Storage Developer Conference - #45: Data Retention and Preservation: The IT Budget Killer is Tamed

Episode Date: May 25, 2017

Hello, everybody. Mark Carlson here, SNEA Technical Council Chair. Welcome to the SDC Podcast. Every week, the SDC Podcast presents important technical topics to the developer community. Each episode is hand-selected by the SNEA Technical Council from the presentations at our annual Storage Developer Conference. The link to the slides is available in the show notes at snea.org slash podcast. Welcome again. Hi, my name is Alan Sarkelli, and I'm representing the LTO program today. And we'll have a presentation today around data retention and preservation. The IT budget killer is finally tamed. So what exactly is the LTO program? So does everyone know

that this is tape at what we're talking about? Linear tape open technology and the technology provider companies that sponsor the consortium that backs this technology are Hewlett Packard Enterprise, IBM, and Quantum. And what those companies do is really set the standards and write the specification around the LTO spec. And those licenses are available to any organization that wants to manufacture. Yes, sir? What's a log-off right here? Say again?

What does LTO stand for? Linear Tape Open. Linear Tape Open. Tape. So? Linear tape open. Linear tape open. Tape. So it's digital tape technology. Okay? So it's been around about 15 years. It's an open standard technology that's on the market today.

I'll kind of go through a little bit of background on that to give you some more information. What the tape and media technology is, it's a high-speed digital tape technology. It's open standard. Those companies are the specification authors technology is it's a high-speed digital tape technology it's open standard those companies are are the specification authors as well as manufacturers of the technology so those are again you can purchase the the products from any from

any company that licenses those are just happen to be the ones that sponsor this consortium okay so again license to any organization you can buy multiple licenses there's multiple media manufacturers in the marketplace to keep the competition for media sales. And we have a roadmap that goes ahead up to 10 generations. Right now we're at Generation 7. So the goal is each generation has some capacity and performance enhancements. And the capacity targets, I'll show you the roadmap in a little bit,

but up to 120 terabytes per tape cartridge. So why is that important? So the challenge, and everyone knows this, is the data growth that you see in the marketplace. There's a lot of unstructured data out there. We're seeing huge growth rates, 42% year onon-year, up to around 2017, looking at about 125 exabytes of data. 80% of that is unstructured. So what we're seeing is a real trend of a lot of this unstructured data doesn't necessarily need to be accessed frequently.

It's in digital archives. And the problem statement is what do you do with all that data? What kind of cost structure do you need to maintain that data for long periods of time? And that's where we see this cold data type of methodology in customer types of situations. So users and operators, they really do see that huge amount of unstructured data. They're trying to prevent sprawl across the across the network and there's a lot of IO intensive workloads out there that is creating more data you can see this chart here just kind of again you know the up and to the right you can see this capacity optimized data growing performance optimized data kind of flat as you look over the years so So there is a workflow problem where

people and organizations, government, businesses are generating tons of data, most of it unstructured and have a problem of storing this for long periods of time. So where does tape come into this? One of these statements I really like here is this quote from a gentleman at Microsoft. He says, where do you think these zettabytes of data will be stored? They're going to have to store it on tape. And the reason is the cost structure of technologies, disk technology, flash, all kinds of storage technologies, is just not cost effective for long-term storage. You have the infrastructure cost.

You have the power, heating, and cooling costs, and this is where tape comes in as still a very viable alternative to having this long-term types of data, unstructured data stored. And I really like talking to people, you know, being a tape person, when they say, you know, oh, I, you know, I don't need to worry about long-term storage, I'll just throw it to the cloud. My response to that is, well, we can let them put it on tape for you. So we're seeing a lot of growth in the cloud infrastructure, having to have that lowest cost per terabyte of storage, reducing hardware, and having that reduced power consumption

and HVAC costs just immediately can buy back the, get your money back on those types of environments. And again, less media disposal is important too. just immediately can buy back the, get your money back on those types of environments, right? And again, less media disposal is important, too. As you see, you know, IT centers thrown away and having guys said they're swapping disks out all day long. The LTO technology right now is specced to last for 30 years per cartridge. So, you know, if you look at a long-term, you know, workflow and long-term storage of cold data, this is really, you is really a big opportunity.

