. Now one of the really interesting things about infrequent access, it's the same durability, it's the same performance as Amazon S3 standard. So our idea is that even if you don't need your data very often, you can get it and you can get it just as fast as you can get it from standard. So when we think about that, think of it sort of as a warm archive. If you have data that you know you're not going to access very often, for example, a lot of our customers in the enterprise, they have data that they know is going to be susceptible to legal holds or HR holds of some sort. That's a great candidate to put in standard infrequent access because when the legal department or the HR department needs access to the data, they need it right away, but you don't want to pay the same prices as standard. In all of our data tiers, we think of the data tiering as just core to the business. So infrequent access, you can tier data from standard to infrequent access using policies and from infrequent access over to Glacier and you can do it all automatically based on things like the date of the object being created. That is so important for customers who have all their data on Amazon storage services. I had the opportunity to actually talk to a couple of customers here at the show that are using this. We're pretty excited about it because they understand when they understand their data, it's like, all right, we know our hot data that I'm going to reasonably access that pretty short term, definitely the first couple of days, then over the first 90 days, depending on I talked to one that said, oh well, I know a certain class of data actually might be in some kind of illegal information, as you said, therefore I can put it in this class, save myself a lot of money. And unlike traditional tiering was kind of tough, I had all these knobs and settings and moving things back and forth and estimating it was difficult as opposed to moving between these tiers. You know, it seems a lot simpler when it comes. That's very central. Just like it's easy for us to get data in and take data out. We want to create the notion of data movement by policy and by automation to be super easy for our customers to do. We had a chat with Andy Jassy earlier and the question was, when you're building these services, does the simplicity and the frictionless nature of it kind of get built into it? I think it comes in the name of your product line that when you're offering it, but yeah. Yeah, that's, I mean, every time we do a design review, every time we look at a new feature idea, one of the things that we always look for is, how easy is it for a customer to understand and to use and this data tiering and data visibility and all of these constructs around a data lake, which is what most customers, how most customers treat S3 are just central to how we built. All right, so the other couple of announcements around really leveraging the data, like the Kinesis fire hose, can you talk about how your storage group interacts with kind of the whole portfolio of Amazon products? Well, I think one of the biggest benefits of just moving to AWS is that you get the benefits of the entire platform. So if you think about Amazon S3 and you think about Glacier, we integrate with CloudTrail audit logging. So you have the benefits of a central place CloudTrail to look at all of our audit logs. We have CloudWatch metrics for size of storage. This is kind of in our DNA because we think this is what customers want. They don't want silos of functionality. They want this integrated solution where they can get the best of security and they can get the best of auditability and they can get the best of monitoring. So what fire hose really did, again, listening to what customers said, which is Kinesis Streams, super powerful. You can build amazing applications. We have a lot of customers like AdRoll that have on Kinesis Streams. But I want to build essentially a fire hose of small objects, any object of the size of one megabyte to come into Redshift, S3, or both Redshift and S3 at the same time. And I want that service to take care of the batch, compression, and encryption for me. So all I do is I set it up and the data just flows. So you're going to see a lot more of this thinking out of storage, which is how can we make it really easy for customers to get the benefits of a solution that take bits and pieces like fire hose, providing the data ingestion for S3 and Redshift and just make it seamless for the customer to just interact with that storage. Yeah, when you look at the analytics solutions, the new IoT announcement today, storage is an underpinning that is just critical to be able to make these solutions not only work, but affordable. Yeah, we're very excited about this. It's, you know, it makes us very proud on storage services when we know that when people talk about IoT and those hundreds of thousands of sensors, it's just all coming back into S3. And that's what the value of our storage services are is that we have this low cost, high durability, high availability. Customers don't even think about it. They just create what, you know, essentially our data lakes that are there today and then they act on it in all the different ways that they want after. Yeah, so last question I have for you. When we look at kind of the storage industry, you know, outside of cloud in general, one of the things we kind of came out of this week saying is not only is Amazon probably the largest growing storage company in the industry right now, but, you know, it's been a long time since there's been a company that said, oh, they're doing over a billion dollars in storage and, you know, our estimates are that you're doing more than that in storage. So any comment on kind of, you know, the role of storage in the IT ecosystem in general and where you play? Yeah, I think the role of storage plays so closely with the idea of innovation. If storage is expensive, if storage is hard to use, then IT departments and developers have to think about it. And we don't want them to have to think about it. We think the reason why we're here providing a storage platform is to make it so cheap and so easy to use that the developers can focus on the innovation that they want to do to go create these amazing things that you've been hearing about all for the last couple of days that customers are doing to transform their different industries of healthcare, financial services, you know, startups doing amazing things and innovation and enterprises just taking off right now. So storage for us, it is absolutely part of the foundation for how we make that innovation happen. All right, well my lawn, thank you so much for joining us. Thanks for showing off the snowball and I look forward to catching up with you more in the future. Thanks for watching, we'll be right back here with more coverage at AWS ReInvent 2015. This is theCUBE.