00:00
Um, Jonathan Karnes, this is Peter Gonzalez. We have a guest speaker today as well. We'll introduce a little bit later, and we are here to talk to you guys about file. So I hope you guys are excited. I hope Accelerate's going awesome, and this is like kind of probably your last meeting for the day,
00:16
but it's gonna be the best meeting for the day. They will. And uh this is kind of gonna be a little unscripted because I don't like to be scripted. Anybody who's watched me, uh, present? I have to do that for my manager so he knows I was actually here.
00:33
Yeah, is that micromanaging? He does micromanagement, yeah, so I get it I get it. Um, you know, Pure has been doing a lot with file over the past few years, and I think it's time for us to give you guys a little bit of insight of what we did the last 12 months, but also I wanna kind of, we wanna show you guys how File at Pure is
00:57
different. Than file outside of here and I think it's really what we have really show people because um that's a message that's kind of hard to tell sometimes, but I think it's really important that people see it. So do you have a clicker? I don't have a clicker. Who has a clicker. Oh, here's a clicker.
01:15
I swear we have slides. Here we go. So. I want to start off. Have you ever heard about this experiment? The jumping flea experiment. OK, I'm gonna tell everybody about this. So it's, it's about Nas, right? It's about mass.
01:35
OK. Well, we, we're related to, so. There's this experiment scientists did. You know, anybody know that fleas can jump really high. They can jump like several inches. And so scientists took fleas, they put them in a jar,
01:46
they watched them jump out of the jar. Then what they did is they put a lid on that jar. They left the fleas for 3 days. They came back after 3 days, they took the lid off the jar. Guess what? The fleas would only jump.
01:58
Up to about where the jar lid was at and no higher and that's all they ever did. They just stayed that way. And so This experiment. I want to be honest with you guys, it's technically not real. It's actually based on many other experiments that are too cruel to present in a meeting like
02:18
this. So marketing people came up with this one, but it's based on true science that you can inherently be taught to live within limits. Anybody been on the highway, seen a car hit their brake lights a little far down and all of a sudden felt your foot hit the brake pedal a little bit, has felt that limit.
02:39
You know, we all get taught that way. And this has happened in file storage over the years. We've had 20 year old companies teaching us how to. Operate file. So what do we do? We have to figure out how to break that paradigm, break this paradigms.
02:56
You know, just a, a little background though on why these 20 year old companies have, I would, I would call it a legacy infrastructure for lack of a better word. And if you think about. I think about right around 2012 is when they stopped the innovation cycle around NAS, and they did it for a couple of reasons. Um, unfortunately,
03:17
I was around back then and uh. If you think about what was going on, they were doing departmental shares and home directories. That was the primary use cases for NATS back in the day. And you also got to think about the infrastructure. The infrastructure was 10 base T, 100 base tea, very, you know,
03:33
kind of slow. Everything was slow, um, and but there wasn't really a need for performance back then, right? And then along came the cloud, right? And the cloud's promise was that we're gonna take your, you know, users home directories, uh, your departmental shares, and we're gonna put them in the cloud.
03:51
So these NAS leaders back then said, well, should we continue to innovate around a market. That really isn't gonna be here in a few years, right? So they stopped. That's what they did. So in 2012, right around when Microsoft released AD, that's when we stopped the innovation cycle,
04:08
and they started going into the life management life cycle management cycle, right, where they're just gonna roll out stuff to maintain the code and that was it. So we saw that. For many, many years, right, go ahead and go to the next slide. Well, yeah, I think the, the, the thing is this isn't have one area,
04:27
right? What did, what did the next competitor that came up, they saw that previous guy and what he had, and they said all we gotta do is what he does. So you ended up with a bunch of these similar NAS systems, all kind of a light blue color, right? By, you know, that you'll look like they're from the late 90s,
04:47
early 2000s, right? And they didn't really change much. So Pure said it's time to apply the pure principles to file, and this is what we came out with. Little different, little different car, right?