And we're seeing a lot of opportunities in this archive market for tape as well as that kind of long-term need for cold storage capability. So again, less resource spent replacing, swapping, having to deal with infrastructure. So one of the things that's come about in the last few years is LTFH, which is Linear Tape File System. What does that do? So linear tape, there's a lot of applications that can write and read data from any medium. IOTA tape, just like any other type of application, the application-specific or proprietary I.O.

can go to that tape as it is any other kind of medium. What LTFS brings to the table is an industry standard that has basically a file system that presents that tape system as basically a mount point on a network. So you can look at, you know, if you were really into a situation where I need to have, I want to be application free, I want to have data on a standard format that I could possibly read in 10, 20 years that's not going to go away, that's available from every vendor,

this is where LTFS comes in. And we're seeing this, you know, and it becomes as easy from an application standpoint as a mount point on your network. So you can simply drag and drop as you would enter files to that tape system. And then have the format being then an industry standard that can be read from any vendor. So you're not attaching yourself to any vendor with the proprietary file system. You're able to access that for long periods of time. A lot of the applications, or very simple applications, to read-write LTFS are free from vendors. You can simply download them and be able to read-write to tape

as it is a mount point on your network. So this is really kind of a game-changer in that the whole long-term and proprietary problem is just one way to handle that, and you can get away from having to have linked to one vendor in these long-term storage scenarios. So again, you get your cost savings. You have your ability to have your cold data stored

for long periods of time on a medium that's very reliable. I'll get to more on that too. So again, the LTFS changed the world we live in. What we're seeing is these application sides. I'm going to run through this here. And data exchanges, being able to have some kind of, I mentioned, the simplest drag and drop, and having a simple archive.

And there's a lot of applications now that can manage that archive. So if you have archive. And there's a lot of applications now that can kind of manage that archive. So if you have many tapes, there's tape library products out there with many tapes in them, you can basically create a cold data storage archive with this tape subsystem and not having to have a complicated interface or proprietary I.O. system to get to it. So there's many applications out there that can write data to various storage

systems. This is going to be another one of those mount points for your final tier in an HSM environment where you have a hierarchy of data going from primary, and you want to get that application then to move it to some other target. This is about every application can write to a mount point on the network. So it becomes a very simple way to integrate tape into that workflow. And you get the benefits of the cold storage then. Yes, sir? I know you're going to get to the roadmap, but is this kind of like the future of LTO?

So LTFS, it's been around for a few years, the format. It's going to be going forward, yes. Yeah, I have LTO4, LTO5, LTO6 format. Do I have to migrate these? So there's always going to be migration as part of a strategy, right? No one has floppy disks anymore. At some point they migrated them. The LTFS does make that simpler.

If you do want to copy volumes from multiple LTO5s onto, let's say, an LTO7, you have the ability to do that without having to necessarily change your file system format. Now, if you have an application that wrote to this previously, you might want to extract that and then potentially put it back to LTFS, and that just depends on what kind of workflow you're looking at as a user of the technology. So it was a good question, like as you mentioned, I have LTO4, I'm going to 5, going to 6. It still can read back two generations as well. So you get at least a few years to figure out your strategy around that.

Always with that kind of equipment weakness issue for long-term formats and that migration. Yeah, I mean, we've seen people do stuff like, we'll take, let's say, the 4s or the 3s, whatever, possibly have to extract it and then bring it back as an LTFS, put it on data, and then throw those tapes in a closet for however many years they want to, or into a vault, more likely. I closet figuratively. But yeah, so this is one way you can migrate that.

