Storage Unpacked Podcast - Storage Unpacked 258 – Introducing Infinidat G4, InfuzeOS 8 and InfiniSafe ACP (Sponsored)

Episode Date: May 22, 2024

In this episode, Chris talks to Infinidat CMO, Eric Herzog. Infinidat has announced one of the biggest upgrades in eight years, with the release of InfiniBox and InfiniBox SSA G4, the fourth generation of enterprise-class storage.

This is Storage Unpacked. Subscribe at StorageUnpacked.com. This is Chris Evans and I'm here with Eric Herzog from Infinidat. Eric, how are you doing? Great. Always great to be with Architecting IT. We love talking to you, Chris, and thanks again for inviting us. And the sad thing is, which we have this discussion every time, is you've got an amazing shirt on with a lovely badge on there too, and nobody's ever going to see that. That's true, but you know, I had to dress. Yeah, absolutely. Yeah, very good point.

Right. So we're here today because you've made a big announcement of a lot of different things. Some are hardware related, some are software related, some are cybersecurity related. And we're going to dig down into what all of those different announcements are, and you're going to talk through them. But just briefly, could you tell us what the four different announcements are that you've made today? Yes. So we've got a brand new platform, our G4, both available as all-flash and as a hybrid, built brand new from the ground up. So that's our system level announcement. And then everything else is software.

So our Infiniiverse software is being extended. Our InfuseOS Cloud Edition, which has been on Amazon, will now also be on Microsoft Azure. We are adding InfiniSafe automated cyber protection that will help you reduce the threat windows. And we're announcing that our InfiniSafe cyber detection is now available for VMware environments. Excellent, so lots to talk about then. Let's dive into the hardware and get started there. So you're calling this a G4 architecture, the new hardware, why

don't you take us through exactly what you've announced? Sure, so what we've done is we've announced a new platform, Ground Up. There'll still be a hybrid version as well as an all-flash version available. We've switched to the AMD family, so that will give us a better number of cores. In fact, we've increased the core count in a single socket by 31%, which of course helps us get the performance we're going to talk about in a second. We also now are PCIe 5 and also DDR 5. So what that results in from an end user perspective is for IOPS and bandwidth, we're up to double the performance, up to double. Now as you know, Chris, our latency already has been 35 microseconds and no one's even close to that. But we've now, with the new generation, the G4, compared to the third generation platform,

up to two times faster on your bandwidth and on your IOP. So real advantage from an end user perspective. Okay, excellent. And it's quite interesting that you've gone down the route of AMD. There are lots of choices out there nowadays, of course. And I think that's a good thing because, I mean, you don't necessarily have to be tied to the one vendor's release of products. Intel have their release, Cycle AMD have theirs. And I think it gives you that choice to decide which direction to go to in terms of, as you said, DDR5, PCIe5,

all that sort of stuff in the architecture. But in reality, this is more about exploiting that capability within software, I think. Absolutely. So right now we're at 2x the performance. We think with additional performance tuning, which we do every time we do a new software release, we always do performance tuning. We know we can get it to 2.5x. We think we might even be able to get to 3 to 3.5x just by tweaking our software. So that really gives the end users advantage.

And of course, don't forget, everybody who's on sport and maintenance, all our software upgrades for free. So if someone bought a G4 next week, in six months when there's a new software release, they just upgrade the software. It's free. It's non-disruptive upgrade, by the way. And then the performance gains and any other new features we've added in the next version, they'll be able to do we are moving from version seven of our software the infuse os our operating system to now version eight and version eight which is now on amd versus version seven which had been on the uh intel ship sets we've made it so that you can communicate between an older generation and a newer generation so let's take an example chris we have a customer who's already placed an order and they want to quote, sweat an asset. So they have a third generation box. And what they wanna do is turn that into their disaster recovery target

and then have the G4s on-prem. So we can replicate from the G4 to the third generation older InfiniBox that they're gonna keep. So we wanna give customers ultimate flexibility. And even though we switch chip platforms, we're able to do that with our InfuseOS version 8.0. Okay, excellent.

So let's just talk about the form factor. And one of the things that would strike me moving to a slightly different processor architecture is you could be gaining additional, let's say, sustainability benefits in terms of power and cooling, reducing the load. And that's a very positive thing at the moment as customers are sort of looking at how much it's costing them to run things in their data center. Has there been

any sort of change in the form factor of these products? Are you doing anything different with them? So we've made a couple of changes. One thing we've done on the hybrid in our 4000 family, which will now be the 4400 family, we've added a 20 terabyte hard drive. So that's our hybrid that allows you to put more applications and workloads on the same array, which of course leads to consolidation of your older generations, which of course not only gives you better OpEx and CapEx because you now have less physical hardware on your floor, but also of course it reduces your power and cooling as well. Now in the all flash, what we've done, we now have four versions of that. So we had historically 100% fully populated. Then last year for all flash,

