Nutanix Weekly: AOS 7.0 Launch Blog

We are excited to introduce the next major AOS release for the Nutanix Cloud Platform solution, Version 7.0. This release is designed to help customers enhance fault tolerance, simplify automation, strengthen security, and improve networking across their Nutanix infrastructure.

Blog link: https://www.nutanix.com/blog/aos-7-0-launch-blog

Host: Phil Sellers, XenTegra

Co-Host: Chris Calhoun, XenTegra

Co-Host: Jirah Cox, Nutanix

Co-Host: Ben Rogers, Nutanix

Transcript

WEBVTT

1
00:00:03.940 --> 00:00:33.670
Philip Sellers: Hey, and welcome again to another episode of Mutanix Weekly. One of the many podcasts we produce here at Zintegra, it is content with context, as we like to say and appreciate you listening in to another episode. I'm your host today, Phil Sellers, I'm the practice manager for modern data center here at Zintegra, and I'm joined with a few friends around the Horn today. So 1st up is Chris Calhoun, one of the solutions architects

2
00:00:34.131 --> 00:00:43.129
Philip Sellers: here. It's integra. Chris. I know you're a a newer voice here. How long have you been with integra now?

3
00:00:43.802 --> 00:00:57.067
Philip Sellers: Started in October? October first.st So 4 months. Yeah, so practically an old timer. He's working working towards that status with us at this point. And former nutanix

4
00:00:57.700 --> 00:01:01.570
Philip Sellers: sales engineer. That's exactly right. Focused in the enterprise space.

5
00:01:02.467 --> 00:01:13.309
Philip Sellers: We're also joined with. Jira Cox. Jira is a chief technologist. Always get your title wrong. This is gonna become a gag. I think.

6
00:01:13.571 --> 00:01:18.009
Jirah Cox: I, we should think of a better gag. But I say, architect on our field CTO. Team.

7
00:01:20.100 --> 00:01:22.660
Philip Sellers: All right and smart guy.

8
00:01:22.990 --> 00:01:27.409
Philip Sellers: That's that's what that stands for, right? So there's there's the gag.

9
00:01:28.653 --> 00:01:37.320
Philip Sellers: and also we've got Ben Rogers. Ben is an enterprise se for Nutanix, based out of the Charlotte market right.

10
00:01:37.500 --> 00:01:42.600
Ben Rogers: That's correct. Glad to be here today from our gts in Austin, Texas.

11
00:01:43.300 --> 00:01:48.429
Philip Sellers: That's right. We've got a remote one. We're all in the same city for

12
00:01:48.810 --> 00:02:03.450
Philip Sellers: a different sort of podcast today. But we're getting ready to to learn what's coming, and all sorts of new things that we'll be able to talk about here in the coming future. But

13
00:02:03.630 --> 00:02:22.570
Philip Sellers: you know it's it's funny. There's no absence of announcements and new things happening inside the Nutanix ecos system. We've got a lot of new announcements and a lot of new versions of software and capabilities that have just come out. So I think it's app that we start with

14
00:02:22.570 --> 00:02:38.023
Philip Sellers: the core. And let's talk about some new Aos caught off the presses at the end of January. We've got the launch of Aos 7 point. Oh, and that's gonna make a lot of people happy.

15
00:02:38.710 --> 00:02:53.218
Philip Sellers: new features, new capabilities. And we'll talk about some of those today. But this is hot on the heels of a another major release version 6.10, which rolled up a ton of capabilities. And so

16
00:02:54.570 --> 00:02:57.779
Philip Sellers: I guess, Jira, I'll throw it to you. Is there

17
00:02:58.380 --> 00:03:05.419
Philip Sellers: any particular reason that we've gotten such a big release so quickly after another big release.

18
00:03:06.150 --> 00:03:09.821
Jirah Cox: I mean it was Christmas time. But

19
00:03:10.510 --> 00:03:11.030
Jirah Cox: Of

20
00:03:11.670 --> 00:03:31.559
Jirah Cox: a better reason to spoil the end of the article. It talks about how this release helps us launch and springboard into our new unified Nci release model moving from the the past of the Sts Lts model to now in the new one with the the planned I think it calls out twice a year. Cadence!

21
00:03:31.839 --> 00:03:49.139
Jirah Cox: Where customers can expect, you know. When? When can I expect releases from nutanix that bring new features? When can I expect maintenance releases? How long can I expect to run those releases? So we're we're moving into a you know, into kind of a 2 big releases per year that are, they're here for all customers to consume

22
00:03:49.818 --> 00:03:59.360
Jirah Cox: now. So that's that's that's the real answer is that 6 dot 10 was our last Lts, and then this is our new our new release model.