05:05
But, but, so let's, let's take another step and say, OK, well what changed, right? Why is it that high performanceAS storage is such a thing today? What changed between 2012 and today? Well, the first thing that happened were innovations outside of network attached storage,
05:21
and the first one that I think is super important for people to understand was networking changed, right? So we went from the 10 100 gig, uh, as a high performance, uh, network to what do we got 1040 and 25 and 100 gigabit today and the prices came down. So everybody's got a high performance network today, right,
05:43
that we can leverage for network attached storage. The second thing that happened, right. Was innovation around uh uses right use cases so we're not looking at home directories and file shares anymore for a directory structure we're looking at unstructured databases we're looking at Vertica, we're looking at AIML workloads all these new workloads started
06:04
coming out and they, they, they were developed in the cloud. That was a cloud first use case, right? So what happened? Why didn't these applications just stay in the cloud? right? Well, a couple of reasons. One, performance.
06:18
You can get performance out of the cloud, but it's gonna cost you some money, right? And that's, you know, the promise of the cloud was supposed to be cheap, right? So the first thing you needed was performance, especially as the data sets started getting bigger and bigger and bigger people wanted more performance, they wanted higher levels of
06:33
availability, they wanted to do more with their data sets, they wanted better security. And they had to pay higher prices to do it, or they could bring it back on-prem. Well, you know, the, the development around on-prem had to improve as well. That's when the networking started becoming more of a factor.
06:51
That's when high performance storage became more of a factor, but it had to be delivered at a price point that customers found satisfactory because we're competing against the cloud. Make no mistake about it, we are definitely competing against the cloud. So our price point, right, has to be better.
07:06
Are, uh, you know, the SOAs have to be better. The, uh, you know, the, uh, management has to be easier. All these things have to come together before companies can bring it back on prem. And the reason you're all here is because you're considering should I be bringing back these workloads, right? OK. Well, I'd say one innovation you missed out on.
07:27
What's that? Flash, flash. So honestly Flash came along, right? And we all saw it like in the past days, like, who here remembers when Flash tiered storage came out and you know your one flash drive was like $40,000 you know, and it was just like, hey, I've got a flash tier.
07:46
Do you have a flash tier? You know, and then all of a sudden Terry came out and it was just like, oh, how do we, how do we utilize this flash? Well, just make it a faster spinning drive, right? So again they all kind of cloned these things they weren't really taking advantage of Flash like they just treated it like another guy in
08:04
the pool, right? He's just a faster swimmer than the other guys in the same pool and I mean, so again like this, this is where Pierre saw the opportunity when we, when we were building file, we knew we could build it for Flash first off and only for Flash just like Pure did for Block. And we could take advantage of all the new technologies and build file that could take
08:28
advantage of all those new technologies. We didn't have to go backwards. We didn't have to clone people. And the industries came together to help us, you know, the networking industry did, the cloud industry did, because all of those cloud applications, what do they run on? NFS and S3 and some SMB,
08:44
right? That's what they run on. So that's what we developed. As we're, you know, what we wanna show you guys is how file at pure is really changing file storage from being this on prem kind of, you know, old school way of doing things to being a new cloud-like file that's inside your data center, right?
09:06
Because we what do we want now? We want things that don't move, we want things that grow where they lie. And who here wants to do another file migration? Raise your hand for me. I wait, not nobody.
09:19
And I think the other thing there is that who knows what is the true cost of data in the cloud? Do they charge you when it's just sitting there? Uh, a little bit. They charge you when it moves because they know that's really the cost. So we don't want you to have to move your data anymore.
09:33
We don't want you have to preplan where your data is gonna be at in 5 years. Put your data where you need it to be. Let it grow there. So some of the items we're talking about is, do you want to go over some of these, or you want me go, let's go go over together.
09:47
So some granum management was one of the items we wanted to make sure we had. Yeah, so this is, uh, fusion fleetwide management, right? And then, um, I like policies. If, if anybody's tried policies on our products, I'm a big fan of those. I do, so just so you guys know, I do a ton of demos.