we introduced it as 60% partially populated and an 80% partially populated. That of course, not only reduced the buy-in price compared to the 100% fully populated, but also you can add the additional storage with a complete non-disruptive upgrade. So your SAP, your Oracle workload, whatever you're doing continues to run while you add new capacity. So we'll continue with the partially populated 60

in the G4, the 80 in the G4 and the 100 in the G4. But then we did something very unique. As you know, most of our customers are in the global Fortune 1000. High-end, we're the high performance guys, high availability, 100% availability is guaranteed. And what we've done is figure a way how to take such a high-end solution

with all its feature functions, including everything we've got in cyber, which I know we're gonna talk about later today, and shrink it into a 14 RACU form factor. So don't think of this as a mid tier array. It is a high end array with all the feature function, no compromises on what we deliver, all the same guarantees, our cyber guarantee, our performance guarantees, our availability guarantees, all in writing, so guaranteed SLAs for the customers, but a 14 RACU form factor. It also, by the way, happens to have less usable capacity than even the 60% partially populated, so that A, allows us to lower the price point. B, this is installable in any industry standard RAC. You can get the RAC from us, but you can use your own RAC. So if

you've got a full data center, you've got some partially populated racks. If you've got 14 rack, you can fit it in there with your servers and networking gear, whatever you're running in that rack. But it also gives our customers two new options. First, co-location. Some companies, for example, of course, we're again, mostly in the global fortune 1000 type accounts. and so those guys might have a core data center but maybe they're expanding in a country or opening up a new factory or something so they might use a co-location facility and as you know co-location they charge for your floor space so now instead of having to bring in one of our racks even if it's partially popular stuff to pay

for the floor space if they've got a rack that they're using in the co-location, you can add it in there. So it gives you that co-location capability. And then also we would call it the edge data center. So let's take an example. Let's assume architecting IT was building something. Since I know you love to golf, you are the biggest golf club manufacturer on this planet. So you have factories all over, right? Building golf clubs. And let's assume that in all those factories, you have what I'll call a small or miniature data center. You have a core data center at corporate, but you also have little data centers. I would call that an edge data center. And so that's another workload we can fit in there. Again, when we had our, you know,

always coming in a full rack, even if it was partially popular, but always getting the 42 rack you, what if it didn't fit in that edge data center so now you've got traditional data center but a small form factor b you can use it in your own rack you don't need to use ours you can do it when you're co-locating which of course keeps that floor space cost down and if you do have edge data centers then you could use that so that's what the 14 rack you delivers. It will be available as a G4 all flash. It will not be available as a hybrid. And that will be available obviously this week.

Brilliant. So I'm just going to sort of dig into that a little bit for people who might not have the history of your technology. I think, cause I think this is quite an important step forward in terms of the way that you've moved the company on, you know, so looking back historically, and I'll link to all the things that we've talked about over previous podcasts and releases and people can go back and have a look at that but generally you know your standard was to deliver as you said into enterprise and to deliver a single solution which was a 42u rack that you'd effectively roll in which is going to be rolled into a slot in a data center that you've

you've made available not everybody can do that that, as you said, from colo perspectives, and not everybody wants to either dedicate a rack or necessarily use the rack that comes from a vendor because obviously there are scenarios around power tracking, all the sort of things that you'd want to do in your own rack, and there's very good reasons why you would choose your own rack rather than somebody else's. This, to me, I think is quite a step away from what you've done in the past in the sense that

you've now given people the ability to deploy this technology elsewhere. And I'm only re-emphasizing that because it is such a difference from what you've done previously. And it does sort of widen that market for the sort of use cases that people could use your technology for. Absolutely. And we had that feedback from customers, which is we love the technology, but sometimes, A, in something like a Colo or an Edge data center, we can't put your whole rack in there when our big data center is sure. And also a few people said, look, I have unused rack space in my big giant data center, right? It's partially populated in my rack. So could I just take your stuff and stick it in there? So that's what we did. We basically shrunk our high-end array with all its software

feature function, the Infuse OS and everything we deliver in the full, our own rack solution, now into 14 RACU. Okay. So that's brilliant. So that's the hardware side. Let's just, from that perspective, remind everybody, what's the sort of availability date from the announcement today? So that is available today. Now, there is one other thing we've done with the G4 family, and we have rolled out a controller upgrade program. Obviously, it's from the G4 to whatever comes next. Let's assume it's the G5 or whatever we call it. It is optional. So in the hybrids and in our all flash, you can get what we call Infiniverse Mobius. It allows you to just upgrade the controllers, which means you would still keep,

if you got the hybrid, you would still keep the hard drives. And by the way, our hybrids actually have DRAM, flash, but mostly hard drives. And in the all flash, of course, it's DRAM and flash. You would keep all of the flash and just put in a controller. So you get to keep the media if you'd like to.