23
00:04:00.440 --> 00:04:25.290
Philip Sellers: That's awesome. I mean, this release is a big deal. We're we're celebrating Ahv's birthday. It turns 10 version 10 is packaged in as part of Aos. 7. So that's a pretty big deal, I mean. Ben, I'll ask you, I mean from from your perspective 10th anniversary, and also ahv 10, I mean, what's that mean for customers?

24
00:04:25.490 --> 00:04:46.559
Ben Rogers: Man. It's a proven solution. It's not something that's just been developed. It's something that we've had developed. It's deployed. It's hardened. It's proven. That's what the 10 years means. And so when we talk to customers about, you know, a vmware takeout, or, you know, needing automation and environment. And they need a hypervisor. Whether it's for containers or something else.

25
00:04:47.060 --> 00:05:08.620
Ben Rogers: we can go in confidently and go. This is a proven solution. This is hardened. We have this in government agencies. We have it commercial. We have it in enterprise. A lot of people are now running this, and since you've seen the broadcom vmware acquisition. You've just seen the adoption of Hv. Just J. Hook. So good stuff, man. So 10 years to me means reliability and proven performance.

26
00:05:09.530 --> 00:05:18.019
Philip Sellers: Yeah, I like that, too. I mean, it's it's battle tested. There's thousands and thousands of deployments running all over the world

27
00:05:18.220 --> 00:05:46.080
Philip Sellers: in all sorts of different places from the data center to the cloud to the edge, as the blog post says. Here. So let's dig into the goodness of this new release. And the launch. We're reviewing a blog post. It's aptly named Aos, 7.0 launch blog. It's out on nutanix.com slash blogs. And this particular one's written by William Parks. So shout out to William Parks, thanks for

28
00:05:46.310 --> 00:05:49.690
Philip Sellers: posting and giving us something to talk about.

29
00:05:50.217 --> 00:06:01.600
Philip Sellers: So so let's dig in on some of the features. First, st up inside of the blog post. We're talking about high availability and disaster recovery, both of those really important

30
00:06:02.317 --> 00:06:16.059
Philip Sellers: concepts for all customers. But, you know, these are things that have long been inside of Nutanix. So what are we actually talking about in terms of new capabilities that we can make use of. Gyro. I'll throw it to you.

31
00:06:16.550 --> 00:06:38.059
Jirah Cox: Sure. A a big one. Yeah. I mean, I don't want to speak for the whole podcast. But I think I'm, I'm safe to say, we're all pro high availability, pro disaster, recovery here right? Not not comfortable topics. But but cool to see them enhanced here for sure. A new way that we can make clusters more resilient right supporting this new failure mode.

32
00:06:38.430 --> 00:07:05.110
Jirah Cox: If you've been with us in a meeting, been with us near a whiteboard you've heard us talk about, you know, Rf, 2 and Rf. 3, and how we can design a cluster to lose like a node or a disk unexpectedly, or to lose 2 nodes or 2 disks unexpectedly, and with this new design we can now at the edge. Design a 3 node cluster, to be able to suffer, to withstand the loss of a 1 node and a drive

33
00:07:05.220 --> 00:07:19.390
Jirah Cox: at the same time, and keep on running so traditionally in the past, on 3 nodes that would have been violating the configurable availability of a 3 node cluster. But now, with this new enhancement, we can actually do 3 nodes

34
00:07:19.390 --> 00:07:37.909
Jirah Cox: and still lose a node and a drive all at once, and keep on ticking. So that's great, making making it harder and harder for the cluster to fall over even as it sustains, you know, the real world events going on around it right? Or somebody in the data center, you know, tripping pretty badly, or some cables and yanking some disks out as they fall down.

35
00:07:38.560 --> 00:07:53.380
Philip Sellers: Yeah. And and it's interesting that you were talking about this in a 3 node cluster configuration. Because, Chris, I know as we're talking with customers. This is a configuration. We're talking about more and more. As we're exploring new.

36
00:07:53.830 --> 00:07:56.630
Philip Sellers: I guess designs with our customers

37
00:07:56.960 --> 00:07:59.909
Philip Sellers: absolutely. And I think that goes back to

38
00:08:00.850 --> 00:08:19.499
Philip Sellers: the nutanix. Solution is a true enterprise grade solution. That's the difference in the focus on these edge locations now, especially being able to right size, the configuration to know that you're getting the best performance and availability and not just a

39
00:08:20.100 --> 00:08:38.030
Philip Sellers: hey? Let's install something and look the other way for our customers. We're really focusing on these remote locations being just as important. And I think that that's really key, too, of having a solid foundation and performance at those edge locations that they can rely on.