10:03
I do a lot of uh customer conversations. I do, I don't know, maybe 4 a day, a lot, right, every day, so. You know, policies really resonate with the customers I speak with, especially when I show them how to use them, right? And then, um, yeah, fleet management,
10:19
who doesn't, who doesn't appreciate something like that fusion for file, which just means now you're like in the cloud where you're managing shares across your cloud environments. Now you're managing shares across your fleet of flash or even flash plates. Right? No more having to log into the Flash Blade and
10:35
then log into the Flash play. If you have file anywhere in the pure platform, you can manage it from one single place. Yeah, agility and automation. We're gonna talk a little bit more about that later on on the deck, but obviously we need to be able to be flexible. We need to offer you flexibility. We need to have a lot of the stuff underneath
10:54
the covers already automated so that you don't have to worry about performance bottlenecks, things like that, right? So all of that's actually already pre-designed and. Um, you know, part of what Purity offers is, as, uh, you know, our storage offering. So we'll talk a little bit more about that and, and other slides later on on the deck, cyber resiliency, obviously,
11:13
you know, we have to have a strong security posture, which we do data protection. We need snapshots. We need replication technologies. We need ways to protect your data with safe mode, right? Yeah, uh, and guaranteed SLAs, right? So this stuff has to perform, it has to perform to your expectations.
11:32
You know, I think anybody here that's a storage administrator or a CT CTO or any level above the application owner has to treat the application owner as their customer, right? And we're all there to provide them with a service and that's where Pure is actually stepping up and saying we're gonna guarantee that SLA for you, right?
11:55
And that's, that's unique to Pure and there, there's real like consequences if we don't meet. There's a whole program around SLAs, so we're talking about agility and we said we talk about it later, you know, unlimited file systems, right? That's the first part of agility, first part of agility. So I mean how many people here have ever filled up a file system on a previous NAS?
12:18
You know, eventually you do. It's like, hey, I can go to 256 terabytes. I can go 300 terabytes, you know, some people are invited like a cluster version of, you know, going several petabytes, but it's, it's complicated, right? It's a complicated operation finding out what to add to it, how to expand it.
12:34
It's not like, hey, I can just go today and say within 5 minutes I'm ready to expand that thing unless I planned ahead, so. Well, and how does it take, what does it take, Peter to expand the file system up here? How many operations? Uh, like none. It's one, right? Well, you create the file system and it's
12:54
unlimited and then and then if you have, if you want to limit it, right, so you have the flexibility of limiting it through then provisioning or, uh, quotas, um, and you do that through policies and you just set up a couple of policies. Depending on what kind of flexibility you're looking for, and then you just add the file systems to those different policies.
13:10
It's super intuitive. I don't wanna say it's simple, it's intuitive, right? So when you're ready to expand your file system to several petabytes, you just go and change the policy, 11 single command. And uh as long as you got enough space in your hard drives,
13:24
I mean, obviously we, we can add space always to all of our systems. I mean look at Pure. Pure is gonna be the first vendor with a 150 terabyte drives. They're gonna be the first vendor with 300 terabyte drives. Why is this so important at Pure?
13:37
Because we want you to leave your data where you left it at and be able to utilize by just upgrading your drives. I think we have 75s too. I don't think anybody else has them. I don't know, probably, yeah, I don't think anybody else has 5 terabyte. And we've had that for like over a year now, right?
13:55
So the next thing, why is unlimited possible is this, right? This is one of the reasons, pool storage. What I love about this is it just shows you showcases pure innovation, right? This was designed from the very beginning to be like this, to be global in nature, so we don't have to worry about storage pools.
14:12
We don't have to worry about raid groups. We don't have to worry about planning any of that stuff. We just write the data to the file system and let the array take care of how it's laid out in the back end, right? Uh, global thin provisioning, global data reduction, so everything on file, the compression and DD that you see on block,
14:29
we get on file, right? um, always on QOS. You don't have to worry about managing particular file systems. The array takes care of that. So yeah, I, I, I think whoever designed this, the smart guys in the back office or in the engineering office, they did a fantastic job with envisioning what they wanted to deliver to
14:47
our customer base. Yeah, and I think the same goes for how snapshots are managed, right? How many people have had to deal with snapshots being the same container as a file system, which then was held to a limit, and now you're planning, well, how much space is snapshot gonna take with the file system,
15:02
how much is gonna grow in the next 5 years? Where are they gonna be placed at? I used to call that playing Tetris with your storage. It's like, you know, I like playing Tetris outside, but I don't like playing with the, the storage. And so at Pure you don't do that anymore. It's like just create the snapshots in the global pool of storage,
15:18
you know, let, you know, let your data protection grow, let your file systems grow. Sorry, click here. So the next thing is you tie all this into Evergreen, right? The subscription for life at Pure of continued new innovation,
15:36
new performance, new capacity. And all non-disruptively, yeah, I, I have a, you have a I have a very different take on it. So, so my take is I, I'm a competitor, right? So I compete against other storage vendors every day and I'm like, well, why don't they just make a slide just like this one and call it something else for their products, right? How easy is this to copy?
