Speaker 00:00:00

I have about 40 IOT devices in my home that generates some kind of data.

How many IOT devices does your organization have?

And do you know how many of them create data?

That's important to you?

Where is that data being stored.

And is it being backed up in any way?

How do you decide which devices to back up?

How would you back them up?

Even if you wanted to.

Uh, this episode's probably going to make you a bit uncomfortable, but I promise you

we'll give you the best answers we can.

To these really important questions.

Hi, I'm Debbie Curtis precedent, AKA Mr.

Backup.

And for 30 years, I've had a single passion for helping

others protect their data from disasters and lately cyber attacks.

Each episode, my co-hosts and I dive deep into one specific area of data protection.

This week's episode will be the internet of things.

Welcome to the backup wrap-up.

Welcome to the show.

I'm your host, w Curtis Preston, a k a, Mr.

Backup.

And I have with me the guy who makes me sweat way

too much.

Prasanna Malaiyandi how's it going?

Prasanna?

I am good Curtis.

Well, remember with sweat you also need to make sure you say, hydrate, which I

think you are currently doing right now.

Right.

Although, I don't know if people will consider that

does not consider coffee

hydration.

I.

it's a liquid of some sort.

it is a liquid, but it's a diuretic, which

I think it means that more water goes out than comes in

So It's not actually helpful.

want.

Yeah.

but yeah, making me get up at God awful hours of the day to go walking with you,

even though I don't live in the same town.

What's that about?

It's good though.

It's good though, right?

Getting out, getting some fresh air, running into spiderwebs,

you're running into spiderwebs is correct.

All right.

it's time to talk about the news the day, backup wise, and I found

two stories, the first of which I'm so excited to talk about.

of course you are.

Of course I am.

And that is that i b m is releasing 150 terabyte, tape.

So the TSS 1170, which is the.

It's the 35 92 form cartridge.

They've gotten a lot of mileage out of that cartridge, the TSS 1170 has

50 terabyte of native capacity and 150 terabytes of capacity with compression.

Now that's interesting.

That number, again, I've been around tape a long time.

Typically that number was like two x, right?

if you look in the L T O world, typically the, you have the native number and

then you have the compression number.

And typically the compression number was two x i B M has a

different compression algorithm.

They've typically done two and a half x.

They're now apparently saying that their compression is so

good, they're getting three x.

I don't know.

and the thing I think we should mention for some of our listeners,

right.

I'm sure a lot of our listeners have heard of L T O, if you have

heard about tape, but this is just a different type of format.

I'd say it, it looks similar in size to an L t O, but it's physically looks you,

you would definitely, like at a glance.

Know the difference

between an L t O and a 35 92, but they, can run in the same tape library.

In fact, there are mixed media libraries that have both l t

O and 35 92 media in there.

I'm sure our friends over at Spectra Logic would be happy to tell you

about that.

I'd say it it's an enterprise.

Class tape drive, that probably beats l t O in a number of numbers, but it

probably also beats l t O in cost.

And by beating it in cost, I mean it costs more

Oh,

one concern is that i b m is the sole supplier, whereas with L T O

you have a number

an open format, right?

it, right?

I, I don't know.

The question is, do you like the, do you like what that tape drive has to offer?

And are you willing to live with i b M as your sole supplier?

It's not like they're a company that is liable to go outta

business anytime soon, but,

Yeah.

the other question I was gonna ask you, Curtis, so you said this

has 150 terabytes of compressed.

Capacity.

And that's quite a lot.

And I know in the article, they mentioned that yes, it is useful

for like archiving use cases for enterprises, other things like that.

If we look at the l t o, like the latest and greatest generation,

how much capacity does that hold?

Do

The LT O nine capacity for comparison holds 45 terabytes of compressed capacity.

So this is

Wow,

times the size, more than three times the size of the most recent L t O cartridge.

yeah,

That's huge.

That.

Yeah, literally

it's huge.