40
00:08:38.600 --> 00:09:00.580
Philip Sellers: Yeah. And you know, I think about also the licensing aspects. You know, we we have the new. I guess it's not new anymore, but a robo, friendly, licensing method that I think is also increasing that interest on the edge. Ben, I see you shaking your head. Are you seeing more interest with that based on the licensing availability.

41
00:09:00.580 --> 00:09:28.740
Ben Rogers: Well, the the robo license is definitely an advantage. But I just see customers, as Chris said, they're looking at edge is more strategic than they had before with AI coming down the path. Man scanning technology, particularly in manufacturing. These are all things that need to be right. Next to the lines or on the edge sites. And so the importance of the edge clusters and their performance and their reliability is just becoming more and more important as we roll these services out to the edge platforms.

42
00:09:28.960 --> 00:09:54.210
Ben Rogers: So, as Chris said, a lot of customers, you know, concentrate on data center edge was just, you know. Print active Directory. Now you're seeing people going. No, I need my edge services to be able to run these type of AI applications scanning applications. They're looking at quality and assurance and stuff like that. So you're definitely see a larger push to edge installs. And the enterprise performance that you need from those installs.

43
00:09:56.225 --> 00:10:02.600
Philip Sellers: You know, as we. We talk about those use cases. You know, we we talk about.

44
00:10:02.870 --> 00:10:09.232
Philip Sellers: you know, other configurations. One that comes to mind is Metro clusters, too. When we

45
00:10:09.730 --> 00:10:26.370
Philip Sellers: start talking about that high availability and making sure that things are always up and available. Metro is probably the gold standard there. There's some new things coming for metro users as well. You want to dig in on that one.

46
00:10:27.300 --> 00:10:36.439
Jirah Cox: Sure. So for to add some context like you teed us up for in the past we've done metro on Nutanix with ahv.

47
00:10:36.840 --> 00:10:59.209
Jirah Cox: we would have you know, 2 workload clusters right where any your vms could float around there and run on either cluster A or cluster B, or both in parallel for active, active, and then prism central lived outside of those 2 clusters and would live in a management site. That was the 3rd site acting like a witness site, and the witness functionality was embedded within prism. Central worked great.

48
00:10:59.528 --> 00:11:09.391
Jirah Cox: It meant that your management cluster, you know, maybe needed to be a full cluster. Maybe it needed to be a little beefier right to run prism central there.

49
00:11:09.780 --> 00:11:10.515
Jirah Cox: the

50
00:11:11.592 --> 00:11:23.679
Jirah Cox: With this design. We can now also accommodate customers that want to run that witness kind of more like as small as possible, right? Like single node, very tiny and just run the witness functionality itself, right? That tiebreaker of.

51
00:11:23.930 --> 00:11:48.470
Jirah Cox: You know, did a site go down, or is there a network link here that's down instead, are we split, brained, and you know, and way way in there, right vote for me as a quorum member. There, on what do you see? Witness there. So now we can run a lightweight witness on that on a 3rd site which now let's just put prism central at each of those workload sites, so you can have prison central that monitors and manages site a prison central, that monitor manages site B, and then they both just chat

52
00:11:48.470 --> 00:12:13.130
Jirah Cox: with a lightweight witness that can run on a single node can run somewhere else. Maybe if you're doing a factory right, you've got data center data room A and B. Then you have a little node somewhere on the factory floor, right tucked under a desk or bolted to a wall, and all it's there to do is be a quorum tiebreaker. So now more flexible deployment options to the way the way you opened it up, Phil. More flexibility around deploying truly resilient architectures for customers.

53
00:12:14.510 --> 00:12:23.858
Philip Sellers: You know this, this one's near and dear to my heart because we did some really early metro clusters at a phone company. I used to work with. And

54
00:12:24.270 --> 00:12:40.940
Philip Sellers: you know, these these designs have challenges that are a little different than other challenges. And so you guys are tackling a big one, which is that witness functionality. And where does that need to live? And you're offering that that flexibility. I think that underscores it really. Well.

55
00:12:41.662 --> 00:12:49.087
Philip Sellers: you know, 7 dot O is also making improvements when it comes to nutanix multi cloud snapshot technology.