It's optional. You don't have to do it. Just as this program is available by many other vendors in the mid-range, we're the only guy who will have it in the high end. There is a fee for it, which everybody else does in the mid-range, but you don't have to buy it if you don't want to. And some people really want everything new, right? Remember in the high end, Chris, as you know, you know, the life cycle of these technologies is five to six

years between major generational shifts. So you may say, oh, you know what, maybe I should just buy the whole thing new. But we don't care. It's up to them. They can do whatever they want, their choice. So Infiniiverse Mobius is another element of our system announcement. That's something we did not have in the past either. I think the ability to do in-place upgrades is quite important nowadays. You know, go back and look at what we used to do 20 years ago, and I don't want to necessarily dwell on the past,

but you look at what we used to do, and very much it would be a case of putting additional hardware next to the previous hardware and then doing a massive migration task. And, you know, those sort of things take forever. I guess that was probably quite logical at the time, because a lot of the architectures

were based on hard drives. And as a result, you weren't going to see a significant gain by changing the controller at the front end, because the full back end disk system was probably the part of the infrastructure that was more likely to be the blocking part of that design. I think when we move to flash, because flash is so quick, and especially when you're using DRAM to cache that and other things, you can definitely get

a performance boost simply by changing the controllers. So giving people that option, I think is quite important in the modern design of storage systems. Yeah, and if someone wants to buy a brand new one, so when the G4 is old and the G5 is a new one, that's okay, too, because one of the things we do provide, in fact, we had a call with a Fortune 50 just the other day, and they are new. They just decided to buy from us, but they were talking about some stuff, and we said, well, you don't have to worry. We mentioned the controller upgrade program. We said, but if you'd rather just get the next generation, no problem, because we have a product called Atom, which is automatic data migration in the background on the fly. It always favors your applications and workloads. And our senior director of professional services was on and was talking about a Fortune 500 financial institution.

And they had done a migration about six months ago, right, from to the third generation. And they had done a migration in one day, all back end, no downtime. So we do offer that if someone wants to do it the old way, as you said, and do a migration. So we give them both options. If they want to just do the controller, not a problem. They want to do it the traditional way, that's fine too. And we do have a technology that will do it automatically. Just set it up and it'll do its thing. And in this case, it was a day. It could take two or three days because we always favor the application. As you know, Chris,

we're all about a hundred percent availability and uptime. And so even during a migration, you don't want to take anything down. And by the way, same thing when you do Infiniiverse Mobius, you're upgrading the controllers, you're doing it live. So the systems are still going to be running. You're not going to take it offline. You don't have to worry about that SAP workload or that Mongo workload, that NoSQL workload. It's going to still work. Now it may slow down a little bit while we're doing the controller upgrade, but it still will make applications and workloads available while the controller upgrade is happening. So we want to give them ultimate flexibility. And of course,

our customers, because of who they are, really demand 100% availability, whether you're adding new capacity and are partially populated, whether you're going to do the controller upgrade with Infiniiverse Mobius, or whether you're going to buy a new one and use our Atom going to do the controller upgrade with infiniverse mobius or whether you're going to buy a new one and use our atom product to do the migration they never want to go down so we've made sure how whichever way you want to do it you don't go down yeah these things are table stakes now eric aren't they really that's the way the best way to look at it okay um right so when you talked about mobius there you mentioned infiniverse i just wonder whether it's worth you just touching on exactly what infiniverse is so people can be clear on that one so what we've done is our infiniverse platform which we've had already gathers incredible amounts of telemetry data and we're expanding that so

we're going to make that a single control plane an infrastructure consumption services model so for example architecting it has bought two infininiBoxes, one SSA, one regular. You now have two G4s. And you're going to set up a bunch of snapshots. Some will be read-write and some are going to be immutable snapshots for our cyber resilient InfiniSafe. So today you would go into InfiniVerse and you could see it all. But with this new generation, what will happen is we can automatically generate reports and send them to you, do whatever you want us to do with the reports. So the idea here is to simplify IT operations, make it easier, make it proactive.

So you could say, you know what, I'd like to get a report every two weeks. So every two weeks you will get a report on every snapshot you have. If you set up new snapshots, you'll see that right in the report. You'll know which ones are read right, which ones are immutable, and you also will see all your schedules. And if you change a schedule, so if you go from every six hours to every four, that will all be in the report. But instead of going into Infiniiverse, which is how you do it today, Infiniiverse, if you

will, will bring it to you, which again, makes it easier for you, cuts the IT operations. And as you've talked about in other podcasts, right now there's an IT skills gap. So the easier you can make everything, not that, you know, obviously, as you know, we have several public references that talk about how they've had an InfiniBox or or InfiniiBox SSA third generation, and they haven't touched it for two or three years, truly automated, which reduces, of course, IT operations expense. And clearly, if you've got an IT skills gap, if the thing runs on its own. Now with Infiniiverse, this new infrastructure consumption services platform, those kind of reports will be proactive. So you won't have to have the storage admin go in. And you can see an incredible amount of telemetry data, whether it's capacity,