Both in the size, you know, and both in the, you know, you know what I'm trying to

say, in the figurative sense of that word.

Yeah.

So it is a very big tape.

The concern I always have with tape is whether or not we can make it

happy speed-wise, Tape has a native speed of 400 megabytes per second.

That is 1.2 gigabytes per second, with compression that, that's that

wonderful three x compression.

Compare it again to the l t o nine spec with it's, it goes up to 400

megabytes per second with compression.

That is a huge amount of data to feed into the tape.

And if you don't feed it that speed, it will not be happy.

It will shoe shine, it will have media errors, it will have all these

problems.

That's what I spent my whole career, you know, working with, was try to,

how to make these tape drives happy.

the reason why it's bigger.

Essentially the same form factor is that they put the bits closer

together on tape by putting the bits closer together on tape.

And by having a linear format, meaning that the, that the, the bits are

recorded in a line on the tape, as opposed to helical, which, goes back

in the time that makes the tape faster.

So my point is, yes, it's very, very big, but it's also very,

very fast, and you really need to build that into the Design.

you need to make sure you have a pipe fast enough

You've gotta have a big ass pipe.

You gotta have stuff that I do not think this is something

designed for backup and recovery.

it's for archiving, right?

Like they said.

Yeah.

Archiving and

mass storage.

Mass storage.

So anyway, but fun to talk about.

I'd love to

have one to

play with.

Feel free to send me one

Well, I was just gonna say tape is not dead.

Although

most people would like you to think that, right?

Exactly.

There's more tape sold today than ever before.

People just don't understand that.

so this next story from TechTarget talks about ai, right?

And I know AI is everywhere, right?

You can't go a single day without hearing something about AI this, AI that, right?

Open AI and everything else, and chat, G P T.

So AI has now started making its way into data protection and backup software.

So Cohesity recently added AI to its support portal for backups.

And it's interesting because I've always thought, what are you gonna use AI for?

Especially in backups, right?

It's Hey, I have this job to do, but I could totally see it being used to help

you because failures happen, right?

And you need help try to figure out what went wrong and.

As a backup company, you have all this information.

Why, as a user, do I have to go like browsing support pages

or filing a help desk ticket?

Why can't you just quickly tell me how to solve my problem based on all

the information you already have?

I haven't tried this particular, tool, but, I, what it sounds like is they're

building it into their support tool.

And so perhaps you could have a.

Resolution offered to you for your problems.

Like I have these following errors, I have this particular problem.

And you could have, if you've got a data lake billed by that, you

know, with all of the resolutions of the past, they could give you a.

Potential resolutions to your problem.

That certainly seems interesting.

my, my own personal take here, I love that they're trying to

use, make use of this technology.

I don't know about you, but when I go to a website and as soon as

I

was just gonna say, yeah.

talking to a chat bot, I'm like, person, person.

Customer representative

Customer representative.

I don't wanna talk to Angie, the chat bot,

Yep.

I'm old.

maybe the youngins, maybe they'll love this idea of talking.

They don't wanna talk to a real person.

I do know that

I prefer chat support over talking to a person.

Yep.

I like the asynchronous nature, of it.

I just, I my only complaint there is make sure your chat bott has a

little ding when you get back to me,

right?

Because I'm gonna move on.

Asynchronous nature means that they're doing like three or

four people at the same time.

guess what?

So am I, I'm doing three or four things

at the same time, so just let me know when you talk to me.

Do a.

yep.

Otherwise you have a tab and you're like, where did that, why am I

not hearing back from the person?

So that I don't suddenly go, oh crap, what happened to that support?

And then you go back and you see messages.

Are you still there?

Are you still there?

I will be closing this ticket due to your non-response.

Dang it.

Yeah.

and I could see like this use case, because like you said, there are

so many other industries, companies who use it for support, right?

It is a very common use case to have a chat bot.

So it's not surprising to see.

Cohesity taking this approach and just focusing it on their support portal.

'cause it also helps 'em with their costs.