56
00:12:50.490 --> 00:13:03.820
Philip Sellers: when when we talk about this, maybe people listening haven't heard of multi cloud snapshot technology. So, Ben, what is that technology? And and has it helpful for us.

57
00:13:04.980 --> 00:13:23.580
Ben Rogers: The cloud snapshot and correct me. Keep me honest here. This is the ability to be able to snap to other clouds so you could be in azure. But you might want to do a Dr. Or a snap out to aws, or you might be on Prem, and you want to snap out to azure. You want to snap out to aws.

58
00:13:23.640 --> 00:13:38.349
Ben Rogers: The big deployment here is now being able to directly do that to objects within Nutanix, and also s. 3 within aws. That was not a feature that we had before, from what I understand, and has been introduced in 7 oh.

59
00:13:39.390 --> 00:14:06.229
Jirah Cox: Yeah, so it it. The 1st offering of this ran in Aws, where you could run a cluster in Nc. 2, and then also have Dr. For that cluster right to take all the data in that cluster and bunker it somewhere else, like in an S. 3 bucket staying in aws with this enhancement. We've now taught the clusters taught all of our clusters everywhere, not just in aws how to send their data both to aws s. 3, as well as to.

60
00:14:06.230 --> 00:14:28.460
Jirah Cox: of course, objects on a 10 x. As well, which, of course, is s. 3 compatible storage. So now I can have a cluster that if if the goal is just, you know, send a secure copy of a snapshot of data right? Which can include all vms if I want it to, or even just one vm. Send it somewhere for security, whether that's for, you know, sending it down to S. 3. And now I can do all kinds of awesome data level protection on the S. 3 protocol.

61
00:14:28.460 --> 00:14:46.660
Jirah Cox: or to your point, Ben, like, send it across an availability zone, right? So from cloud region to cloud region from on prem to cloud. It gives us more more ability, more flexibility, and even gives maybe some of our smaller customers or customers with like lesser sensitivity, environments and ability to do Dr.

62
00:14:46.660 --> 00:15:00.739
Jirah Cox: Maybe at an even more economical price point right? Because now there's no compute. I'm literally just taking bits somewhere, either in the cloud or on Prem, and just sending them to the cloud as storage versus needing to run a full Dr. Cluster. So there's a lot of attractiveness there.

63
00:15:01.190 --> 00:15:15.170
Ben Rogers: Let me let me ask a question. You gave a little bit of a history lesson here, being a new Tanix employee, I think, is kind of fascinating. This technology started out in our cloud clusters. And then we built that back into our on Prem environment. Is that is that what you said.

64
00:15:15.470 --> 00:15:37.040
Jirah Cox: Right. We we shipped earlier versions of this last year as Nc, 2 cluster protect right where a cluster running an Nc. 2 could predict protect itself. Poster running an Nc. 2 and Aws specifically could protect itself to S. 3. Storage in aws. Specifically, this is now kind of an enhancement and more broadly applicable sort of second coming of that technology.

65
00:15:37.370 --> 00:16:05.149
Ben Rogers: See for me as an employee. I I think that's awesome, that things were developed in the cloud and we're backfilled into the on prem environment. Usually it's the opposite. You get it on, Prem, and then you try to force it out to the cloud. So to see that development path to me is pretty cool, man to to take the clouds resources and go. Hey, people need this on Prem, or we need to be able to bounce from one cloud to another to on Prem again, just the development process is just neat to me when I heard that that history lesson that's cool.

66
00:16:05.450 --> 00:16:15.789
Jirah Cox: Yeah, when you don't differentiate and you just develop one cloud platform. It gets easier to kind of offer, you know. Goodness broadly to more customers all at once.

67
00:16:17.350 --> 00:16:25.590
Philip Sellers: And Jara. I think I remember this being a topic from the dot next release this past year, and obviously to see that

68
00:16:25.710 --> 00:16:52.230
Philip Sellers: implemented and used in a way to help folks focus on disaster. Recovery in 2025, I think, is super critical in the sense that you know there are customers that don't want to invest in a full, complete second data center for resiliency purposes that really just sits there. And I think that this offers a good flexible option, so that you can

69
00:16:52.480 --> 00:17:14.569
Philip Sellers: know that your data is backed up as well as being recoverable in a more flexible manner instead of brick and mortar. We've got to have that second data center. We've got to have a like for like infrastructure. And I think that that from a financial perspective really does lend itself for future improvements to for the Nutanix platform.

