00:01
Mm Good morning, good afternoon. Good evening everyone. Thanks a lot for joining us in this session today for accelerated breakfast 2022. Today we are going to talk about business continuity. Business continuity at no cost for your business with active cluster,
00:28
let me introduce myself. I'm veronica Gruel, I'm the senior product manager for replication at pure storage and Zane. I'm Zain Alan, I'm a principal technologist with peer storage and work with with you've been here and pretty closely on this replication. Glad to be here today. So let's get into it Zane before we go into
00:49
much detail about active cluster, I think we should first talk about business continuity itself because you know, there are different aspects of availability continuity, disaster recovery and why don't you tell our viewers how is business continuity different from the expectation and understanding of disaster recovery?
01:20
Absolutely. So yeah, this is a great place to start and what we like to explain from a a solution standpoint, when customers are thinking about how to provide, you know, what am I going to provide for my business business continuity, disaster recovery, these the same things and it really comes down to, as it says here, the the scope of the failure really, when I think of business continuity,
01:44
it's I would more localized, more rack to rack or regional, you know, so the scope of the failure being single point and eliminating that single point of in your in your data system and and really going towards that zero, zero ish data loss or zero data loss if you can or zero downtime. So so how how close can I get to a failure that things just continue over if there's a single
02:09
point or Iraq failure power failure 11 local aspect and you know, so I could, I could lose that that piece that one piece and still have no impact on my business operation. I think that's the other crucial part when talking about business continuity is no impact of the business things just kind of run over really automatic if you will right automatic to this other fail over scenario piece you will
02:34
see and and really just keep things going. Should should I have a lot somewhere It really classified for your tier one applications but the applications that drive the business that again, you know, I I don't want any impact. Should you know, some single point of failure happen. You kind of contrast that with disaster recovery and we kind of spread this out,
02:54
this is more regional failure I think would be the so local localized failure if you will, maybe it's a metro region but you know, if that whole regions failed or that whole building natural disaster an event like that, you know, what am I going to do with that data then write that still trying to get to, it's close to maybe zero as possible maybe. Right, depending on the system you choose, but it's not gonna be a zero downtime,
03:19
right? There's gonna be some sort of failure. A low touch recovery is a good point here to say to maybe, you know, automatic no touch on on business continuity. Low touch on disaster recovery, meaning maybe all the parts are in place. I know what is going to happen is predictive fail over to some degree.
03:37
Maybe even something I test, but I do maybe have to hit start, have to hit, go, make that decision and then all those pieces fall into place. So, kind of a stand by if you will with the man you think of. So that's that's great. Essentially, what it then comes down to is business community is continuation of business.
03:58
Even when there is a failure of some scope. And we're saying the scope is comparatively smaller. Whereas with VR, it's a huge disaster. Like maybe a natural disaster. And business would not continue automatically. So it would require some sort of intervention. Yeah, bring it back.
04:21
All right. So, how do you suggest suggest are understands that customers decide what kind of their applications would fall under the business continuity realm versus which ones were b falling under the disaster recovery category? Yeah, It comes down to a few things 1st. What is the the the monetary
04:53
impact? Right. What what is the what's the dollar amount that I can put two if, during the disaster of failure. Right. If if this part of the business application, service of data, whatever it is, has a failure, you know, what does that impact and maybe minutes seconds our you know whatever that is depending on the business and I think we'll have some examples
05:14
later that kind of kind of illustrate that and then what is the most, you know from from those systems, you know, what is the most critical age, be it information that needs to be provided for for the internal business or for my customers or even just business processes, right? If if this were to stop what down the line goes away right. I I worked in newspaper in I. T.
05:36
A long time and I used to say that if one part of the chain like all days things are just kind of moving out the door to a newspaper, if one part of those change kind of as a hiccup has a stop it just escalates right, it's like dominoes and critical information could be like also the stock market news or stock market information, the tickers that go and those decisions that are making, why is it value? You know based on sales and and and yeah
06:01
absolutely. And then how does I. T play a part into that? And I would probably say it probably plays a large part because you know, most of everything is in the computer systems now but you know how does it play a part in that meaning how much does the business rely on information services, information technology to provide those services, those applications and information.
06:25
So you know you know having this understanding of you know what's what's our impact from an I. T. Professional standpoint then of course what you know now that I have this information what recover capabilities, availability what we'll hone in here capabilities do I have at my disposal? You know that's justified to cover those those first thoughts that we have here?