And it helps 'em with the customers.

Quickly finding the answers that they need.

Yeah.

For the customers that want it, it'll be . It'll be

interesting to see if customers enjoy it.

Well, that's the news of the day.

We need a thing.

We need a little chime, bong, binging bong.

So in our continued backup to basic series this week, we're gonna talk

about the Internet of things or i o t.

do you, how many i o OT devices do you think you have in your house?

14.

But that's a very specific number.

Do you know how many I have?

Because I know, because I've been working on my wifi

127.

no, it's like 37, but it's because of

all of the plugs.

It's because of all the smart plugs that I went around with,

many of which I'm now dcom.

But by the way, if anybody's interested in buying a bunch of smart.

Reach out to me.

When we say what is what,

What is that iot device?

Yeah.

when we talk about IOT devices, right?

It meets a whole bunch of different things.

Most people think of it like you said, right?

It's your streaming devices like your Apple

TV or your ROKU Box.

It's

your TVs if they're smart, connected, right?

Your smart TVs, it's your smart switches or smart plugs that

you use for turning things on.

It's your ring device, it's your security cameras, it's your Alexa

slash Google devices, right?

All of these things are what people normally think about as iot, which

totally makes sense, but there's a whole bunch of things that people

don't always think about, especially in a corporate environment, like your

sensors for your doors, you have various things measuring, say, humidity or

things that are used for your factories.

That are all internet of thing devices, right?

There are devices that are getting a task done, but people just don't

consider those to be internet of things.

And all of these things are gathering some data or doing something, generating

some data and sending it off somewhere.

And usually these are sending it off to the cloud, right?

Some backend service, right?

Or to some server sitting locally on premises.

But It's doing something.

It's creating things and it's sending it somewhere, right?

Actually, raise your device right now.

cause that is an

iot.

a smart coffee mug, that keeps the temperature of my coffee at

the perfect temperature.

so I would say that IOT devices, any device with a network connection.

How about that?

I and typically we would exclude from that definition, normal computers.

Like we're not generally talking about servers and computers

and laptops and cell phones.

We're talking about, like you said, like ring cameras and smart

streetlights and my, my coffee mug.

Right.

how's that definition?

I think that works.

Yeah.

But in our world, I'm only concerned with one category of these devices,

and that is devices that create data and send that data elsewhere.

Right?

Which is probably most of them.

Yeah.

Well, your coffee mug, may not be

that type of device.

yeah, it's not right.

I'm thinking about like, well, I, I think they all send data.

It's just, it, would that send data that you might care about

Yeah,

exactly.

Yeah.

my smart TV sends data, but it's just data about what shows I watched

or

which you don't care.

And my smart plugs send data about my electrical usage,

Yeah.

which is interesting.

But if it all went away, I really wouldn't care.

Yeah,

and I don't think my coffee mug is storing the historical use of what I drank.

that might be scary.

Let's think about a variety of devices that do this.

So things like, obviously your ring camera, your, um, what, what

other, you know, uh, and, and, and various other security cameras.

This is a big iot

category,

Use case.

Yep.

what other kind of smart devices that we can think of.

there's like things that detect humidity right in a factory

or other things like that.

Things that might be detecting sensors.

Like even if you think about like cars, right?

Smart cars, right?

Those are IOTs, right?

So

a vehicle generates a lot of data.

That data.

Could be useful to you, right?

As you wanna know your driving habits and things like that.

But it's definitely important to the manufacturer as they start to build.

I know we talked about AI earlier, as they start to build models and

other things like that, right?

You need that data in order to be able to do that, so that becomes

important for the manufacturer.

exactly, and again, Like I'm thinking about like I have

a smart thermostat, right?

My smart thermostat sends data, but again, it's just data

that allows it to do its job.

It's not data that I think of in terms of data that I'm, that

see, I'm the opposite of you.

really?

I actually like my thermostats data because I actually go back

and see how often did it run?

when did it run right?

when I want to look and see, okay, how many hours and how does it compare

to like the average that they see?