And the LTFS format is going to be going forward with every generation. So it came about with the Generation 5, and it's planned to go on for the rest of the generations of the LTO roadmap. Okay. Okay. Okay. Okay. Yeah, so one of the things, too, is, you know, that's my place there. Yeah, so one of the advantages of tape is, again,

we kind of talked about this indirectly with the question, you know, we have this ILM or information lifecycle management type of, you know, problem where you have, you you have clients and different users and applications having the hierarchy going from a flash, the dollar signs indicating the expensive to the primary, to traditional disks, with some kind of period of access ability, which is basically the definition of an HSM, or a hierarchy management system. So basically, you want to have your expensive stuff period of access-ability, which is basically the definition of an HSM, or Hierarchy Management

System. So basically, you want to have your expensive stuff for the stuff you need quick and accessible very quickly. You then have a tier of possibly traditional disks or some kind of hybrid where it has some less hot type of data on there. So typically, we kind of put a temperature on this on this hierarchy, right? From the hot data, which you need quickly and fast, to the warm data, to, you know, some kind of what they call sure-nothing clusters where there's possibly like an archive environment. And then when that data gets to that cold tier, it really becomes, here's where your advantage of cost

comes in. You can throw it on tape and then once that cold tier is, really becomes, here's where your advantage of cost comes in. You can throw it on tape, and then once that cold tier is, you can vault or maintain that in a library with very little cost overhead on doing that. So it really becomes, the tape becomes the reliable and the cheap way to get from your expensive flash onto an archive environment. We're seeing more and more of that. As this becomes more of a backup type of application, I like to think of archive as data that you don't need to backup anymore. So it cleans out the clutter across all your primary storage and saves you money. So this is just some kind of density trends. So what this is showing is how much growth across this is in gigabits per

square inch on the y-axis there. So you're looking at kind of growth of various mediums in the industry from DVD to hard disk drive. You can see getting more data onto that media. And you can see tape continues to grow here, which is the key point of this slide.

I have all kinds of dots. But there's a lot of room for more data per tape cartridge per medium. Improvements in the servo mechanism, the media itself, and other aspects of the drive design make that growth possible. And it's going to continue to go, as I'll show you in the next slide here.

And this is where I wanted to show it. So if you look at, without giving you a headache on these little dots here, but trying to look at the scale bits across a flash technology, this number right here says 2,150 gigabytes per square inch. So they're packing a lot of data onto that very densely onto that medium. Tape is a relatively smaller four gigabytes per square inch. So 200 times

whatever the math is, 500 times less. So what I'm trying to show here is that there's a lot of room to pack more density, more bits onto that same medium and that same footprint. So with simple improvements on, not so simple always, but improvements on the drive technology and the media materials and such, you can get a lot more density. There's a lot of room to grow to fit those bits into that environment. So I want to show that the reason I'm going through all this is you hear a lot of times, well, tape's an older technology.

Tape is dead. What I want to show you here is there is still a lot of room to grow, and we're seeing with those cost benefits and ability to pack more data onto the same type of media footprint, there's a lot of room to grow, and dead isn't the right word. And as far as reliability, there's no better technology. This is an error bit rate chart, and what this is is showing hard error rates in bits. So for every write cycle, they can basically statistically count how many times you'll see an error in a bit read or write operation.

And this hard error bit, you can see various technologies on here, SATA Consumer, Enterprise, Enterprise SSDs, and you can see the bit rates and their format. Enterprise Tape and LTO, which is what I'm talking about, I'm an LTO guy, you can see it's a magnitude difference between that and the other technologies. So between a flash and an LTO tape, 10 to 16 versus 10 to 18, so you're more likely to get struck by lightning twice than to see an error bitrate on a tape cartridge on a write cycle.