15:58
So I started thinking about it and when I looked underneath the covers on something like uh Evergreen Forever, right, in array that you can update in place, um, the media kind of caught my attention when we design our DFM, our Flash modules, they're designed with a 10 year life cycle in mind 10 years plus, right? If you don't have a media like our flash,
16:21
you can't do a program like Evergreen forever because the cost becomes prohibitive for the vendor trying to imitate this solution. This is why it comes back down to the innovation underneath the program really Evergreen is a reflection of what we can offer because the innovation in the technology is there, whether it's the DFM technology or being able to upgrade the systems across multiple
16:46
generations, right? All that comes from Pure and only Pure can deliver it and the reflection of that is Evergreen Forever, Evergreen One, and these kind of programs. That's how I see it. I don't know. So making making files simple, I know files is not simple.
17:04
Files never been simple. Making it simple is very hard, and this has been a lot of work that we do at Pure is making it simple. And one of the biggest questions of the PM I get asked is when are you gonna do everything that everybody else does? When are you gonna have all the stuff that that guy has?
17:18
And I'm like. Oh well, did you really need all that? Did they do that because they had an inferior hardware or an inferior software or they had some kind of thing that they built, and they're trying to, you know, get around that with a different feature? Well, I'll tell you why I didn't want to end up happening and we didn't want pure and pure does
17:35
not ever want this to happen to a customer is that their storage looks like this. I mean, who wants to manage their storage like it's a 70 70s aerospace capsule? You know, there's a ton of buttons and things I gotta touch and monitor and I've got to figure out what this feature interacts with that feature. Next thing you know, you stir the tanks and half your your spaceship's gone.
18:01
Yeah, OK, you got it. I actually said to John, uh, we went to the moon in this, but I didn't talk about all the ones that blew up on the way to the moon. Oh, do, do. Let's see, honestly, pure is all about simplicity and,
18:16
and, you know, we know you need access to, you know, maintain, monitor and change and configure, but what we wanna give you is only we wanna try to remove anything that Pure can automate and self-drive so you don't have to worry about it. Why is that? Who knows what the majority of outages are caused by in storage? Mistakes.
18:40
Somebody made a mistake. I pressed the wrong button. I did the wrong config. I set it up wrong. When you have more configurations, more buttons to push. You have more risk to mistakes.
18:52
So immediately the way we do this, the way we get rid of that is we lessen the stuff you have to touch to maintain your storage. So this is what I want your storage to look like. Oops. Like this. Anybody seen a modern day dragon capsule or something like that,
19:10
you know, it's like a little touchscreen in there, right? Uh, and you have access to what you need to maintain the spaceship, maintain your storage, but it's modern. And why isn't everybody doing this? Yeah, they're trying to do it. They're trying to throw it up in the cloud, they're trying to throw it into like this other
19:26
thing. Pure's doing this on the array. They're doing this where it needs to be done, where you're touching it all day long, and fusion's adding to this to make it even easier. Yeah, versus the guy that starts off with the capsule trying to make this, we actually started much later than they did, so we were able to learn from what previous
19:46
vendors were doing and actually look back and say, you know what? Number 1, should we make it? And number 2, how do we make it better? Apparently NASA lost the plans. What's that? Apparently NASA lost the plans. They lost the tapes. OK, we're not going there.
20:02
um, so when we talked about civic layer management, this really comes down to policies and being able to to do this across many objects and many things in file, so. You know, I think the best thing about policies is that you learn them once. You learn one policy. You learn them all on file.