70
00:17:15.359 --> 00:17:37.959
Jirah Cox: Yeah, I think I mean, it's it's sad to say, but we all watch the headlines every day, right? And we can spot companies that we wish had had better protection, better resiliency, more options for recoverability. And really, it's all of our day jobs right to meet with, you know, customers like that and say, Look, I know the tech. You know your business and what you need. Let's find the right solution here and now. We have even more tools to bring to bear, to do that.

71
00:17:39.160 --> 00:17:47.260
Philip Sellers: Well. And and I think that the snapshot technology being able to store this long term in an S. 3 compatible object store.

72
00:17:47.320 --> 00:18:04.429
Philip Sellers: opens up a lot of, you know, freedom on. Prem. So you're not having to maintain long term snapshots and pay a penalty on your very expensive hardware, maybe on, Prem. And so you've got another way to do that long term

73
00:18:04.847 --> 00:18:21.530
Philip Sellers: Archival of those snapshots. And and I think that's a great use case in addition to Dr. From a compliance standpoint, from any number of reasons being able to to keep this online and available. But on a

74
00:18:21.620 --> 00:18:25.949
Philip Sellers: less expensive tier storage is an advantage to the customer.

75
00:18:28.200 --> 00:18:37.149
Philip Sellers: So let's let's roll forward. So we've talked a lot about high availability and disaster recovery. But that's not the only goodness. We

76
00:18:37.410 --> 00:18:58.510
Philip Sellers: have some additional automation and management things that are coming. This blog post doesn't go into a ton of details around automation and management. But there's 1 big factor that we are getting with. Version 7. 0, and that's new Apis. What? What's in this release that that we can expect.

77
00:18:59.620 --> 00:19:00.070
Jirah Cox: So.

78
00:19:00.070 --> 00:19:00.750
Philip Sellers: Where do you.

79
00:19:01.160 --> 00:19:27.939
Jirah Cox: So with so with 7 0 our, our V 4 Apis are now generally available. What to say about this is, you know if you've been watching Nutanix for long enough. You've heard of at least a handful of Api versions from V 2 and v. 3, I think back in the day, like version, like 0 point 8, and all kinds of of wild versioning to this what this is is really a wonderful, grand, unifying of all of our Apis. So now anything with Nutanix.

80
00:19:28.190 --> 00:19:37.489
Jirah Cox: there's 1 common Api standard to do that thing, and one place to go. They all flow through prism central. So that's awesome. It gives a lot of clarity, gives a lot of agility

81
00:19:37.490 --> 00:20:00.839
Jirah Cox: to our partners. Right? Think about anything that I do on a nutanix cluster. Even our actions right like when prism central talks to a cluster to build vms or replicate them. Now we even have better Apis that we can consume for our tooling, but also for things like, you know, all of our backup partners, disaster, recovery partners, monitoring partners. This all gives them more powerful and more centralized and more streamlined Apis

82
00:20:01.176 --> 00:20:30.113
Jirah Cox: to make integrating with our stack so much easier as well. So it's it's great. It's more like a cascading release of like. It's 1 thing that we do that, then, has a hundred other things that unlocks, and and you'll see more goodness coming here both from us and from our partners. But it is wonderful to to now have it out out the door. If this is something that you care about. As a listener like check out dev, which is where we have all of our api documentation code samples, sdks, all documented.

83
00:20:30.470 --> 00:20:36.859
Jirah Cox: so it's easy to go go in there, take a look around and start kicking the tires on these more easily and faster.

84
00:20:37.530 --> 00:21:04.120
Ben Rogers: That's that's been one thing that has fascinated me about the platform is not only from an infrastructure stance, but what it brings to the development side of the house, and this consolidation of the Apis and the version 4 coming out is just another testament of that we're not only keeping up with infrastructure. But we're also keeping up with developers need to develop applications and services on that infrastructure. So just another commitment to us as an it community to stay ahead of the curve.

85
00:21:05.820 --> 00:21:25.170
Philip Sellers: Well. And and I think this is also underscoring kind of A and Jerry use the word unification a couple of times as we've talked today. But this, this also is kind of that migration towards prism. Central is the place to end. All be all more and more functionality, moving up

86
00:21:25.410 --> 00:21:35.590
Philip Sellers: out of prism element into Prism Central. Now you've got that central point for all Apis that there's a lot I think that

87
00:21:35.740 --> 00:21:53.179
Philip Sellers: that gets unlocked because you're thinking backups and things like that. I'm I'm now thinking, you know, what can we do from an assessment standpoint? We've got a central place to tap into all of that goodness and that data and help our customers. As we kind of look at your clusters, so.