whether it's what we've done with snapshots, what you're doing with replication, what you're doing with performance. We gather all that, but right now you have to launch Infiniiverse and then you look at these incredible screens. So now it's all moving to proactive. And as you know, some of,

and I know from your past, some customers are very conservative. So if they don't want any of the reports to go out, then you turn it off, right? So you can either do it the traditional way, go look at all that great telemetry data in Infineverse, or again, this new proactive approach is designed to simplify IT operations. If there's a skills gap at that company, hey, automated reports are coming to them. And so that's the idea is to help them with the skills gap and just overall make it just more efficient. Lovely. Okay. So I think this for me is your sort of fleet management, your platform

management solution. And ultimately these are, or this is another, what I would class as table stakes nowadays. This is the sort of thing where really, if you're not giving the customer the ability to manage their environment, and let's face it, you have a lot of customers who have got a lot of hardware sitting on the floor, so they need this ability. These sort of things are just, I think,

just critical to the way that you manage your infrastructure going forward. So it's good to hear that, you know, that's gonna be expanded a bit further. Yeah, I mean, we're going from an incredible amount of data that's clearly and easily displayed but still you have to launch infiniverse and then you can see everything you want to see and the amount of telemetry data we gather is incredible our customers use it as you know chris we provide technical advisors whose job is really about application and workload optimization they're not tech support they're free

some of my competitors offer technical advisors, but they charge. We provide every customer with free technical advisors. But also remember, we only do level three tech support. As you know, the tradition is level one, level two, then you go to level three. Because of who our customers are, they go right to level three.

So these are not people right out of university, right, with their computer science degree or electrical engineering. Nothing wrong with that, but they're inexperienced our level three tech support guys have an average of 10 years of experience doing support so they use the infiniverse right now as well so it'll make it easy for everyone our tas our customers remember we also do 80 of our business through the channel and of course our channel partners like to charge for services managing. So this way, reports could go to them and they could call up and say, you know, Chris, did you know, blah, blah, blah, that, you know, in your factory making the golf clubs XYZ, you don't seem to have any immutable snapshots. They're all read right. Would you like

me to set some up, you know, with cyber resilience? And so that's something that partners like to see because allows them to proactively then approach their end users as well so everybody benefits from this expansion of infiniverse excellent okay so that's that's uh infiniverse let's move on and talk about your operating system update so infuse os now the last time we talked i think might have been the last time or maybe the time before i think you just started to release uh infuse os into aws and that was a sort of a first step of that and i that's there's now an update on that isn't there yes so what we've done is we had aws which according to the guys who track the market numbers is the largest public cloud but microsoft is right there too not as big but they're close So we've now done is gone from not only having AWS,

but now having Microsoft Azure. And of course, because our customers are so large, they may be using both of those clouds in a hybrid cloud configuration. So now you can have hybrid multi cloud, the Infuse OS, which is the operating system on every InfiniBox and every InfiniBox SSA is a full version sitting last year in AWS, now a full version sitting in Microsoft Azure. So the use case is pretty straightforward. Replication for disaster recovery and business continuity. While you're physically going to Microsoft or to Amazon, you're actually going from an InfiniBox to InfuseOS sitting in the cloud. So it's going literally from an InfiniBox to an InfiniBox. Okay.

Second is backup. As you know, we partner with all the major backup vendors. We have about 150 customers who use the InfiniBox as a backup target. As you know, and we talked about this in other podcasts, we do have a purpose-built backup appliance called the InfiniGuard. The difference is in the InfiniGuard. The difference is in the InfiniGuard side, the data reduction is done by us. When you use an InfiniBox, the data reduction is done by Commvault, Veeam, Veritas, IBM Protect, the software vendors do it on the server side. So we do it both ways. And if the customer wants to do it on the server

side, then today they would back up to an InfiniBox. And like I said, we have over 150 customers that do it that way. Now you could go not only to Amazon for your backup, but you could go to Microsoft Azure. And when you do that, you think you're backing up, you know, your backup software, Veeam for example, if you had an InfiniBox in data center A and in data center B and C, you want

to go to the cloud, one to Amazon and one to Microsoft, you'd point it there and it would think it was going to an on-premise InfiniBox. That's what it would look like. So that makes it easier. You don't have to learn anything. Obviously Microsoft and Amazon have good solutions, but they have their own user interface. So if you're going to go to them today before this announcement on the Microsoft side, you'd have to use the Microsoft storage user interface. In this case, our customers who, as you know, our average customer has over 16 petabytes installed.