20:21
It's like, hey, you learned how to do a snapshot. Oh, look, do you want to configure and export? Oh, it's the same policy type. It's a different process but the same a rule and then add the file system. That's how you do a policy and then you can add as many file systems as you want and it allows you. To grow whether you're doing a system with just
20:36
terabytes or you're rolling out a NAS device with petabytes, it's the same management motion. That's what I like. I don't want to change anything. I don't want to come in after a month of going on vacation or whatever or not not being, not managing my system on a day to day basis and look at the screen and say,
20:52
OK, what am I doing here? I don't want to do that. I want to look at it intuitively understand it. If I've never seen a system before, I wanna figure it out, right? And that we can do that with our array. I think it's. Who knows about cellphones here?
21:06
Do you need a manual when you get a new cell phone? Does your kid need a manual when he looks at the cell phone? Because everything works the same as the last cell phone, right? You learned it once, you learned it on. If I grab my Android phone and he grabs his iPhone, I'm pretty sure I could figure it out.
21:20
That's what policies is supposed to do for you. That's what Fusion's supposed to do. Make it so you're doing things over and over the same way in a in a like way that way you don't have to retrain yourself every time. And everything we do will continue to follow this as we had other solutions. Oops, I went too fast.
21:38
It's OK. We have to probably move a little faster. Oh yeah, so there is a session on perfusion. I'm not gonna cover too much, but Fusion is gonna be for file. Um, you're gonna see the cross flashing flash blade. If you really want to go into this, I would go and view the session on fusion,
21:53
so I only I can save time and all our Green management goes to NFS data source. So we did data sources differently. We said we want to make them. Almost viva like. So if you're one of those guys that uses NFS data stores on another vendor, you're gonna get a lot of added control and features using them on FA file,
22:12
for example. This is only on FA file and it's not flash blade, but, um, definitely here you still get like, for example, an unlimited data store size, no more planning where everything's at, like again, no more Tetris. And there are policies for this as well.
22:28
And then again, we just, you know, since we built it, we had to integrate it with the rest of our our visibility platform and that means it's in virtual machine analytics and pure one. OK. So we talked about making it simple, we've had all this stuff, but we still have to keep files secure because files inevitably get into the wrong hands or
22:52
they're attacked by a cybercriminal. How do we protect you guys? How are you, how are we doing it differently than You know, the competition. And it is not. Social Security came there we go, there we go so.
23:11
We have recently released on flash ray file servers and we've already released on flash blade file the servers. This is the ability to isolate data access to different users and groups and identity sources. So anybody familiar with this is a pretty commonAS ideology, you know, we all had, you know, Windows file servers in the past and you know they connected
23:35
to different things. They showed different name spaces they had different export names and things like that. Servers on pure file allow you to do all that and some. What we changed is that in the industry, everybody's made servers, direct and only the direct owner of data.
23:57
They weren't there to just control access. They were there to control the data. And what that caused is it made so you end up with these silos where data was not accessible to other resources. We're not doing that on Pure. We've made it so you can take multiple servers and actually share data the same data set
24:16
between them if you choose to. This means you're no longer having that data siloed to one server. So in many cases where you know people are trying to build environments where they have a DMZAD for external users and an internal AD for internal users, they, they need to have two servers accessing the same data.
24:34
That's not possible traditional legacy file. But it is possible with Pureile, yeah, or maybe you're an MSP and you have a lot of ISOs that you wanna roll out to your people and you just wanna have it. That's another way of doing it, right? Yeah, or you're trying to reCL something. You're trying to get rid of an old domain that you do not want migrated into your domain.
24:52
Everybody understands the the concept of multi tenancy multis server, right? Just so we're all clear. OK. So, Who wants to see it in motion? Yeah, to see it? It says me fat fingering uh you gotta talk over this because I don't have yeah, OK, um, yeah, so I'm gonna go into servers. If you're not familiar with the top,
25:14
it's, uh, this is a feature release code, so this is the latest feature release. This is going into multi-server. That's the base array. Right, so if you wanna see what the base array is exporting, you just click on that. Um, I'm using this, and this is what I'm showing you here.
25:28
These are the exports. This is the interfaces that the base array is using. You can see it's got, uh, that IP address. It's added to, uh, this domain, right? Um, you can see, you know, where it says remote directory services, you see the DNS configuration, local domain information, which is local groups and local
25:47
users. All right. So the base array, everybody knows that we've found a single array on flash array file, flash wave file for a long time. We wanted everybody to be converted and know where that that server's going. It becomes the server.