88
00:21:53.450 --> 00:21:53.890
Jirah Cox: Sure.

89
00:21:54.096 --> 00:21:54.510
Philip Sellers: A number.

90
00:21:54.510 --> 00:21:55.010
Jirah Cox: We'll end up.

91
00:21:55.010 --> 00:21:56.840
Philip Sellers: You can extend that.

92
00:21:57.200 --> 00:22:04.719
Jirah Cox: Yeah. Terraform ansible also will, you know, are powered by the same kind of behind the scenes enhancements as well. So yeah, it's it's very exciting.

93
00:22:06.090 --> 00:22:33.480
Philip Sellers: And back a quick story back. Whenever I was working at Nutanix I had a customer looking to integrate service now with some of their nutanix platform, and and at the time I think the word unified makes sense. To Jair's point is, they were bouncing between version 2 and version 3 Apis. And now we've got a solid foundation to point all back to Version 4 with specifics around prism central. And I think that

94
00:22:33.640 --> 00:22:51.229
Philip Sellers: our tie-ins and our offerings from service. Now, customers that are looking for Zintegra for guidance around, that we definitely can help with that, and it helps to have a seamless platform with Nutanix to now tie those 2 and integrate those 2 together.

95
00:22:53.090 --> 00:22:59.249
Philip Sellers: and as we talk about everything that's going on from a ha and

96
00:23:00.700 --> 00:23:16.069
Philip Sellers: disaster, recovery standpoint. Security is a huge part of that. And we really didn't focus in on that. But there's security enhancements that are also coming with version 7. So let's dig in a little bit there on the security enhancements. Jairo, I'll throw it back to you again.

97
00:23:16.810 --> 00:23:21.099
Jirah Cox: Sure. So thinking back to kind of how we started our

98
00:23:21.650 --> 00:23:50.159
Jirah Cox: life in the data center as a product, right? And thinking about data at rest encryption? Or how do we? How do we write customers? Data securely? Encrypted at rest on a storage platform? You know, when we got our start. What? 1314 years ago? It was almost like that Silicon Valley joke about this is where your box will sit right. And that was where the keys in the data center lived like you could point and say they're either in this box or that box or some combination of the 2, and like that was where the keys lived.

99
00:23:50.380 --> 00:24:01.530
Jirah Cox: and fast forward now to 2025, and most customers are not wanting to invest in like a hardware. They're called a key management server like that. A Kms now is usually more like a cloud service.

100
00:24:01.610 --> 00:24:31.389
Jirah Cox: And so with this, it's I was delighted to see this move from, you know roadmap into a Ga feature. We can now use the azure Kms functionality right? So that this can. Now you can use a cluster on Prem store your data securely. Use the cluster. Excuse me anywhere, right? So on. Prem is one way to do it, but it could be a cluster anywhere that can reach azure and now use their Kms service right? So now, what was a hardware defined box of a workload, you know, in in decades past. Now, as an Api call into the cloud service.

101
00:24:31.390 --> 00:24:46.789
Jirah Cox: But it's still that secure key vaulting somewhere else. Right? So then you get the outcome as a customer of either checking that compliance and regulation box, or even the true mitigation of like. Now, the the threat of hardware theft does not expose my data to bad guys.

102
00:24:49.850 --> 00:24:56.250
Philip Sellers: That key management is not always a very

103
00:24:56.960 --> 00:25:02.200
Philip Sellers: thought of process, but it's important where those keys are stored, because.

104
00:25:02.778 --> 00:25:15.270
Philip Sellers: just like your house, you lose a key. You're locked out, and I think a lot of customers don't spend time thinking through the implications of where their keys are managed. But having

105
00:25:15.920 --> 00:25:39.499
Philip Sellers: unfortunately seen the effects there, it's it's something we definitely need to talk more about from an advisory standpoint with customers key management where we're keeping those how we keep them secure. There's a whole top track, I think, just around that topic that is not really getting as much airtime as as maybe it should.

106
00:25:39.960 --> 00:26:06.970
Jirah Cox: Yeah, I mean Nutanix. As a cluster for hosting workloads is wonderful at so many things. I think one thing that no one wants you to ever do is have us host, that key management server on us managing our own keys. If that doesn't make any sense, give me a call. We'll have a little chat about chickens and eggs, and how that works. And so, as a as a as a simple solution to a complicated requirement here. It's a it's great to have this available now.