06:48
Right? The criticality the amount of downtime what that costs? Right? So kind of balancing that cost benefit. Oh I see how much are you willing to spend two to achieve that makes sense. Criticality of of your business if I spend this doesn't make sense because I'll lose this if it's down.
07:08
Yeah so for business impact analysis if these are the critical components how would our customers go and kind of do it? It's it takes discovery right first and foremost and that's tex technical discovery is a data gathering coming out of the I. T. Room and understanding these components that we outlined above and really I think that comes down to again coming out of those silos right
07:39
human conversations talking to the other parts of the business right? When I was in I. T. You know I had to realize that early and it's actually kind of a pivotal moment in my in my career when we had we had some down time we had some data loss and you know this was a while ago so systems weren't as good and and we as an I. T.
07:59
Department we're we're covering this right? I'm having I have backups I have a way back but in this scenario when we did recover it wasn't the expectation that the the business had right? I think in this instance that I'm thinking of it was from tape is a while back. We some departments lost an afternoon some departments lost a day of data right? Or and it took us some time to come back. We were really proud of ourselves because what
08:24
the tools we had at hand but that is not what the business expected right? But that conversation Probably wasn't really had right. It was it was my junior very junior 19 it was a very pivotal moment too come out of your silo right. It's very important to evaluate the business impact not just based on the monetary numbers that you might have but also validating
08:45
and having conversations with the right teams to further substantiate that I. T. Team is in the right direction and right understanding of what those would be. Yeah if I fit into my budget this my backups this is the time I want to spend. Okay that's great. But does that that the business need?
09:03
Did you did you take the time to do the above steps? Yeah so we discussed a bit about you know business continuity versus disaster recovery and then how somebody would kind of go about bucket ng those in terms of what falls under, what kind of application falls in the reward. But at the center offered, what would be those metrics that you were
09:26
kind of drill down to finally that everybody has to agree upon as A S L. A. And then stick with it. And and that metric kind of measures what are the most important business critical apps, sources, you know, comparatively less critical apps? Sure. Well, I think first foremost it comes down to
09:49
this and and and everybody listening might might be familiar with, this might not be, but really? R P O and R. T. Oh right. Our recovery point objectives and recovery time objectives. So if we look here like when the incident, incident happens,
10:03
when that disaster happens that downtime that lost whatever it is, we first look at our P O how much data can I basically lose? Right. How, how far back can I go to get me back up and running? Right. So you know, what, what systems do I have in place if I were to recover to it or restore to that place,
10:24
could I lose that day that half a day? Those two hours, whatever that. So fitting applications into that. Right. And applications and information kind of spans throughout an organization usually from none to forever. And of course the answer originally is going to be none.
10:40
Mhm. And I used to tell that's great. You know, but how much money and how much time are you gonna give me this is what if that's going to cost? Okay, let's really have a conversation. Right? So really really defining that helps with S. L. A. And then of course there's RTO.
10:53
So the recovery time objectives. So I have my, how much data would I lose when I do a recovery and then how long is it going to take me to get the system back up and running? That's just as much. Right. Maybe I have only a couple of hours, lots of data, but if it takes me eight hours to get up and running well now I've lost eight
11:09
hours of time. Right? So those are that's a very crucial crucial piece to and fitting your applications and services into those two is a great place to start former defining sls and picking the solution and then just find the cost based on that escalator. I see. So, so at the end of the day, basically kind of
11:28
categorizing your apps from an IT perspective and from a business perspective and coming to a conclusion that what apps should be in what kind of R P O R T O requirement buckets will be the way that a customer decides business continuity versus a little bit er recovery. Absolutely. And then and then fitting is eventually mapping the solution to each one, right?