26:01
Now you can add new ones and, and I'm gonna add one here. And so I'm clicking on the plus button over there and so what I got define here is the local domain, right? So this is my local users and local groups and so I'm gonna go ahead and the base array is called domain. So I'm gonna call this uh serve or domain A or should I call it server A and then,
26:20
um, the local domain, right? I'm gonna create a new one that's the only option in that box and then I'm gonna give it the domain name, right? And that's where I'm gonna select domain A. Have to define that. I'm saying that up there, but you know, a lot slower.
26:37
I actually uh did this video and I rolled it out. It's faster under pressure. Yeah, I, I know I'm nervous, man. I didn't have an audience while I was sitting in front of my laptop there, you know, so I'm gonna create a new virtual, uh, interface, right? So it says create new.
26:52
I'm just gonna leave that there. I'm gonna call it server AI. This is internal, right? So I know what it is, OK? So server A is my internal name for this system and then, um, I have the option of creating a VLAN. I'm obviously gonna be in the same domain, so I didn't create the VLAN.
27:06
Um, for this particular instance, and then I'm gonna give it a gateway, uh, for this particular server, and I'm gonna give it an IP address. And I'm gonna pick the sub-interfaces. So this particular system only has two interfaces that are available because it's a lab box, um, but, uh, typically you're supposed to have 4, right?
27:24
4 ports, right, for H8, uh, port failover as well as controller failover. Right, and then I'm gonna pick an MTU size, so I'm just gonna use 1500 to default, and I'll just note you can create any config from a single config from a single page. You can create LACPs, VLAN tagging, combination of those two, anything. Yeah, so I, I could configure DNS if I wanted
27:46
to use a different DNS server if I was in a different subnet or if I'm doing a different domain or a different forest, maybe I'm gonna be. Using a different DNS, uh, in this case I'm gonna use the default DNS because it's already been predefined by the base array and it's already pre-filled out, right? So if I just select base, it's gonna use
28:03
whatever the base array had as the default and it's using the management interfaces. So that's, that's I said cool I I like that. That's why I'm gonna go ahead and do that because you know I only have the same domain, so it's not like I'm gonna make a lot of changes. So source network interface just allows you to say where the DNA traffic will be going over.
28:22
So if you wanted to go to the same interface for the server, you can do that. If you wanted to go out some other interface, you can set it to that. Yeah, you go out the file interface, for example, that I just created. And so in in this case I have, I'm setting up the Active Directory environment,
28:34
right? So if you, if you know Active Directory, you have to set up an internal computer name, right? So I'm gonna give it server A I think is what I called it and then. Yeah, give it a array. Then I'm gonna pick the domain, the DNS for the domain name uh the domain DNS name,
28:50
and that's what I'm gonna put in here, right? So in this case I think it's uh Chi Lab something, yeah, Chilab PureStorage.com, and then I'm gonna, I could define the OU. If I wanted to, it's going to go into computers by default, and if I wanted to, I could just go into Active Director users and computers and move it to the appropriate OU.
29:09
Uh, in this case I just left it blank because, uh, the default is exactly where in this particular domain we put computers like this. So, um, and then I'm gonna define the computer name which is the object name and active directory, right? So this is what my clients are gonna be looking at and, uh, I think I called it I we have a naming convention.
29:28
I think I call it whatever the base name base of the array is, uh, plus the letter A, and then, um, come on Pete talk faster. You can do that. There he is. And then um I'm gonna put in my user and password to add the system to Active Directory. So the computer name and this name above don't have to match.
29:53
One is just the management name on the array. So you can call it, hey, that is the application team I don't like server. These are the meany guys. I mean guys. All right, and then I'm gonna pick TLS, right? I'm gonna uh say it's required, and then I'm
30:16
gonna use the default file network. So what they're asking for here is which, uh, source network interface do I want to use? I'm just, if I leave it at the fault, it's gonna use the one I created, right? So this is what I want. I want to use the one I'm creating, but it'll also let me pick one out of the drop down.
30:33
Right, so I could use the one I created Server AI or I could use file VI, which is the one that the uh base rate server is using. So I'm gonna just leave it as the default and let it use the one I created when I was on the second step called Network. This is just in case you got one of those network teams that isn't,
30:49
you know, picky and doesn't tell you what to do all the time. You know, this is last chance. Go back, go back, fix it, or just let it go. And so I just went ahead and graded it and it checks everything to make sure it's correct. If it's wrong, it will put a little red mark there and make you go do it again.