107
00:26:09.630 --> 00:26:22.490
Philip Sellers: That's not the only thing that we're talking about from a security perspective flow network security is getting some love. We're getting multi environment isolation policies which

108
00:26:22.920 --> 00:26:36.279
Philip Sellers: I'm not a networking person. And I didn't say the Holiday Inn express last night. So I'm gonna need Jaira to jump in here and exact. Explain what this multi environment isolation policy is.

109
00:26:36.540 --> 00:27:04.169
Jirah Cox: That's a great question. So we've had the ability to, you know. Go in there and and to do the easy se level demo say, like isolate prod from dev vms easily with like a checkbox. Right? The real world. I don't know if you know this tends to be a little more complicated than the average se demo. And so now, like, what if I have a test and a uat and a pre prod and a smoke test environment. And now we can have multiple environments that we can simply say, all of these can't talk to anything else. Right? If you're in one of them.

110
00:27:04.220 --> 00:27:26.930
Jirah Cox: You can only talk to the one you're in and not the others. Right? So that can now be. I don't know what 2 points to find a line. So 5 or 6 points is like a cloud. Maybe. But you can have much more complicated isolation environments here, right, so that you know not just prod and dev, but now prod and dev and Uat, and and whatever can all be isolated from each other.

111
00:27:27.550 --> 00:27:51.969
Jirah Cox: And then there's also some some some goodness here that came to even the ui right around. You know, helping our our customer administrators be able to quickly take stock of what is this flow rule doing? What's it blocking? How would I add? New allowed sources or destinations, or protocols? Or or even do things like reuse like, if I define my active directory domain controllers one time

112
00:27:52.220 --> 00:28:09.810
Jirah Cox: I should be able to use those in multiple policies right? And use that kind of almost like a variable expansion within. My firewall rules more repeatedly. Right? So kind of, you know. A little bit more blending, maybe traditional firewall management ux with Vpc. Management and micro segmentation on our platform.

113
00:28:11.718 --> 00:28:24.570
Philip Sellers: I like the way that it sums it up here. Less complex policies can be updated faster with more confidence. So again, one of the core tenants of nutanix simplification.

114
00:28:25.228 --> 00:28:31.160
Philip Sellers: Kind of ringing. True as we we see this technology get implemented. And I mean, I guess

115
00:28:31.490 --> 00:28:44.649
Philip Sellers: you know, as the outsider to Nutanix. I appreciate the simplicity that gets driven into flow network security. It's it's approachable and a way that

116
00:28:45.203 --> 00:28:54.759
Philip Sellers: makes it easier for a customer to be successful at the end of the day as they approach micro segmentation. So that that's a really welcome

117
00:28:55.438 --> 00:29:01.990
Philip Sellers: thing to see that continuing even as we do more from a functionality standpoint.

118
00:29:02.580 --> 00:29:12.209
Ben Rogers: I think one of the cool things that's mentioned. This article is about just the power of our categories. I talk to customers all the time about how they can use categories to organize their environments and

119
00:29:12.210 --> 00:29:37.140
Ben Rogers: be able to kind of protect yourself. So if you've got categories implemented, if you have, like your your replication policies and categories. And you put a Vm in that category. It's automatically gonna fit those replication policies. The same thing here for the security policies, you'll be able to inventory and be able to also take multiple multiple services or multiple apps and put them under one category so that they all can receive the same

120
00:29:37.140 --> 00:29:51.900
Ben Rogers: policy. So again, it talks about that here in this article. But, man, we're always talking to customers about how they can use categories to stay organized streamline again, simplify the environment. So you know what's going on, and you can look very quickly if you have questions about it.

121
00:29:53.390 --> 00:30:22.190
Philip Sellers: Well, and it doesn't stop with security when we're talking about networking in this release. There's more things happening also in the flow, virtual networking, which is the overlays portion. Traditionally, we think of that as the overlays and the Vpcs. But we're also getting load balancing so integrated layer 4 load balancing built into Ahv. That's pretty exciting enhancement.

122
00:30:23.910 --> 00:30:26.710
Philip Sellers: The fact that we can do this inside of

123
00:30:26.830 --> 00:30:30.799
Philip Sellers: one system as opposed to having to layer in other solutions

124
00:30:31.090 --> 00:30:48.961
Philip Sellers: and the program ability that it should bring us. Both are exciting to me, because that helps in the journey around Cloud Native and Kubernetes, because you you rely on services like load balancing within their frameworks as well.

125
00:30:51.190 --> 00:31:00.200
Philip Sellers: I do want to kind of focus a little bit more on the performance enhancements that come with Aos. 7, you know, as I talk to customers.