11:55
And I think it's a great example of it right here at getting those. So having that first step those priorities and we can see this is mostly about recovery but this can also go into recovery point objective but a recovery time objective. Okay. Here's the priority one. These are the class of applications and services that our immediate recovery,
12:14
high availability really probably no data loss or as much as little as data lost as possible. And then you know down the line, 2, 3, 4 and then five of the best effort. Right? This those really going to D R in my opinion too because you could say with with a plan, I should say like a run book, got my high availability and then as I go down the line these are the systems I'm bringing up
12:36
first these are the systems I'm bringing up second these are the ones and you might even go into that to go a little further depending on how long the outage is, if it's kind of a temporary fail over maybe priority five, I never kept to bring back up because it just wasn't crucial to business. I bring it all back to the regular data center, fix the problem and then everything goes back
12:55
to normal there. But but having this S. L. A can help you map the solutions. Are you going to solve each and then put your services into each one of these buckets and makes it really easy. Alright. So now we come to the important question, we discussed business continuity, we discussed our customer can identify what applications
13:18
are important for business continuity. Now let's talk about the solution, you know, what is available in your storage to address business continent he needs. Yeah. One of my favorite subject to talk about with with flash array here is these are the out of the box protection features, right? That we as soon as Pierre provides on our flesh
13:42
race so you can see it here in the context of our P O versus Rto. So we talked about that concept. They classify those in those sls and first we start with our feature called active cluster. That is business continuity. That is R H A. That is the right near the incident automatic fail over correct Iraq.
14:01
It could be metro, You know, usually local if you will in a region but keeping the RP 020 recovery time is Basically zero. Right? It's flipping over, you might not even notice. Oh, that single failure happened, no impact to the business. Everything keeps on running my most crucial application of up and going and I've protected
14:21
that with the built in feature of flash. Right? Then we go down the line, what we call active D. R. You see that line kind of going out a little bit more so maybe the R P O S. You know, we say nears Near zero but it's a little more the R T O. The recovery time might take a little coordinated effort.
14:39
Again, this is that er scenario and that's why it's VR that's why it's D. R. And, you know, you might have a system on their if it's true D R. That you wish was active cluster but you have the kind of event that needs that D R. That's fine. What that means is you'll have a probably a i very much scripted a run book of how you're
14:57
going to do that recovery and we set up a lot of great tools in there too. Make a lot of the pre work done, you just have to bring it online and choose to bring it online then of course a sync replication, This is our snapshot replication from one side to another and you can clone or or reuse, take it to any distance in the array, even off the cloud. Right. Just for storage,
15:18
but more coordinated effort to get it back. So more of a near term recovery. And then of course you can offload what's called snap off load for very long term archive. Now I had that data kept for a longer period of time, I got to coordinate it to bring it back if I wanted to use it. So really kind of getting away from D.
15:37
R but, you know, still something we can do and then and then local snapshots at the bottom, that's more about our P O. And then recovery time and recovery time is actually better as you see from a long term or off in the cloud because it's local but it's a R P O might be less often. It might be twice a day, three times a day every once,
15:56
maybe once an hour. Right. But different R P O setting but it's local to the array is something I could flip back or you know, clone off into another volume. So this session is focused on business continuity at no cost and for that based on the previous discussions we just did. It's essentially that a customer would want High availability with zero downtime,
16:26
zero data loss or RPO automatic and automatic. So active cluster is the solution that fits into it and I want to just highlight that because it is out of the box and because the session's title is at no cost while all the features that Zane discussed, all of them are free. But I do want to highlight again that active
16:55
cluster itself is also free and I want to share that. It's not easy investment for pure rising, it is a lot of engineers that have worked to deliver active active synchronous replication capability for last five or six years, that's when it came out and we have made that free and included impurity for you for all of our
17:29
customers compared to there are some other products, some competitors that don't give it out for free. So no additional licensing here. Right. Yeah, no additional licenses you have it today. No additional license, no additional hardware. And I feel very proud to say that. And to repeat that,
17:51
that it's a lot of investment dollar wise, engineering wise, and we have built a very good product, right. It also isn't that we just did the bare minimum to do a check mark and let it be for the sake of it. We actually built a really, really resolute and really, really good solution that is available for our customers out of the box without no extra hardware and license.