31:06
So just, uh, you know, I had to do it like a few times. I didn't see that in my video though. All right, and that's it I had to bleep out what he said in the video. So I created it. It's, it's up there and it's done. And now what I need to do is just export some file systems from it,
31:29
right? So I can't create the file systems in here, but um, if I have file systems, I have an example of both where I have an existing file system that I decided that it's already exported somewhere else and then I export it again here and then I create a new file system and export it here just so you can see how to do that as well in case you're curious and then I look at it,
31:47
I go to browse to it through a client, right, which I'm gonna happen now. OK, I think we're gonna go forward to this. I think we're gonna I think we're just gonna pass it, let me go past it. Yeah we see it, we're, we're, we're short on time. Yeah, that's, that's how you do it, um, you know, so you will see servers appear in realms,
32:07
which is released last year for blocks. You'll eventually see them in realms. It adds another layer of multi-tenancy to servers where you can isolate for delegated access and things like that. I'm gonna be rushing a little bit here? I think we're gonna have to rush, OK, yeah, security.
32:21
So we did release a in last year, um, in addition to all that, and then obviously we support safe modem files. So, so you have a story for safe mode. Well, we're gonna, yeah, I know, but let's talk it's it it's worth it, I promise you. OK, so we're gonna go through these. So, right, well, while he's doing that, um, I'll, I'll go ahead and share my safe my story.
32:45
So I was actually at the EBC executive briefing, uh, a couple of weeks ago and I was talking to a customer and they just bought a system. Um, for Vertigo, right, a big S 500, and I was like, wow, that's pretty cool. We didn't have to do a lot of selling, right? You just bought it.
32:58
I was like, wow. And I was like, well, so what made you, what made you make this decision? Why did you reach this decision? I was really curious to see what it was, and he said that, uh, several years ago, I guess he had a Gen one flashlight and he had safe mode enabled and he had a security, uh, breach and his data was corrupted and he was able to roll back to a
33:19
safe mode copy and recover from it. I was like, oh well, thank you for your business. I appreciate it but you know that's what convinced him to do it because he says everything that you guys roll out works and, and we've had, we had the ultimate test there with safe mode because that is a hockey stick move that's a
33:36
hockey stick save if you're if you're going to safe mode just to recover your data really is right, so anyway, I I just saw somebody get attacked today and we were able to recover. It was literally in a chat today it's, it's amazing so. You know what, you've heard enough for us. I think it's time to hear from somebody who uses this stuff. So Jeffrey Wilson has joined us from One Gus.
33:59
Thanks, Jeff. So you know we're gonna just kind of sit here and chat a little bit. Uh, we had a little bit of time to hang out with Jeff earlier and you know discuss you know some of this, but you know Jeff has adopted a FA file, um, you know, and I wanna give him a chance to talk to y'all. So Jeff,
34:17
why don't you tell us a little bit about one guest. Uh, one guess, we've got, uh, we now have 16 pure flash arrays, um. We uh We started, uh, testing with file services probably a year ago because our old system was going end of life and was no longer had software support, so we started,
34:41
uh, moving test shares over to Pure did some testing, then we moved all non prod and then on, uh, Easter Sunday, the old solution died and could not be resurrected. That's always a nice way to force your way. So instead of having to negotiate outages with all the production teams, we were able to get all the production NFS shares moved to Pure in about 2 days from.
35:09
The, the DRA and at our DR site or from backups. So, so what's I, I. How many people are on your team? How many people in the system? Me. So how important is it to you that something's like easy to use,
35:30
kind of, you know, doesn't break a lot. That's important. How important is for one gas? Oh, it's, it's very important. I mean, we, we provide natural gas to, uh, customers in, uh, Oklahoma, Texas, and Kansas, and you know,
35:45
without natural gas they could freeze, not produce electricity. It's pretty important. It's pretty scary. Um, so what has, how have you found the file services so far? Uh, it's been great. It's, I mean, it's been great. We've got 20 terabytes on there. Um, uh, one of the benefits,
36:04
uh, is gonna be performance. We had a 3 terabyte share that backing up to Commvault used to take 30+ hours. Now takes an hour and a half. Wow, hm. Um, With um, with, uh, yes, we,
36:26
we do use Intelisan. So, um, what do you think your favorite part of file services is? Uh, well, performance, reliability, uh, we get all the benefits that come with being on pure. We get D dupe, uh, safe mode snaps, uh, have you done any upgrades? Um, for firmware, yeah, we've, we've done a few, yeah, I,
36:56
how's that going, uh, seamless as, as always. I do it through the edge service myself. Oh, you do the self-service upgrades. Oh wow, OK, OK. I didn't know you. I, I think we didn't even mention that. So you know, if you were to go back and do it all again.