Am I using more heat or more air conditioning than other people?

Those sort of stats I get information about.

All right, so that means that you value that data.

The, and this is really, I think the key here is that people

and businesses are different.

They have different preferences about the data that they want to keep on one end.

I really want to keep the, The footage from my ring camera, it's been

invaluable multiple times in the last.

I don't know, a few months where it's like Amazon said they delivered a

package and we can pull up the ring camera and see that they did not right.

If the, if that data's not there, I'm not able to do that job.

The, uh, and of course if, if something more nefarious, I just, as we're

recording this episode, they just caught this murderer after a two week, uh,

he, he was escaped, convicted murderer.

Stabbed his girlfriend to death.

He escaped and he was on the lamb for over two weeks.

They just caught him.

And the way they caught him was ring camera footage.

He would walk by people's ring cameras that would get reported.

actually he broke into a house that had an alarm, that's

another iot device that set off.

A bunch of alarms and, while they didn't find him at the house, that allowed

them to circle their, uh, to, I'm sorry, to concentrate their, their efforts.

Right.

They ended up finding him with a heat seeking drone.

that's another one we did talk about.

Yeah, drones.

yeah.

there are so many types of these devices.

I can't possibly fathom 'em all, but I want to focus on, I'm gonna

see if I can categorize these.

You tell me how I do here.

Okay.

Police body cams versus my ring cam.

Okay.

Okay.

And, and I'm gonna make a third category just on-prem security cameras that that

may have a, an internet connection, but they're not uploading, they're storing it

only locally.

So there's three categories there.

and here's my definition of these categories.

So ring that thing doesn't store anything.

it

The device,

What's that?

The

device.

the device does not store any data.

I have specified.

I only want it to with a ring.

you can actually zone in.

So I've got it.

My ring camera sees.

My neighbor's house and the street, but I don't care about my

neighbor's house in the street.

I've got the little motion detector thing set.

I only wanna see people to come in my yard.

I only want it to kick off and record people to come to my yard.

And obviously people that ring my doorbell.

Then it records that data and then that data is, is, uh, sent

up to the cloud.

The point is that the data is continually stored in the cloud and a cloud service.

That's one discussion.

The second is the police body cams.

it's a locally recording device that's recording, the, what's

going on with that police officer.

And then at the end of each shift, they, uh, have a device that they

Pop that body cam in and then that transfers that data to something magic.

And I'm sure every police station does something.

And by the way, the only way I actually figured this out was by

watching t modern Police shows.

They show them at the end of each shift was popping the camera.

And uh, and I realized, I was like, oh, I never thought about that.

of course that's the way it is, right?

because you want all of that footage.

Right?

Um, and then the third is, This idea of a device that is recording

data but doesn't have the capacity to get, at least there's no process

created to get that data uploaded.

How did I do?

Did I leave?

What do you think?

And by the way, it doesn't have to be video data, it's just any type of

data,

right?

And these are the categories of devices that we actually care about because this

is generating some data that we care

Yeah.

So the categories are, the first two are basically ca devices

that they're all devices, only

devices that create data that we care about is

Yeah.

Okay.

Yeah, that's what I, that's

So the first two are devices that there's a, there is already a process

defined to get the data somewhere.

And then the third one is the data's only locally stored and held.

and then

there is a process

Yeah, no, I think those are the three options.

okay.

All right.

So what do we think about that first one from a data protection standpoint?

My ring cam, let's not say ring cam, but let's say it's a more important device.

Yeah.

That is creating data that is now being stored only in a cloud service.

What do we think about that?

So this, I go back to all the discussions we've had about SaaS,

right?

Which is, yeah, the data's up there.

Hopefully they, the . Company, the vendor who owns the camera, right?

Or who created the camera, is providing the facilities to make

sure, okay, are things protected?

Are they available?

Are they being backed up?

All the rest of that, right?

But the question is for these, right?