18:18
So enough of that, why don't we get into the actual active cluster and how it works? Yeah. No. And to your point, all that engineering effort has removed all the complexity. We've done the complexity for you. So as a customer and I know you'll get into it can very easily use this and have that H A that we talked about out of the box. So, absolutely,
18:43
you know, like we're going to go through the customer success stories in right in this deck for our viewers and some of the key highlights that will come out of that is it does what it's required to do. But it also does that in a very simple way. And what that delivered as a business value to the customers is requiring lesser game and less people to manage synchronous active active symmetric
19:15
solution. So enough of that cluster? Yeah, yeah. As as as a product manager replication, Yeah, why don't you and dive into it at the active cluster unless you can tell everyone how it works. Cluster is a product which provides symmetric active active capability. And what that means is when you have your two
19:39
flash arrays and you have them in a setup of active cluster, then read and write. Both are possible via both the arrays. So we have let go of the traditional concept of active passive synchronous replication. We big and synchronous replication with active active read write capability so that you are able to leverage your
20:07
applications on both the arrays if you want and rights can happen on both the arrays as well. The second aspect is transparent failure, which kind of goes to your point saying that in high availability with this part of the solution, the expectation is zero downtime and zero downtime is only possible when there is no manual intervention and the failure is automatic. So the apps and the host that are using the
20:35
array for their storage requirements does not even know that there has been a failure and the workload just continued to operate with this automatic transparent fail over to the second array that exists in this active cluster combination. The third aspect I want to highlight is while you can do Active Active between two arrays, there have been real world practical
21:09
requirements that the customers also want to protect their data in the concept of you know, D R for example or backup to a third side which is further away regionally so while Active Active is available within the same data center or in very close by data center, within the same metro region, you have the capability to pan out, you're data in terms of snapshots and using really
21:41
replicate those to a third side which is far away in the distance. And we begged that in the workflow of active cluster to achieve that kind of a use case, you still, you still might have to protect against the regional failure. Yes, more rare case. So I'm always up time with my H A but I still might need to do that and we make it very easy to to add that on.
22:03
Absolutely. And then I mentioned there are no bolts on and licenses, which means that once you buy the flash array and you are our customer, if you are subscribing to the subscription, you get purity and the purity, you get active cluster. So there is no additional cost. There is no additional license needed to
22:26
acquire and implement active cluster. I also want to add, there's no additional hardware also needed to implement active cluster like you don't need a separate dedicated hardware to just set up replication between the two arrays. It's all digged in the Flasher itself, you know, which is something that was quite different
22:49
when I joined Pure because I've seen other solutions which require a separate piece of hardware to enable replication, if Well, that goes to the next .2, or even to mediate it right, you don't have to provide anything there. Yes, so, so there's obviously simple management and I'll show you how the next few slides. But another important concept, which is very
23:16
crucial is the mediator and mediator is our technology that provides the resolution for split brain situation in terms of failure. So when there's a failure there has to be a decision that which are a actually takes the workload going forward and for that, there is a cloud oriented mediator that your provides for active cluster which is hosted
23:46
by us in the cloud and the customer as in you don't need to maintain or set it up in the third side in your infrastructure as well. So further simplifies your high availability journey with cloud based a mediator. To break ties Between the two. Flashers in a split brain failure situation.
24:12
Yeah, no, I think that's a that's a great part of it. So why don't you, if you wouldn't mind walk us through what that looks like, like absolutely an array goes down, yep, so we're going to go through three steps that I wanted to showcase here. Step one is, I'm just going to show how the whole layout of active cluster looks between the host and the arrays And how kind
24:38
of data is being replicated between the two are mirrored between the two. The second we'll go through is what he said in case of failure, how does a split brain situation is resolved? And then the third I'll go through Is when there's a failure can happen but operations continue on one array and then afterwards the second that it comes up,
25:02
how you don't have to do anything and the array kind of catches up on its own when the failed that it comes back online efficiently is the third aspect that I'll show you. So beginning with the first one, how does active cluster essentially work in this diagram? You have your host or hosts and they are connected to all flash array which is on our side and then that host is running a couple of VMS apps or whatever is your workload.
25:33
The first thing you would do for active cluster and for any replication solution, just logically, is get a second array so that it could replicate between, you know, two pieces of hardware. Once you have the second piece of battery, what you're gonna do is you're going to create a part and parts are the container that we provide in purity software and flash
25:57
rate that contains volumes of your interest that you want to protect. So you will create a container called the pod and then that party will move in the volumes that you want to mirror and provide high availability for the next step is uh stretching that part. So that operation means that now you are telling the flashlight that stretch
26:26
this part or replicate this part From your flash array on side one to your flash array on side two to mirror the volumes that are contained with that within that part. And once you do that, the flash array itself starts replicating, do the second flash array gets to a point where both of them are consistent and at that point the volumes inside this pod are also considered
26:56
stretched and what that means is now they are consistent on both the arrays and they're available to serve IOS from the whole and for the apps and they're highly available at that point of time. Now, one important thing to remember that when these volumes are exposed, they are logically the same volumes. So the serial number of a volume on flash rare one is the same serial number on flash radio.