So what that means is the reliability of this is rock solid. We don't see a lot of data problems. The LTO is a very reliable technology. Like I mentioned, it's been around for 15 years. We have a long roadmap ahead of us. And it's going to continue to improve. Like I said, as they pack more density on there, there's a lot of room to be able to not only improve

the reliability to make it even better, but there's room for more data per medium as well. You can see the bit error rate continues to go down. And the capacity is going up. So that's what I want to show there. Efficacious storage, that's a great word. So in a 10-year total cost of ownership study, we looked at this media cost and how they were going up

and how just operating costs continue to increase. LTO media costs are going down. So as the technology makes improvements, you get a more mature technology, and you're seeing costs go down. And If you look at some of these numbers, incremental user benefits, I mean some of this I've described. You get the benefits of HVAC costs, the whole idea of cold storage, not having to maintain expensive medium like flash for long-term storage. It really makes a lot of sense to compare your media cost

over a long term that you're going to be storing those tapes compared with what kind of primary storage you might have to invest into to have that data infrastructure carry your primary storage. So basically there's the advantage of cleaning out the clutter on your primary, using that for what it's designed for, which is your hot data, and using tape for the back- designed for, which is your hot data, and using tape for the back end or the cold data archive.

So, talk a little bit about the tape market. So, you know, I said, I hear this once in a while, yeah, I don't want to use tape instead. We're seeing growth in capacity shipments on the order of year-on-year 17%. We can seriously see an uptick. You can see this. This is a compressed capacity shipments on the order of year-on-year 17%. We can seriously see an uptick. You can see this is a compressed capacity shipments by quarter of tape. You can see it's up and to the right going into the third quarter of last year. The LTO7 introduction is actually driving more media LTO6 sales, and the reason is LTO7 can read right to LTO6,

and the cost of those, you know how it is, when a new technology comes out, the older technology, the cost goes down. So we're seeing a very cheap way to invest in new technology with the latest generation of drives from LTO, at the same time reaping the benefits of the last generation's cost decreases over time. And we expect that to continue. When LTO 8 comes out, LTO 7's costs are going to drop, LTO 9, et cetera. So just like any other technology,

the next generation drives the cost benefits of the previous generation. And these being backward compatible, users can get the advantages of having that cost savings into their new technology without having to worry about a backwards compatibility issue. And tape media, again, shipments running relatively flat.

These are just basically unit shipments. So we're seeing customers are still buying and users are still buying a lot of tape out there today. Capacity going up, but as the tapes themselves get bigger, you can kind of see it flattens a little bit on units. So a little bit of arithmetic there. So here's the roadmap. Like I mentioned, this is, you know, Generation 3. We're going back quite a few years,

and I just want to kind of show the growth rates of what the spec has brought to the table on the units. You know, 800 gigs doubled to 1.6 terabytes, to 3 terabytes generation 5, 6 and 1 quarter terabytes generation 6. Here we are today, LTS 7, 15 terabytes compressed, 6 terabytes native per tape, 300 megs a second native, and 750 megs a second throughput compressed.

So it's not a slow technology either. It can outpace many of the existing disk systems out there today. The next generation, and typically there's usually around a couple years between generational releases. So like the gentleman asked, how am I going to plan for that? That's something you can kind of look at, read back two generations and read right back one generation. So you have some of that planning available to you as a user to be able to determine how much data you need to store and what kind of roadmap you want to have for your data center.

Generation 8, I mentioned, is probably going to be into late next year, possibly, maybe a little bit after that, but that's in the works for planning. Specification is out there for vendors to look at today. And then moving into generation 9 and 10, moving out some years, you can see the target maximum capacity as we get to generation 10, 120 terabytes per tape. So, it's very high capacity. And up to 2750 megabytes a second. So that's the throughput per drive onto that tape.

And those are two defining characteristics of the generation. Is this capacity and speed? It is. There's been some features, like I said, when we were talking about the LTFS, they came out in Generation 5. So there are feature sets. Worm came out in Generation 4.