Usually it's such a low cost thing that . You're not always sure what they're doing.

And there are so many vendors, like typically, right?

They might charge you like, I think Ring charges $3 a month per camera

if you want the capabilities to record and go back 30 days, right?

But there are some companies, right, that charge less.

There's some companies that offer it for free.

And so the question always becomes, if you offer it for free,

what corners are you cutting?

and can you trust them to actually protect your data?

Now, I think a lot of these vendors give you the option to download

the recordings if you want.

But who actually does that?

Nobody.

Literally

no one.

Not even me.

Exactly.

So I think that's where it's good that it's out there.

So it's not a single point of failure on premises.

So you have the data up there and it's available everywhere, right?

You could pull up your footage from your phone, from your

laptop, from wherever else.

But there are some downsides.

Yeah, so I, I'd say my general advice there, assuming again, we're

using Ring Cam, but only to help you understand the device, there

are probably hundreds of devices

that create really important corporate data.

For an organization that is then copied up to only a SaaS service.

And my answer to that from a data protection standpoint will be the same

as every SaaS service that we talk about.

so the question is just make sure you're having a conversation with that vendor and

see what the data protection options are.

Don't just assume that because it's a SaaS service, they're protecting the data.

'cause they probably aren't.

Yep.

Okay, the second one, it's a little bit harder, I think, to advise because you've

got

a system on, you're getting the data.

if it's like what I described, and again, we're using these just as an

examples, but if it's like what I described with the body cams where

at the end they pop the thing out of their vests and they pop it into this,

there's like a wall of these things.

They just pop it in there and it does two things.

It charges the, the camera

for the next shift, and it also downloads the video onto their video system.

I think in that case it's probably being stored in a locally accessed

system, which kind of makes it like the second or like the third device.

So what, we'll get to that in a minute, but I'm, what I want to focus

on here is that you have this device that's creating data out in the field.

That isn't that unlike the first device, that data isn't

automatically being transferred up.

I think, again, an example that I can think of, the first device is I knew

an engineering company that had an iPad based system that they went, when

they went around, they took pictures of what was going on in the field.

Those pictures were automatically uploaded

to their cloud system that they write.

So that's the kind of thing

it

was instantaneous.

Yeah.

Instantaneous.

But the type of device that I'm talking about here is you have a device that

creates the data out there in the field and it has no internet connection, and

then you have a way to suck that data in.

just make sure that whatever that is, whatever that system is,

because it's going to be critical, that system is being protected,

here's the question.

How many police stations, since we're talking about body cams, how

many police stations do you think actually back up that body cam footage?

To a reasonable degree,

I got nothing.

I

You have no data one way or another, but.

yeah, I don't know it, I think it depends on, Some of the police

stations were forced to do these body cams and others are like really

thinking of them as a really good

tool and others are like hating them, like

any IT budget, it's a prioritization thing,

Yeah.

Yeah.

But I was just thinking about like, how many times have you heard of

like body cam footage gong missing,

yeah, I can't think of another example of this type of device.

It could be like iot devices.

Sometimes you're out in the field like gathering results, like you're in a

remote location and you come back to a central base and then it uploads the data.

It could be even pictures, right?

Like your

Yeah, yeah.

The pictures.

Yeah.

Except that in your scenario, the pictures aren't automatically

uploaded with an internet connection.

Instead of buying iPads with the mobile signal, they bought wifi only iPads.

So the data all uploads.

When you get

back, just make sure that wherever that data's going gets protected.

It's or another example is like you see a lot of like

professional photographers, right?

They shoot with digital s SLRs, right?

And typically they don't have wifi connection, so

it's stored on a card, right?

They take the card back in the night to the camp, right?

Then they upload it potentially, or we'll get to the third case, right?

Which is probably more likely what they end up doing.

Yeah.

All right, so the third case, and what I really am thinking about

here is I really am thinking about digital security footage.

Of which there is so much now, right?

It used to be not everybody could afford, right.