So do they want to learn all that stuff? No. This way, the Infuse OS is sitting atrosoft or at amazon or i think with many of our customers possibly in both and they could back up to it they could do disaster recovery to it we do think burst capacity if they need burst capacity obviously a lot of our customers are very high performance and as great as clouds can be cloud storage as you know is usually not very performant right and if you performance, they charge you for it, right? You got to pay for it. So it allows you, we think, burst capacity. And obviously people do do test

dev and proof of concept. So this case, if they're doing development of this new Oracle workload, for sake of argument, they can do it and do it on the InfiniBox, right? They've got an InfiniBox SSA on-prem for this high-velocity Oracle. Now they're adding a new application that they're going to put on that SSA. They can do the testing and development on an InfiniBox in the cloud. So there's no issues of, hey, the software worked fine when we were testing, but then when we put it on the InfiniBox, it didn't work right, or when we put it on the InfiniBox SSA.

So that's another use case. Those are four key use cases for the cloud. And last year, you could do it with Amazon. Now, this year, you can do it with Amazon and Microsoft Azure. You pick whichever you want. And we do think big enterprises probably may use both. And this gives them that flexibility to use either. So I entirely agree. I think having spread across the different cloud vendors is very important because people have enterprise licenses and they have other reasons why they might want to use one platform over another. And not only that, but as you said, they probably could well be using both. And there's a degree of resiliency there if that's really something you want to consider.

But I think from the other side of this, I think, which is probably more important from my perspective, people might say, well, why would I be backing up, you know, something like Veeam into your platform first before placing it on the cloud? And I think it's all about data portability and the ability to look at it and say, well, if I'm using, you know, your snapshot technology, your data movement technology, and I want to move that data from Azure to AWS, maybe you can do that, you know, move it from one platform.

We can do that. There you go. This gives you that platform independence piece, you know, so somebody says, well, we've got this running in AWS in this region and then somebody else comes on board and says, well, we need to be able to do that on Azure because we've taken on, you know,

this new business we've just acquired and they're all Azure. Now we need to be able to implement this technology across that platform. Data independence becomes very important at that point. So that, for me, is the sort of reason I would sort of use that technology, to be honest. So we see this as an important additional step. We already had Amazon. Now we've got Azure.

So those are the two biggest clouds by far. You know, there's Google Cloud and IBM Cloud, but they're way smaller than Amazon and Microsoft. So for most of the customers, we're set. And remember, it depends on the geography too. As you know, in the United States, AWS is probably bigger. And in Europe, in the enterprise side, Azure is bigger. Now, both are in both locations. But from an enterprise perspective, Amazon is bigger. Now both are in both locations, but from an enterprise perspective, Amazon is bigger in North America and Microsoft is bigger in Europe. So again, a lot of our customers are global, right? So we got to be able to support both. What if they're

doing Amazon, let's say they're doing backup with Commvault in the US to Amazon, no problem. They're backing up to an AffiniBox. What about in their European operations? They're using Azure. It's the number one cloud in Europe. So they're using Microsoft and they want to replicate. So you don't have to do the same thing, right? It is basically replicating, backing up burst capacity to an InfiniBox instead of doing it to the native user interface of either Amazon or Microsoft. So it simplifies everything. It's seamless because it's essentially like going from one InfiniBox to an InfiniBox SSA on-prem. It's exactly the same way you orchestrate and do things.

It's the same. It just happens to be sitting in the cloud instead of in your data center. Yeah, perfect. So that's great. That's good news. That's good to hear. Now we're going to talk about InfiniSafe and ACP, as I think you call it, automated cyber protection. I'm just going to start us off by podcasts people can go back and listen to. Having said that, I'm going to obviously ask you to briefly summarize what you do today, and then we can dive into what the new stuff is. So

off you go. Great. So what we do today is InfiniSafe has two components. Base InfiniSafe, which is free, it's part of the InfuseOS operating system, no charge. It works on the InfiniiBox, the InfiniiBox SSA, and on the InfiniiGuard. It does four things. A, immutable snapshots. And you can schedule those however you want to do it. We have a master scheduler. You figure what you want to do. Second thing is we do logical error gapping, locally, remote, or both. Third thing, as you know, it's not if you'll be attacked, it's when and how often. So if one of those attacks is successful, you need to recover and you need to recover quickly. So but you have to have a known good copy. Remember, the cyber mafia is not out there pounding their chest like King Kong or screaming like Godzilla

in some science fiction movie, right? It's all done surreptitiously. So you could be making snapshots of malware or ransomware on your Infinidat storage, or by the way, on anybody else's storage. So you want to get to what's known in the industry as a known good copy. So we create a fenced forensic environment, and that way you can figure out whether it's good or bad. And if it's good, you recover. We then do near instantaneous recovery, which is guaranteed. So on our cyber guarantees, we guarantee two things in writing. First of all, guaranteed SLA that the snapshot is indeed immutable. Second thing we guarantee is the recovery time objective, the RTO.