27:26
So from host perspective, these are not two volumes but the same volume with just more paths or multi paths to now going to both. There is rather than just wondering but that's the simplicity offered from the, from the fact of how you configure and manage it and how the app looks at it because that's where that's how the failure becomes transparent as well. That's the big part, right as the host and the application, see this as one volume Living in 2,
28:00
2 places essentially, but that's all transparent um and like you said just it's just more paths to that to that volume, yep, The other thing also do note is for this setup, you really don't need to install any software on the host as well. So from maintenance and management perspective, your replication is well contained within your
28:26
flash Aries and for that matter contained within purity. So you're not going to upgrade or maintain a specific version of active cluster that is separate from purity, you get your flash array, it has the purity US security software and actor pluses is part of it big then you will upgrade it with purity, you will run it with purity.
28:51
The other thing that I also wanted to mention was I have seen in the field and across our customers that because the host sees these volumes as just multi party volumes like connected to the volumes. This technology is also used for moving data from a flash array to a second flash array non destructively.
29:22
So think about it that you have your flash array one On site one, you do the active cluster setup, you create the board stretched apart, stretch the volumes in that part. Now they are being shared between flash and a one and 2. What now you can do is go to flash every two and say I want to unstrap chit and make it
29:44
local now to the second flash away and because the host was just working on it as multi path volumes, but the same exact volume when you stretch it on the second side of flash array or the second flash array it becomes local to that naturally the volumes become local and you're hostile continues to connect to it. So from host perspective there is no disruption, that's like the second side use case
30:12
which started to get adopted and used by the customers using this technology. But high availability is the primary use case and and this kind of delivers that very simply very nicely non destructively to fail over to the the post an application now to go back to the engineering effort, removed all the complexity for you so you can exactly operations very simply and that's why I feel so proud that we build a solution,
30:43
we offered it to our customers with living by living with the pure value of simplicity and delivering innovation to our clients 1%. The next aspect that I wanted to share was the follower situation. Right. And how cloud mediator comes into place in this picture, if you see the cloud media is represented by the cloud and then there are two
31:10
flash arrays and the cross in between kind of simulates that there was a network connection that was severed between the do. Now in that situation let's say you have apps connected to boast of Flashers right and now because the flash arrays cannot talk to each other, they are not able to ship the rights from one to other synchronously.
31:36
And for that reason, if the apps continue to write on both the applications on, sorry, on both the flashes, it might end up corrupting the data and diverging the data on on both the flashes as well. So at this situation you need a mediator, a third party to say which of the flash arrays is going to win in terms of going to accept the IOS going forward.
32:05
And for that we have the Pure One mediator. So as soon as the connection is severed between the two flash arrays, what happens is both the flash arrays Send us a communication to the pure one mediator and say I'm not able to talk to the peered flattery am I the winner or can I be the winner?
32:30
And when they race the mediator then besides which of the area is going to be the winner. And then that array starts accepting the IO going forward. And the other array does not accept I. O from that point forward. And that's how we resolve the split brain situation. And from there on mediator
32:59
is has done its work. Also mediator is in listen only mode. So it does not communicate to the flash arrays or initiates the connection to the flash arrays from if if there's a security concern, is there anything to add here? No, it's just that it's again, take moved complexity, done all the work for you,
33:20
um cloud media helps, you know, kind of thought of all these scenarios to help make that decision again to get it more of a to an automatic fail over. So that decision doesn't have to be made manually and you don't waste more time or take more time to get to that recovery time. And I think I think that's the good stuff when you're saying the recovery time because that's
33:39
the third aspect which I was going to mention is let's say right let's say everything the media to raise happened. You got your one flash rest still operational for high availability but then your second flashlight came back up and now it's time to kind of restart it in terms of start to serve it, serve Iowa's from it. And this could have been a complicated task because your hardware is coming back up.