There was some partitioning on the tape itself for ability to have the LTFS. And then you can see there's possibly other features go, but you're right, though. the basic defining features are capacity. Typically you look, you know, it's about double between each generation on capacity and performance. Yes, sir? What's the timeline here? Is it Generation 8?

So Generation 7 is released. Products are available today. That was released in late last year, so typically two years between generational releases. You could estimate plus or minus months to see products into late next year for 8, following two years after that for this and two years for that. We're looking into 2021 for the last generation.

Those are maximum speeds, so read-write. Maximum compressed throughput. Typically you get it faster on a write cycle. And again, that's compressed too. So the spec calls for native maximum throughput and also then on a read write and then also compressed. The compression algorithm for instance LTO7 allows 2.5 to 1 in the hardware itself so that algorithm will be able to compress the data as it comes in. And depending on the data set, as the disclaimer always says,

it can be up to 2 and a half to one for that compression algorithm. So typically if you see a lot of small files, a lot of unstructured or simple data, it's going to compress very well. You see more complex files like video we're seeing a lot of these days, it's going to compress much less since there's already compression algorithms in that file structure, especially like video and audio and those kinds of rich media files. Okay? Get on time. Okay, so here's a nice little acronym here, PROPEL, PR.

So protected, we like to think about this when we're talking about LTO and tape. Protected, there's an offline content, creates that air gap. Basically, ability to vault data, keep it safe for a long period of time. It's reliable, we talked about the reliability. There's not a lot of bit rate errors. It exceeds every other media type on the market for error bit rate generation and build the

transfer from that archive data into from content into some kind of revenue generation in that your cost savings will exceed the primary storage cost over many years especially it gets better with time reduce environmental impact less replacing that media less waste that goes for you know the especially. It gets better with time. Reduce environmental impact, less replacement of media, less waste that goes for you know the HVAC and other IT cooling and heating considerations you have around the data center. You know a tape in a vault doesn't use any energy, it doesn't spin any media, doesn't have to use any power. So it's a very cost efficient way to store large amounts of

data. Optimize, so you can reduce time to your data up to 3X and streaming applications. Again, the performance of the tape today exceeds many of the requirements for these kinds of applications where high rewrite or fast rewrite cycles are required. Portable, again, you can share content. The LTFS, the linear tape file system, is an open format. So one piece of media that can be used in another vendor's hardware, you don't have to necessarily be stuck to any particular vendor

when you're using LTO. And we see that a lot too, especially in the rich media content where they want to have the ability to move to another system or another vendor or some kind of outsource vendor. Everyone can read LTFS. There's no restrictions on that. Expandable. So limited capital expense.

If you have a tape investment, adding media is the way you can grow that archive or that data set. You don't necessarily have to add more controllers and hardware and this subsystems to that environment to get that extended archive to grow. And low cost. I've talked about this many times. This is one of the most important ones. You can definitely reduce your IT infrastructure costs by considering tape as that tier where the cold storage can live for a long time without having that overhead and extra capital So propel with the two in the middle. Okay. I want to kind of quickly here, we're already at the Q&A, so I just want to recap.

Again, LTO program, we're here today. We're going to be at our booth out there speaking about LTO, talking about details around the roadmap. Feel free to come by and chat with us and the team. Again, I want to really kind of push it in there that the LTO is a very viable program. We want to keep everyone in mind that we have a long-term roadmap for this, the vendors and very stable suppliers in the marketplace, a lot of media vendors, additional features and application support, generic support with the LTFS and other types of investments around

the tape community. So with that, kind of quickly, I'll open up the Q&A. Yes, sir. The drop guys talked about online, gear line, and archive. Mm-hmm. I was kind of listening to you. Like archive drives. Archive drives?