On the InfiniGuard, we guarantee 20 minutes or less. And just to give you an example, we now do cyber webinars every quarter because it's such a big deal. And because I'm a crazy old man, you've known me for years and years, we don't do any demo that isn't done live. So we demoed live last time, a Commvault 20 petabyte, 20 petabyte backup repository. We recovered that in under 13 minutes. The quarter before at the end of 2023, we did the same and we recovered a 20 petabyte Veeam backup repository in 11 minutes and 55 seconds. But we guarantee an under 20 and nobody's doing that. Then on the primary storage side, we guarantee regardless of the size of the data set, we will recover that snapshot in one minute or less. And we did that demo, of course, both in Q1 of this year and in Q4 of last year, 200 terabytes, 175,000 files,

200 terabytes in three seconds. That could have been four petabytes. We guarantee no matter what in under a minute, and no one's doing that as well. Then the second thing, which we're going to talk about a little bit today too, is InfiniiSafe cyber detection. And that's the capability of using AI and machine learning

to identify malware or ransomware. You pointed at the snapshot, you could point it at a database, you could point it at a file system, and then it will scan it. And that allows you to use the AL machine learning. And then there's two use cases. Clearly recovery. You've got to get a known good copy. So you want to make sure what you're recovering is

good. The second thing is it could be used as an early warning system. You could do the cyber detection once a day and if you see something anomalous through the API integration, you can send a note to a security operating center if they have one, or you could send it to some global cybersecurity monitoring package that InfiniSafe Cyber Detection saw something. Cyber Detection is a charge. It is a storage software as a service, but it's only what you scan. So for architecting IT, the golf company bought, of course, the 17.1 petabyte hybrid for your factories.

If you're scanning 200 terabytes, you only pay for 200 terabytes. So that's the InfiniSafe and InfiniSafe server detection before today's announcement. Okay. Wow. There was an awful lot in there. And I definitely recommend that people go back and look at that because this area is becoming, I think, pretty critical as a differentiating factor of how vendors are working in the sense that the first stage was definitely to allow you to see what was going on and to unmute those snapshots so that they could be protected and not destroyed. But in reality, we now need to go to the next step, which is we need to try and automate this. And I think that's

what you're going to talk about as the new set of features. Yes. So with automated cyber protection, ACP, which is free, by the way, no charge, it allows you to integrate with cybersecurity software threat packages, Microsoft Sentinel, IBM QRadar, Splunk. So what you can do is those packages are looking for threats. If they see a threat, they send out a warning. It'll come through the API and it will automatically kick off a snap on InfiniBoxes or InfiniBox SSAs that you so deem. So why would that matter? So let's take the perfect example, architecting IT, the giant global golf club manufacturer. You're using Infinisafe today and you're doing snapshots

every six hours. Okay. So today I'm coming to you live on the 22nd at 4.38 AM in the morning. Okay. And the snapshot just finished. Okay. So I'm the guy running your factory in California. Now it's 5.38 an hour from now, and you are using Microsoft Sentinel, and it detects a threat. Even though we just did a snapshot an hour ago, we'll automatically kick off another one. How does that help you? Most of our customers, when on their snaps, whether they're read, write, or mute, they're on a schedule. Every four hours, every six hours, once a day, and it varies by the different workloads. You could have certain snaps on certain workloads that you do every six hours, and then

other snaps that you do once a day. But in this case, we will snap that InfiniBox SSA. And then if you want to use, you can use InfiniSafe cyber detection to scan scan it and it helps you reduce what in the cyber security world is known as a threat window. If you have that set up to replicate every six hours, like I said, well, the threat was detected by QRadar or Sentinel and it was detected one hour after you've done your schedule. So if you stay to the schedule, you're not going to do a snapshot for another five hours, which makes no sense because there's a threat detection that happened by other cyber security software which you purchased so in this case our storage is working with the cyber security software that you've bought and then automatically

takes some proactive action you can do it with our reference architecture there's basically a recipe of how to connect these things and then you just set up what you want and if you want to do scanning then of that immutable snap you would implement cyber detection capability with infinisafe and do that and if you don't want to do that that's fine but you've got an immutable snap and you can look at that in other ways you don't have to use infinisafe cyber detection but the point is helping reduce the threat window and that's an important thing. As you know, the average enterprise will suffer over 1,200 attacks a week. And the cost to enterprises of cyber, a large healthcare company recently, they had to file something publicly in the US. The cost of their cyber attack between the actual ransom they paid and everything else, 1.8 billion US dollars. That's just one instance, one instance that happened at the end of Q1 of this year. So doing this, and you've already bought the cybersecurity

software, let our InfiniiSafe take advantage and leverage what that sees. And we can do that now with ACP. Yeah, brilliant. So I tell you, there's a number of things here. So I think just sort of breaking it down into, I think, a number of different areas. First of all, I think storage is becoming, let's call it the last resort, the backup or the protection copy of last resort. So you're trying to make sure that you try and mitigate all of the issues before they even get to that point, but you still need some sort of last resort thing.