34:10
It has fallen behind so you need to catch it up, you have to figure out a way of how to move your delta data. Now that has been on that has been incrementally written on the are they that was leading or that was accepting IOS you have to move those those data sets to your clashes that just came back online. You have to make sure that they're consistent and you have to also re induct them to
34:43
each other so that they can continue in high availability manner going forward. And these could have been very complicated steps if we didn't have active cluster. So what active cluster is going to provide in this kind of scenario is as soon as the second flash that it comes back online because it was already part of the active cluster pair. The leader Flash it in at this point of time and leader I say,
35:11
because at this point of time it was the only one accepting IOS it will start shipping the changes that it has accumulated over time to the secondary to the second flash rate, which came back and they had fallen behind and it will do that in an intelligent way using asynchronous snapshot based replication to re synchronize the data so it can send bigger bits of changes towards the other side, rather than
35:44
consuming the bandwidth to continuously send everything all at once. You have to go back to that slide earlier when we talked about all the different functions and in true pure fashion, this has just been built on the innovation of the past. So yes, it definitely, I think that's a good point to mention that when we talk about active cluster, it does not mean that it's like its own just siloed technology,
36:08
we have reused all of our technologies bar addressing these changing components and these changing needs in different products. So, Active cluster also uses a sink in this kind of scenario, we're trying to, you know, make the other flashes to catch up and and and and and then flip over to a C. Yeah, automatic, Exactly, and that's incredible
36:32
because but if if we wouldn't have big that in then somebody had to do this manually or in an inefficient way. But for with active cluster it's just all happening by itself, yep. And then, and then the flash race catches up there in the same consistency and you're ready to go and you're again highly available.
36:54
So I think that covers the introduction to active cluster and I hope you guys like uh info, I hope you had some understanding, like I hope you have now some understanding of active cluster and how it can provide high availability business continuity at no cost but Zane, I think we should now go into where active cluster has actually proven its mettle and then
37:26
how demanding situations it has proven its mettle. So that those two concepts together, right? Making those escalates for your business impact, the technology of Active cluster and and and and some real world use cases here. Absolutely right. And that it is mature and has proven successful in delivering extremely critical business value to some of our customers.
37:52
So with that, I am going to ask you that, are you a fan of F1 Formula one racing? And I mean, who isn't right, even if you're not a fan you say, who doesn't find it exciting to watch those cars going around the track. Yes. Ever since I was a kid, I loved cars and when you love cars you always chase the fastest car and F one is the fastest car.
38:16
So I love the sports and to my ah surprise when I joined pure I learned that The Mercedes AMG Formula one team relies heavily on technology for analyzing data both off and on the track for their car. We're not talking about web apps, we're not talking about their website,
38:46
we're talking about there Car that they used to win the Formula one racing championships. They collect millions and billions of data points in that car. That that sounds counterintuitive but now that you mention it I mean cars are Especially if F one cars I'm sure are very have just as much technology and computing them nowadays is they
39:11
do mechanical technology. So that that makes a lot of sense. A lot of sensors and things that they need to be read And as you point out real time as that thing is trying to win races. Absolute. So think about it that the F one car is running on the crack and a bunch of data points are being collected and the team is analyzing those data points to you know, get into the details of every aspect that can provide that you know,
39:39
just a little bit more impact for them to win against the really next competitive real time decisions they have to make as you said that when it stops when the slowdown tires need to be rotated? Well yeah, that make that competitive edge right there. And then now what happens if suddenly your data isn't available? You know, the car is running around the track
40:02
and the data isn't available. All the dashboards like show up empty and you're calling the I. T. Guy wondering why, why the system is down, open up a helpdesk ticket to get your race car. That's speedy around at 200 plus miles an hour. Yeah, so I'm glad that it doesn't happen for the Mercedes scheme because they use active cluster for making their data highly available while it's on track and off track.
40:25
Yeah, it provides Mercedes a competitive advantage for performing those analytics and active cluster works fine for them. It delivers what it says, high availability for them with zero downtime, zero data loss, help them win races. I love it. I love it too. Now let's talk about the next one.
40:50
So not far away from a speedy car, but in the realm of automated vehicles, why didn't you tell me about the Empresas Kharrazi case? Yeah, I like this 12. Very interesting. I mean a little bit of a contrast from a race car going around but still these real real time decisions we should say. But real time decisions made by robots I guess
41:14
as they and in this case automated processes inside of a warehouse. But these are vehicles, things moving in this case manufacturing, but food and processing across warehouses and and from place to place. And you can imagine if something were to go down with all those, I'm sure there's just as many sensors, if not more in this,
41:36
in this factory, this manufacturing plant. If one of those were to go down in those automated processes and robots and things speeding across floors, couldn't understand what the other thing is doing or when to catch up or you know, when to be ready to maybe take something from some place to the other or not crash into each other as their automated across. Even for even for,
41:59
you know, mapping and yeah, you're right. Even for mapping for knowing where the other automated vehicles our while while they're moving around the Yeah, while they're moving around the contents of the warehouse for the food processing plant, I think this is again the real time, decision making use case and can't tolerate downtime, right? I mean, because it kind of goes back to my
42:22
example, that's more extreme than, you know, when I was in the newspaper industry, one little blip would cause a ripple down the line. So, if, you know, I don't know that the things that right here, I can't take it to the next place. If there's a collision, just all that would impact business and I'm sure they measured downtime and minutes.