37:12
Um, Would you waited so long to adopt it? Definitely not. It can save that. Few days of uh scrambling to recover. OK, no, I think it's great. I mean, um, do you guys have any questions for Jeff, what made you choose the file on the FA rather than going to the file.
37:39
We didn't have any flash blades and we had some flash arrays that had capacity. So, so just everyone knows there's no charge to enable file on a flash array. If you have an existing flash array and you just wanna do file like Jeff did, just, uh, get it enabled, talk to your SE and he'll have it enabled for you and you're ready to go. That's it.
38:02
I enable um let's say 8. So 6 of 683 or 682. It's enabled by default so you can just go create a file system run through the but if you're on the 65 or you're on 65 or 67 support as a tunable. um I would suggest you be on um 66 or 67 because we have dynamic allocation
38:28
of resources then. So turning it on is not a there's no like there's no penalty to your regular workloads. You want to be on 67 because that's a long life code release and we're still working on it, you know, bug fix and stuff like that. 66 has been deprecated so you know there's everything's gonna be. 68 for the for the code,
38:48
you know, for the engineering focus so you wanna be on 67 if you're gonna roll this out, OK? And then if you really want to, you've got an existing array, you wanna go play, you wanna put a you you're ready to adopt the workload on there. And i one, you can access workload planner and files supported on there so you can just sit there and say,
39:05
hey, I got this existing array. I wanna add a little bit of this file workload to it. What does that look like? And it will tell you and it takes all the costs involved so and if you want someone to, if you wanna, we have a team, um, FSA team like myself, and I'm happy to help you walk through it. So if you wanna learn how to use it, just want
39:24
somebody there to help you add it to the domain. Uh, help you set up the, uh, uh, interfaces, anything like that we're happy to do it, so just, uh, reach out to us and happy to walk you through it. Usually in about an hour you're like a professional. You don't need me anymore, right?
39:39
Or if you do, Pete's really good at repeat, repeat. I repeat this guy with a short jokes. I got all day. I got all day. I sit back down because I don't wanna keep you up all day. Do you have a you so if you guys want to hear the flea.
39:55
I didn't give the flea joke because John was late, but, uh, you guys want to hear the flea joke, OK. So there's a, there's a, it's not a very good joke, but it's a great flea joke, OK, so. Uh, there's a scientist and he's working with fleas, so he teaches a flea how to jump on
40:12
command, OK, so he says jump and the flea jumps. One day he decides to cut the flea's legs off. OK, this didn't really happen, by the way, so don't get upset. We cut the flea's legs off and then he tells the flea jump and nothing happens. So he goes to his colleagues and his colleagues ask him about uh the flea experiment and he and
40:31
they said well what happened with the flea experiment? And he says as soon as I cut the legs off the flea got deaf. OK, you guys flee out of here. Now I know why you didn't get the PhD. So back to my flea story now that we've had the flea joke,
40:51
you know what I really want you guys to understand is that we are a pure we're all about enabling to do things that you didn't think you could do before. That's really what we're trying to get across to you and that flash, you know, flash array and flash blade file are there to enable you to do things that you didn't think were possible with mass.
41:09
Because we built them in a way that could not be done before. People were still building things for spending drives. They were still building things for things and funny enough, all those vendors, guess what, they still have those systems in place. They're still supporting spending drive systems. They can't change it because they break all
41:29
their existing customer base, so they can't even do what Pure did. And so we want you guys to be able to see that you, you know. If you haven't tried file Services, or if you are, let us know what you think. If you haven't, I think you should try and see how far you can jump outside of the jar.