And rather than saying, well, we've got backups that are sitting on tape or we've got stuff that's sitting over there and having to do mass restores, you want the storage solution to be integrated into that process because it's quicker and it reflects what you just said, Eric, about the speed at which you could recover things like Veeam recovery and the integrated into that process because it's quicker and it reflects what you just said, Eric, about the speed at which you could recover things like Veeam recovery and the files and all this stuff you talked about a moment ago. I think the second thing is that dynamic nature. So if you

rely on a snapshot that you took six hours ago, well, you're six hours out of date if that snapshot was valid still. So that's really quite important but of course these threat actors are really clever so they may well have got into your system and looked at your backup environment and said do you know what i think they're backing up every six hours so let's go and apply the attack that we apply literally two minutes after the last backup was done so you know they're not they're not stupid they know how to sort of make the the impact of their of what they do really uh dangerous. So I think that dynamic nature is the really important piece there because it's taking away their ability to

try and attack you in the worst possible way. And I think third is, of course, nobody can monitor and manage manually dozens of systems and be able to see this sort of scenario going. Sometimes other systems detect the issue and other systems see the problem and think, right, okay, we need to do something now. It's not all about one part of the infrastructure working in isolation. It's bringing all of these things together to protect against. And I think that's where the APIs come in. So, you know, the way you've implemented that, this is actually a holistic approach rather than it just being one separate thing now. Well, we, you know, when you think about it,

when you talk to a CISO or you talk to a CIO, and I have been doing this, as you know, almost 40 years, I have never met a CIO who used to be a storage admin. They kind of hate storage. They know they need it. They know it's important, but boy, do they hate it. They're almost all software guys. The days when it might have been the mainframe vice president or the network VP, that passed about 25 years ago. They're mostly software guys. There's nothing wrong with that. So we focus on optimizing applications, workloads, and use cases.

When we talk about our performance, we don't usually talk about IOPS, bandwidth, or latency. We talk about Mr. CIO, that SAP workload, how long does it take? Oh, it's five hours, Mr. Herzog. Well, what if we could do that in 30 minutes? That's what gets them excited. In this case, on the cyber side, they usually don't think of storage. They think of the edge, of course, they think of the network, right? Cisco and other networking vendors have all kinds of security stuff. And of course, there's that whole wealth of cybersecurity software vendors like Microsoft, like IBM, like Splunk. There's, you know, Fortinet, right? And these guys specialize in doing data center-wide cybersecurity packages. So in this case, you're getting, as you said,

not only dynamic action from the storage side, but it's holistic approach. It basically means that by including storage, you have a comprehensive cybersecurity strategy for your data center. If you don't include storage, and quite honestly, most companies don't right now, you're leaving a wide gap. And in this case, we make it simpler because they've already got the cybersecurity software, most of these companies, right? It's a huge, huge business, the cybersecurity software threat detection business. So in this case, whatever they bought, we integrate. And by the way, if they do it themselves with a Syslog

and they've created a security operating center, doesn't matter, that will work too. So the idea here is the ultimate flexibility. And as you said, dynamic action from the storage, not just passive and waiting. And by the way, you can set up a schedule. You and I discussed this one year when I was there at a dinner. We have a customer who does immutable snapshots every 15

minutes. But that's not the norm, right? The norm is four hours, six hours, eight hours, right? You used to be the CIO of an end user. So you know how it works. So in this case, it's proactive. It's dynamic, as you said, and it's done between the integration of us with a cybersecurity software package that you probably own. So it allows you to get best advantage of us. And also that cybersecurity software, it's going to send a notification. If we pick it up, boom, we cut that snap. And if you want to go to the next level, we can do the integration so after that snap infinisafe cyber detection can scan it and you just configure and you can all do that with our reference

architectures a recipe book just like you're doing a cookbook it's the recipe just follow that and you figure and by the way it doesn't have to be the same for every infinibox or infinibox ssa you may want to do five of them and not do the other five you may want to do two you may want to do a couple hundred terabytes because you have that's your most valuable data and not do the other five you may want to do two you may want to do a couple hundred terabytes because you have that's your most valuable data and not do the rest you can do any of that you decide what you want to do and then once you've used our reference architecture it's now fully automated and fully integrated with sentinel or qradar or fortinet or splunk or any of the other big packages that you're going to use, right?

And it's ubiquitous these days. I can't think of an enterprise that didn't go buy some sort of cybersecurity software product, right? So we can integrate with them. Perfect. Okay. And there's one other, I think, one other new feature related to virtual machines for cyber detection.

Yes. So what we've done is on InfiniSafe cyber detection prior to this, you could do file systems, you could do files, you could do databases, you could do volumes, but anything that was in a VMware data store, we couldn't touch. So now not only can you do files and file systems and volumes and databases, but everything now that is in a VMware data store. Again, it is a billable model. It's the only part of InfiniSafe that we charge for cyber detection. And we only charge you for your scan.