42:41
If not seconds, right? If I have this many minutes of downtime, this is going to cost us this and our food, you know, the cost of our product is gonna have to go up and we're gonna revenue is going to go down because of it. Yeah, Well I'm happy because they were our first customer in latin America using active cluster. And they did test intensively active cluster by
43:02
disrupting one side of their flash array to make sure that their robotic vehicles still operate, didn't notice and and they did without disruption Providing a 100% availability. Look at that automated too. Right. As we said, so it fits that criteria. So this is a perfect example of tier one. Now let's get into the realm of sports. So we talked about race cars.
43:28
We talked about automated robots. Are you a betting person? Well, I feel like with the stock market nowadays I am but I think I understand sports betting a bit. Yeah. So our next customer is sport radar and they are worldwide
43:49
platform for providing sports information and sports betting and gambling platform. Now, how would you feel because you said you work in the stock market a little bit, how would you feel that you have your, you know, a bunch of million dollars that you're making already. You have those in in your stock market account and you see that the market is quite
44:15
volatile and there's a good opportunity to make money but suddenly your quotes are coming up empty because the system is down the website that you use for your brokerage is down and you can't get the information of what the stock ticker is, you can't submit an order, you can't execute an order. How would that make you feel?
44:36
Yeah, yeah, we've become so reliant on in the world in general, but definitely something like that, especially with the fact that I can see it on my phone, right? And I can make those decisions and, and you know, I'm not not day trade or anything by any means. But you know, when I when I do make those decisions, it is based on information,
44:51
I could have, I want to put the order in and then I expect it to happen and you're right, if it if it couldn't, that couldn't the timing, there could be crucial to the why I made the decision in the first place, right? You know, things could change by the time it catches up if there's down time. Absolutely. I'm sure that's the same,
45:08
especially in the sports world where sounds like this company does a lot of data points and makes odds change and things like that. I mean think about a basketball game, how long is there then? Few hours you're gonna ask me and ask me basketball questions. Been told me, how about, you know, how about football, how long is that?
45:32
Yeah, that can be a few hours. Well it's yeah, depending on how long and commercial breaks and all that. But yeah, it's it's, you know, it's a definite set of time and it opens and closes and yep, there's time zones and so think about like, you know, it's like 90 minutes worth of game and then you are trying to place bets for the game, But suddenly for 10 minutes or 15 minutes the
45:56
platform is out. So you will not be a happy person. Yeah, I understand this company to when we were talking about earlier helps with the other side of the house as well. So coaches and and the way they do practices and for the scouts and giving them that real time information actually impact their decisions, right? What does what or who they choose or so
46:21
absolutely, they really interesting company kind of at the intersection. I think it actually says here is the sports intelligence from the, from the bedding side really. So, so how how would that young player field who's in the field playing his basket, his baseball pitch and there's a scout on the, on the, you know, on the benches and his he makes a good shot.
46:49
But the data isn't available for the scout and something goes wrong with the future of that player. That's real time, real time information again, right there, they're they're expecting expecting that up to date and and and being accurate and so especially if you didn't know there was downtime on the other side,
47:10
you just have the old stale information, you may make the wrong decision depending on which side of the which side of the house you are here, yep. Alright, so moving on, I think so this was sport radar and they use active cluster as well and they use active cluster for their business continuity needs and they're happy It works. And again, active cluster provides zero RPORTO. So zero Downtime
47:37
for and Transparent Failure Work. Moving away from sports betting, let's talk about life hanging in the balance of faith. That's right. So we have maybe we. Mm hmm. Excuse me. So they are a company since 1985 and they specialize in insurance.