That's where we're seeing the biggest growth. So if we go back to the HSM slide, if I can find it. But I think you're, yeah, I haven't heard the, I don't know, is it cost difference or are these? I think it's an, what I'm gathering is, I don't know the exact terminology, what I'm gathering is it's an application use case difference. So you might have, you might have, you know, we're seeing a lot of disk subsystems doing backup where tape had traditionally been, you know, you had your server, you had your backup, and it was a tape, right? And a lot of people still do that.

They still put single drives in the servers, you know, as a backup. But there's a lot of these disk subsystems now that are basically replacing that backup tier, and they have features like the deduplication and be able to have better compression algorithms where that backup then is basically compacted into some disk subsystem. Where the tape piece comes in, this is where I think you meant by, I call it the archive tier, maybe someone referred to it as archive drive, is you have that disk system then that can basically take that data that's not needed in that cold tier,

dump it over to that tape archive, and then you don't have to back that up off the primary anymore. So you're saving money in your disk backup costs as well as your primary storage costs as well. And again, I always say archive is data. You don't need to back it up. So it becomes that long-term tier where you need to have it. You've got to keep it for a long time by policy or regulation or what have you, but you don't necessarily want to have it on your expensive disk, and you certainly don't want to have to back it up and manage it every day.

It's also offline, so for the tap, you have that air gap. Yeah, that's a good point. Yeah, so the idea of the vault that in some safe location is a consideration, too, where you have the ability then to cheaply vault large amounts of data for a long period of time. Yeah, I think it gets down to kind of the, you touched on it, the temperature of your data. Yeah. The hot data you want to have, you've got your fingertip access, but you start getting to the point where,

yeah, I touch it once a month, I touch it three months, and then the whole data, I never want to touch it again. You got to have it, right? And that's what the good thing, like some of these LTFS applications, if it's as simple as a mount point,

and a lot of these apps are coming out now where they'll actually kind of stub the data on the primary. So the user doesn't even know it's on tape. And if it's in a library, for instance, one great use case is you're in the legal department and you say, IT manager, my disk here, don't erase anything ever. And you go, well, the guy hasn't touched it in five years. But he told me not to erase it ever. I can put that on tape,

stub it on the primary, have it on tape in the LTFS format in the library. When he does need it for whatever reason some years down the road, he can double click on it. Basically, the operating system will try to bring back the stub, the primary data.

Just pull it off tape. No one would ever know, except that it takes a little bit longer for the tape to actually load into a library. So, I mean, that's a great application where you see this. It's just been sitting there for years, and there's policies designed. Every legal department user is over two years old.

I'm throwing that to tape. So it's a pretty good use case for having the ability to then have a long-term, true archive in a cold storage environment. Yes, sir? How does the media management piece handle LTFS? How does it know? It's like, how hard is this to handle?

And then what take-through or activity make you have a big silo of tapes? Yeah, so the management... So the LTFS is the file system, right? So there's various applications that can do the management piece for you. Again, the format is going to be standardized. How you manage it, I mean, it can be as easy as, you know, with the spreadsheet,

or you can have some application that basically has a database of what that archive looks like. And some of them are clever enough to, like, for instance, I know there's some of these archive apps that will use LTFS as a format, but they'll basically write their own database of the table of contents, if you will, to each tape. So any tape can bring back and it says, oh, this is tape one. It has the whole D drive legal department,

and these five tapes are associated with that. So there's different ways that the applications will manage it. The format on the LTFS is just that. I'm trying to think of it as a comparison between the back of software and the LTFS. Yeah, so yeah, the back of software, I mean, the big difference is, you know, and the back of software does that kind of effort.

That's what those applications do for the LTFS. The difference is the proprietary, right? So the proprietary format. Yes, sir? You mentioned legal. I understand some legal requirements now for certain documents to be maintained for 75 years or more.

I assume there's a refresh cycle that you're storing on tape all the time. Yeah, absolutely. So, I mean, that's going to be part of your management policy, right? Kind of like we alluded to at the beginning, you know, no one has floppy disks anymore. At some point, they migrated those. They may have thrown those disks into a vault. They're not going to work anymore. You know, they're 30 years old.