Um, you know, uh, security cameras, right?

But now suddenly everybody's got HD security cameras and they have, it

doesn't cost a ton of money to create.

Um, you know, basically there's companies that for a couple thousand dollars you

get a, a, you know, a purpose-built system that has a, has a computer with, with a

disc drive on it, and then it has you.

And, and it supports, like one, I was just looking at, um, the, the

church where my wife and I met.

They, they have a school there and they had an I, uh, they had a.

This digital camera system that supported up to, I think it was 16 cameras,

and you just buy the, the p o e cameras,

right?

that's power over ethernet and, piece of cake, you're off and running and it, and

it's sort of self-contained.

The concern that I have here, and this is what I wanted to talk about, So having

worked with a number of these systems, I.

Murphy's Law applies the chance that you have, the footage that you're

looking for is inversely proportional to how much you want that footage.

I can remember many times where vandalism happened, theft happened.

Smash and grabs happen like full on like hardcore vandalization

of the inside of a building.

And 90% of the time, we'd go, we're

It

oh, the camera was, wasn't working that week.

there, the reason why, without going into detail, the reason

why, was looking at the video.

System was, there was an incident that happened at the school and we

went to pull up the camera footage and we found out that of the 16 cameras,

the one camera that wasn't working,

was the

Yeah, that

you wanted.

yeah.

the concern that I have is, If this is a device that's creating important data, and

video is an example of, it's, it can be, it's data that no one cares about until

Until they care about it.

until they really care about it.

it's interesting.

It has the opposite value of most data.

Most data is really important while I'm working on it.

While it's being created.

And then the moment I'm done with this document, I couldn't care less, right?

I'm gonna put it in Google Docs.

I'm never gonna look at it

again.

This is data where we're making so much data and it has no value.

And then three weeks from now we're like, Hey, do anybody got any footage of

that, that room when this thing happened?

And you're like, let me see.

And the answer's no.

What do we think about this?

this?

is a real problematic,

Yeah.

And I think, like you said, you should back it up.

The challenge becomes whether it's obvious when you are using the systems that, is

there facilities to tell you how to back things up, what you need to do, or is it

just yeah, everything's self-contained.

Nothing I need to worry about until things break.

Yeah,

I think it's, I think it's like the SaaS category.

I just want you to have this conversation upfront.

Yeah.

Is there any availability built into the system?

Is there any ability to have more than one copy of the data?

it can be problematic to have more than one copy offsite because video,

it's, they have 16 HD video cameras.

Do you know how much data that is?

And if they had a, an, and I, an internet connection, and then that data was

uploaded to a cloud server somewhere.

They'd hit their cap, their data cap within a day.

Yeah, exactly.

So it can be real problematic and DDU doesn't help.

Why?

Why does DDU

not help Prasann?

Dupe does not help with video.

Why

Of how video works, because typically how video works, right?

You have one frame and then it's a changes to the next frame.

And so there really isn't a lot of duplication between frames.

D ddu?

I would say DDU doesn't work 'cause there's no dup,

right?

There's no DDU looks

like finding.

Yeah.

There is compression and there are compression algorithms that

help with security camera footage specifically, that is able to detect

the similar things between frames.

I think they call 'em Codex.

There.

It can be really problematic.

So the best I can say for you for many of these is to see if

you can create an on-prem backup.

This is, I'm gonna violate my usual rules, right?

Basically because the alternative is near impossible.

It's not possible.

Yeah.

so see if you can create an on-prem backup.

Many of these servers, I'm gonna say all of these servers are either based

on Windows or they're based on Linux, and hopefully you will be able to like

install some kind of something on that box to then copy the data to another box.

The facility, and then it just becomes in the event of something really bad,

like a fire or a flood, then, you grab that box and get out of there.

No one's gonna remember to do that, but it's gonna, in the midst

of saving their life, no one's gonna go, Hey, make sure you grab the video backup.

But that's the best I

can do.

at least something.