47:58
They specialize in life insurance savings, provident insurance, all kinds of insurance and they have somewhere around half a million and use their customers. They manage More than 0.5 million contracts And somewhere around €700 million euros on their business. Obviously when life is in the balance, when there are claims that are going to come in off
48:23
this, half a million people, The platform that you used to provide services has to be 100% available. And for something that crucial. Maybe if we relies on active cluster. So that makes me wonder if my insurance provider also uses actual question or not.
48:47
Yeah, this is the they help with the sale as well. So if you know those, those brokers or whatever, you can't get access to that information and make that sale for someone that's trying to get those policies coverage is important and timing of coverage is important. I'm sure the other side about claims and all that that that could be very disruptive
49:06
especially with numbers like that. Like I said over half a million contracts. 600,006 hundreds of thousands of contracts. Half a million customers you know not too shy either hundreds of millions of dollars at stake. Right. That could be crucial. Again our measured right hours and minutes and all that.
49:23
You can you can actually measure that impact If you didn't have that 100% availability that this I. T. Provides to these companies. Absolutely and that's why active cluster providing high availability for such crucial businesses is what makes me I feel happy that we are solving real world problems for
49:47
our real customers. Yeah with that let's talk about our last case study which is the S. I. L. M. E. I. T. Department for Ireland monarchical. So this island is a self governed island. I think it's in europe. Right. Yes. Mediterranean I think is where it's it's
50:11
a good place to make have vacation. Yeah and I think we need to go see this one personally. Yes we need to check it out. It's working as expected. So they use compare shortage and they use look last term for their accounting systems. So they have a bunch of databases because
50:36
obviously it's a tourist destination but it also has local residents. Their municipalities and their whole taxation system and things like that. And their accounting system included within their databases run on pure storage and for the high availability of that they use active cluster. So again when I think about the impact, how angry or upset would you
51:04
be if your county's website is down when you're trying to submit a bill? Well maybe that time you'll be happy that you don't have to submit double. But if you really need to get approval for new plans for your house, if you need any approval for changes in the architecture, any permits that you file, you need the websites of your county working.
51:30
But it's simply payroll right? That's probably impacts with the people that work for the municipalities. Right. If I can't get that thing that I expect because of some downtime. That's a lot of angry internal customers for them. So absolutely. I don't know why that scared me. It all comes down to I'd imagine.
51:48
I mean accounting system the first thing I think that you know that might be a smaller bits. Maybe not the whole island but still it's an island. Probably a community. Yeah. I mean that could be a serious impact delay somebody doing it longer hours because you get there to get their paycheck or you know to your point is well paid in the end customers as well.
52:09
So they rely on active cluster for her availability as well. And they've been using it for some time. So I'm happy that we are solving a lot of different kinds of use cases, a lot of different problems for a lot of critical requirements for our customers but I think but I think it gives you a pretty good example though to kind of sum it up what we
52:34
mean by that critical applications and then providing a service that is actually just a cost effective, right? It's free, right as we mentioned, but cost effective to to put in to provide those services because the impact of those would be great if he didn't have that high availability that zero downtime that automatic fail over. So just hopefully it shows a pretty good
52:56
example of types of use cases across different businesses that that fit in that bucket, you know, as as somebody we're watching here goes through it on their own organization, they might classify their systems. Absolutely. I completely agree that these are the applications that you can see how much of a high impact they will have if they are not available. And these are the applications that require
53:21
business community and they can get it with active cluster that's right out of the box. Alright Zane. So our last slide is about resources. So for the viewers who are interested to learn more, I have put together a comprehensive list of resources for for you to go and look at so more detailed solution, our view some documentation.
53:46
Also, there are some explanation, explanation via light board which is similar to whiteboard videos on youtube, a live demo of how to put it in place, how to configure it, Everything is in this slide in these links and if you want to test and try the feature yourself, we have something called Test Drive which is available on pure storage website so you can just type in Test Drive pure storage and it will show up,
54:16
you can sign up and try this feature by yourself with that. I would like to thank Zane for joining us to explain the importance of business continuity to help me essentially go through the presentation and be a good partner in this whole conversation. Yeah, no, but glad to be here very, very proud to be part of this and uh subject as you
54:43
know, I like this. We'd like to talk about a lot on top of all these. If you get a hold of me as well, I'm easy to find as the allen and pierce storage. Happy to help with any questions, awesome and thank you everyone for joining us for the session today. I hope this was useful and I hope if you have any business content ID requirements and you
55:06
are interested please learn more and explore active cluster. Thank you. Thank you everyone