You know, one of the benefits of the LTO medium is, you know, it is specced to last 30 years. So you can vault it for a long period of time. Typically, five to seven years, you're going to try to find some kind of cycle to basically bring that media onto the newer. So that's just rewriting. It's changing media. Exactly, yeah.

And it may be as easy as, oh, I wrote this. Let's say I wrote this in 2016 in an LTFS format. In 2020, I can basically copy that LTFS to basically a larger LTO, so that stack of media becomes one piece of medium. So it is part of your recycle strategy. But one of the advantages is it does have that long-term capability, so you're not pushed to do that. And you can still keep that medium for a long period of time without losing the data if that's an option.

Obviously, 20 years from now, who knows really what kind of hardware is going to be available, so you would want to make sure that you have some kind of strategy around keeping that for a long period of time. But you're right, those 75 years, I've seen this with EMR, the medical records. You know, video surveillance is a big one now where they want to keep, you know, police videos for many years just in case there's litigation or other legal considerations. So, I mean, there's all these applications, be they, you know, legal or policy, where it's the exact same scenario you described.

You know, many years I've got to keep this. And, you know, we feel that, you feel that the LTO is a very cost efficient way to do that. Other questions from the crowd here? How are we doing on time? We got time? We got a few minutes if you want to chat more tape? Like I mentioned, how many folks here are actually developers? There's one in a couple here. How many are IT managers or infrastructure people? Okay.

And how many people use tape occasionally or more than occasionally? Have it in your infrastructure? Okay, a few. I only recommend it when you have to. What's your alternative, typically? To me, every time there's a new technology, like SSDs or flash storage, it pushes everything down on the stack. So hard drives, nice to be hard drives, used to be,

you know, the ultimate performance. And now they're second tier, third tier type of performance. Tape gets pushed out. They work with a lot of media entertainment people,

right? And the smaller shops, because of the cost, put a lot of stuff on tape. Because they just have so much content and they can't afford the online. But, they lose a lot of stuff on tape because they just have so much content and they can't afford

the online, but they lose a lot of productivity in looking for things that they need for various projects. It's really a treasure hunt. Yeah, I have seen stuff like that, yeah. But for, like I said, for LTO, what I see it as is that really cold storage. And another person, you kind of touched on it a little bit, but another presentation was talking about data information protection as well, ransomware and so forth. Someone hacks into your network and they encrypt all your data and you have to pay the price to get it back.

But if you have it on the air gaps storage, it's like, OK, mine has a storage problem. So we see it more at that insurance level that would be cool. Like I said, if I touch it, it's variable kind of stuff. Right. And that's valuable because it's typically IP, right? Especially in M&E, like you mentioned, that's their money.

And if someone steals that, they're stealing their money. Yeah, and who knows? You're doing something today, and then 10 years from now, you're doing something related and you want to access footage from 10 years ago to do a human interest story or whatever. Right. So you need to be able to find it and access it this footage from 10 years ago to do a human interest story or whatever. Right. So you need to be able to find it and access it and have reliable, you know,

I've heard a lot of people that have those stacks of hard drives. Sure. And they cross their fingers every time you plug those things into a SIDRAM for five, six, seven years, right? Yeah. So it's kind of the same thing. The tape is long. And it's still kind of hard to find. And that's kind of what the gentleman to your left kind of mentions,

like, well, how do I manage this stuff, right? And like we talked about, the typical backup apps, I mean, they have kind of a database or contents where to do it. I mean, it's not rocket science to just say, well, I have a bunch of C drives, one from 2015, one from 2016, one from 2017, to manage that. So there's not these huge databases growing out to worry

about. So, like I said, there's these applications that will write to the LTFS format and just kind of manage that content. And that's, I mean, it's not overly complicated to do that. The media is growing dramatically because of just the file sizes and the resolution of the cameras and all that stuff. But there's a huge explosion