So for sake of argument, since we're talking about your giant golf club manufacturing company that you own, we can say on this InfiniBox SSA, we've got a bunch of VMware. We want to do sub-detection on 100 terabytes of that. On another InfiniBox SSA, you could say we'd like to do 100 terabytes of VMware, but we'd also like to do this volume. That's 200 terabytes. So it doesn't have to be the same on every array. You can choose what you want.

You can change it. It's software as a service. So if you had 100 terabytes of VMware data store, and then you add more to that specific array, and now you want to scan 200, you just let us know. We just change the billing and it's done as a service. So it's easy to do. So you have ultimate flexibility. And again, VMware, it's here for a long time to be. And we have an excellent integration. You know, 95% of our customers have VMware installed right now. So, you know, not doing a VMware data store, it really opens up another avenue. And of course, they can be attacked too.

You know, VMware and all of the other, what I'll call the interstitial software layers, to me, containers, VMware, Hyper-V, they're all subject to attack too, right? I mean, they're going to attack whatever they can get. Backup data sets. So by the expansion, we've now added that. And the last thing we've done with cyber detection, today it's on the InfiniBox and the InfiniBox SSL Flash. At the end of the year, not now, but we're announcing it now, at the end of the year, we will have InfiniSafe cyber detection on our InfiniGuard purpose-built

backup appliance. Right now, you can't do it on our backup appliance, but you will at the end of the year. So that's kind of a minor thing because it's not available today. But everything else we talked about today from the G4, Infiniiverse, InfuseOS Hybrid Cloud, ACP with InfiniSafe, Cyber Detection with InfiniSafe now on VMware, that's available right now. You could place an order, talk to your local partner, call us up, go to www.infinidat.com, whatever you want to do. But we want to make sure that, again, this is the biggest launch in the last eight years of this company. And again, as we said at the beginning, we changed our system with the G4, but the bulk of what we do, the magic of Infinidat is our incredible software. When you look at our patents, which

is over a hundred, we have a couple for hardware design, but the bulk of our patents are all about software. Infinisafe, InfuseOS, NeuralCache. That's where our patent value is. And that's why we make sure that not only did we refresh the platform, but we've added all this additional software value, which is what we're known for. People buy us for our software and our ability to, as you know, to do white glove support and service, which we talked about. So, okay, great, Eric. Thanks. That was really useful to understand the stuff about the virtual machines, because that's going to be, I think, quite an interesting one for people. So let's just summarize what we've talked about. We talked about G4, we talked about Mobius, we've talked about new InfuseOS for Azure, and we've talked about the ACP stuff. So just remind us, when will this stuff all be available? And when can customers go online and order it after the announcement day to day? So everything is available today with the exception of cyber detection for Infinigard,

which will be at the end of the year. Also, we're going to do a series of public webinars in June. We'll be doing two in the beginning of the month, one on European time and one on America's time. You can sign up at www.infinidat.com. At the very end of June, we're also going to do a LinkedIn live. So you can sign up for that as well. And we'll be doing a roadshow. So you and I will have to grab a lunch or a dinner. We'll be doing a roadshow in Europe. We'll be doing Madrid, Milan, and we'll be doing also London, and we'll be doing Tokyo. We'll also be doing in the US, we'll be doing San Francisco. West Coast, we'll be doing the

Southeast. We'll be doing the North Central area, and we'll be doing New York. We haven't picked all the cities yet, but that will be in September and October, and you'll be able to sign up for those. We'll be doing them live. In the webinars and this, which obviously we can't really do in a podcast, we will be doing some live demos because I'm a crazy man. We will demo the cyber detection for VMware because VMware is so ubiquitous. It's here for a long time. So we'll demo that. And we will also do a demo of InfuseOS Cloud Edition going to an Azure configuration versus an AWS configuration.

So again, www.infinidat.com, you can get all of that. And again, thank you very much for all your time. As always, I've been doing these things with you for years and years, not just at Infinidat, but my older companies as well. And it's always a great pleasure working with you. And yeah, we get great feedback from both the end user community and our channel partners about these podcasts that I've done with you over the years. So again, I want to thank you very much at Architecting IT for all the stuff you do.

And you do bring unique twins because you're not an ex-vendor. You're not a press guy. You're not like me from Infinidat. I was, as you know, CMO of IBM Storage, a senior VP at EMC. So I'm the vendor side, but you're the real end user. You are a former CIO now doing this. So you really represent the end user side. And that is unique in

the analyst community, as you know, they're mostly ex-vendors. So we really appreciate that you can see things in a different light because you used to be the customer. So I would say that if we had the cameras and I'd show you all the scars I have from all of the work I've done over the years, but as we haven't, then I'll save people that. Eric, thank you very much for your time again. It's been really good to catch up with you. I will make sure that we get all these links put into the show notes and look forward to catching up on the next release or when you're next in London. Great. Thank you very much.