Yeah.

yeah.

I think that's the best possible thing, because otherwise it's better

than not having anything at all,

Yes.

Never let the perfect be the enemy of the good.

Good, better, best.

Do what you can until you discover something better.

If there is a way the the thing like with ring, for example.

Ring has figured out how to minimize the footage, how to minimize the

footage that matters, you could do that somewhat with security systems.

You can turn on,

motion detection.

yeah, motion detection.

Thank you.

You could turn on motion detection to minimize the number of hours of video you

are creating, and maybe the system can use a codec to reduce the size of the file

Yeah.

and, Then maybe it's possible you can put some kind of backup system on there, like

a, something that I would think of as a backup system to get that data offsite.

That's a lot of maybes in there,

but just realize that what you're looking for is one of the most

challenging things in it right now.

A really important data intensive.

Iot device that can create a ton of data that is only stored in

one place and they don't send it anywhere because it's just too big.

Yep.

, I got an example for the second one.

My meter, my electric meter.

So my electric meter stores my electric usage for the entire month.

And then once a month, they wirelessly go by and they pull the data in, right?

Oh, yours isn't always connected.

no, not like yours.

Yours

are, you have, like, you have, uh, what you call it.

Yeah,

Yeah.

But they recently, uh, the same with my water meter, right?

The water meter.

They went around and they replaced.

The water meter with a smart one that has on the top, they have

a, little round metal thing.

I don't know

what to call it.

right?

then they drive down the street and it gives them all the, all

the data from all the makers.

That's a, that's another example.

Yeah.

but yeah, so I guess the general advice here is to just think about

all of the iot devices that you have.

a good way to do that is to use your, your,

your network.

look at your network

right?

And, look at all the DHCP addresses being given out to random devices that, and then

You have no idea what they are.

in the case of mine, what's super annoying is when you have an iot device

that doesn't give its name like it, it just says it gives you its Mac address

and you're like, that's not a name.

it is a

name, but

it's not

a

not what you're looking for.

Yeah.

I've been going through a lot that lot lately,

as you know.

So, oh,

one thing I wanted to add that I don't think we covered, but is also important

is, I know we talked about video, right?

, but there are also use cases where you're streaming data.

For instance, like my thermostat, right?

It's probably sending, Hey, here's the temperature, here's what I turned on, and

sending all this data at probably, let's just assume it's say every minute, right?

Now as a vendor, right?

You may not need to care about keeping all that data, all the

raw data for every minute, right?

You might aggregate the data up and say, okay, instead of every minute data, I'm

gonna aggregate it to every 15 minutes.

That aggregated data, though, you still need to protect that data

because that's like what consumers like myself see, but you may not need to

necessarily keep the raw data forever.

And back that up.

So that's another important distinction as well, is understand how that data is

going to be used, what's coming back from the IOT devices, what you need to keep

and how long you need to keep it for.

if I could summarize what, what I'm hearing you say there is.

Look at what you're going to do with the data that you get, and then consider your

data management practices based on that.

Right?

So the, the, I, I think, uh, a correlation or a, or a, a similar device

to what you're talking about is sensor data and manufacturing facilities.

They've got

all these sensors and this is, this is kind of like the, some of those sensors

are directly driving what's happening.

Some of those sensors are used later in investigations, and so just realize what

that data is being used for and then, store it and protect it accordingly.

and I'm just saying just have this discussion.

Just, just think about all the IOT devices that you have and, sometimes

we spend a little bit too much time just talking about servers, just

talking about SaaS devices or SaaS services and, we don't spend enough

time talking about the millions of iot devices that are out there.

Yep.

All right.

And with that, thanks for having the

chat again.

Prasanna

Anytime.

Thank you, Curtis, and looking forward to our walk in the morning.

Absolutely.

I'd like to thank the listeners for listening to us.

We'd be nothing without you.

And remember, this is an independent podcast.

The opinions that you hear are ours and not necessarily an employer.

That's a wrap.