in the metadata side of it as well, right? It's all the descriptor data for that data is just growing more and more thematically. They're struggling with ways of being able to manage that for writing standards for the metadata piece. Right. Yeah, and that's, again, like with the application side, there's applications that specifically manage the tape piece of that as well and kind of show

that mount point. And then that operator or whatever might want to have their own metadata database for their primary that doesn't necessarily have tape as a consideration for tier. So it might be a 1-2 type of leap from the primary to the archive. And there's applications that do that. We can talk about those at the booth during the things,

but I can list off a bunch of the ones that do that. Yes, sir? I'm just curious. I know what flashes do and get to get smaller and smaller in density. What do you do to a tape? There's a medium and there's a magnetic stuff. It's really, yeah, there is improvements on the media materials. You see different kinds of, you know of metal oxide that's being used.

The head design, the number of tracks, basically compressing tracks, that's one of the ways that the growth of that LTO capacity is growing. Improvements on the head and the track count that they're laid down for tape. Like I mentioned in that one slide where it had the density, there's a lot of room

left on that metal piece of medium to lay down more tracks, lay down more, have improvements on the head. So there's really this improvement on design on the basic head and that spec that have driven the growth across the medium. You're talking about lightning? Yeah, exactly. Flash and stuff. Yeah, flash. And that was kind of the point, that one where you saw little dots.

Flash is packed on there pretty tight, and you're paying a lot of money for that. Tape's got a lot of room to grow, so that's what we're really trying to. Those companies I mentioned at the beginning are investing in trying to continue to increase the capacity, increase the reliability, et cetera, as it goes forward.

Any more questions or comments? the reliability, et cetera, as it goes forward. Okay. Any more questions or comments? I have a question for you. In the media entertainment, are you seeing anybody using tape to get information, get videos across town as opposed to, you know, if you're case setting, I don't know if it's one-offs or... Typically, you know, if they're going to do the secret F thing, it's the hard drives, it know, one thing I realized

or I learned about video media and entertainment workflows, they say they're like snowflakes. No two are exactly the same. Right. But in general, for sneaker net kind of stuff, they're using portable hard

drives, flash drives, that kind of stuff because of just the speed of getting content out of tape is, again, it's like, I'm done with this project. I hope I never touch it again. Yeah. We were at an ABA. I had a few folks come out, and I was wondering if they were just going to want us to do that.

Yeah. That's a great thing you brought up, though, Manny, because I have seen some use cases where, you know, data has weight, right? And rather than, like, a company might be migrating from one location to another, and company mergers or different stuff happens. And rather than having to dump all kinds of data over the network, they basically FedEx a bunch of tape.

And one of the engineers, I remember, he was saying, he's like, when that FedEx truck crossed town with all those tapes in it, that was basically six megs a second or something, right? So, I mean, the data does, and that's saving them money on network bandwidth. The Internet is still faster than Ethernet, right? Yeah, exactly. And you look at some of those, like, there's some of these companies,

when there's a merger or split, and they want to basically want to split them off. And we've seen this. The FedEx trucks go back and forth with the tapes, and everyone gets their data. And the network bandwidth to do that would have been ridiculous, and they did it overnight with a bunch of guys and a handful of tapes. So that's a great one. Any other questions or discussion? We've got a few more minutes for the next one.

Okay. So a lot of the work papers and TCOs in the booth, the staff might have got those turned over and rolled one. Okay. So a lot of the work papers and QTOs in the booths that I thought was turned off and you're all done. Yeah.

All right. Well, thanks, everyone. I appreciate your time and enjoy the rest of the show. Thanks for listening. If you have questions about the material presented in this podcast,

be sure and join our developers mailing list by sending an email to developers-subscribe at sneha.org. Here you can ask questions and discuss this topic further with your peers in the developer community. For additional information about the Storage Developer Conference, visit storagedeveloper.org.