126
00:31:00.340 --> 00:31:04.049
Philip Sellers: Chris and I both. We we were with a customer a couple of weeks ago.

127
00:31:04.670 --> 00:31:18.419
Philip Sellers: We were touting the fact that as Nutanix continues to develop and release, things are getting faster, things are being optimized. So you're getting more performance out of your storage, you're getting more performance out of the Hypervisor

128
00:31:18.740 --> 00:31:27.880
Philip Sellers: Aos 7 has a lot of things here that we're we're starting to tap into what's happening from a performance standpoint. Jaira.

129
00:31:29.200 --> 00:31:36.680
Jirah Cox: So when you're using our flow networking stack right with Vpcs, which is like our software find networking functionality.

130
00:31:36.940 --> 00:31:52.975
Jirah Cox: There's ways you can choose to deploy that either for Nat or for no Nat functionality. Right? How? How should our software defined networks connect to your let's say, hardware defined networks right as a routed decision right over either static routing or Bgp things like that

131
00:31:53.310 --> 00:32:09.509
Jirah Cox: or over like net gateways. And so what's cool is released now is that ability for those gateways to be highly available right? We can scale up and deploy multiple of those for you, both for more throughput and for also faster failover, if one of them were on like, say, a piece of hardware that does go belly up. So so

132
00:32:09.510 --> 00:32:24.129
Jirah Cox: more good things, not the traditional area where we thump our chest around too much around. Performance is like network packets themselves. But it's a critical functionality to deliver right as part of being, you know, being acting and delivering like a cloud operator and a cloud platform.

133
00:32:26.410 --> 00:32:41.300
Philip Sellers: Yeah. And and as the the blog goes on to say, we're also complementing those changes with new hardware, you know, Intel, their 4th generation, sapphire Rapids and you know.

134
00:32:41.880 --> 00:32:52.890
Philip Sellers: Amx. On. Hb, those are also driving additional performance. What's what are we tapping into there? That that's helping us out.

135
00:32:54.191 --> 00:33:02.579
Jirah Cox: Tapping into global demand for AI on all of the things. I would say as a simple, a simple answer, right?

136
00:33:02.580 --> 00:33:05.670
Philip Sellers: Just reacting to. Then, instead of tapping into.

137
00:33:08.650 --> 00:33:12.669
Jirah Cox: the word is inferring. So the the

138
00:33:12.970 --> 00:33:36.750
Jirah Cox: inference is the name of the game. But if AI in the data center stays defined by expensive gpus that consume all of the power and generate all of the heat and are the constraint. Then there's a scaling limit, right when there's other places that AI can run in the data center like powered by Amx on the Intel chip itself in the host.

139
00:33:36.750 --> 00:33:43.765
Jirah Cox: Well, now, it's more interesting, right? Because guess what? Every data center has, kind of by definition more cpus than it does Gpus.

140
00:33:44.700 --> 00:34:05.350
Jirah Cox: I'm sure there's some exception that I'll hear about after this podcast goes out, but for the most part more cpus than gpus in the world. So the more that cpus can do AI offloading from Gpus. Well, now, there's just more more computing available to AI, broadly speaking, in the world. So it's great to see that. And I can think of cool stuff like even at the edge, perhaps right like the edge is, gonna have cpus

141
00:34:05.607 --> 00:34:15.380
Jirah Cox: and when those when those can do inference. Well, now, I have more AI closer to maybe my pieces of the business that generate data that I want to feed into an AI model for smarter, faster inference.

142
00:34:17.100 --> 00:34:31.719
Philip Sellers: Yeah, I see a complaint from a Bitcoin miner coming your way. But outside of that I think it's probably true. Most of the times we've got more cpus than we've got, Gpus, and that

143
00:34:32.199 --> 00:34:45.660
Philip Sellers: that is an interesting topic. We were talking a little bit about sustainability, so I'll ask about this, too. You know we've we've got more and more demands, both from CPU and Gpu for power. And so

144
00:34:46.389 --> 00:34:54.650
Philip Sellers: that's becoming a more limited resource, and we're starting to hear people talking about planning for the power consumption

145
00:34:55.159 --> 00:34:59.290
Philip Sellers: at a at a customer level. Chris, I mean, what?

146
00:34:59.620 --> 00:35:08.809
Philip Sellers: What? What are you seeing? I I guess, kind of from that perspective, as you're working with our our customers and talking to them around power consumption in the data center

147
00:35:09.570 --> 00:35:22.200
Philip Sellers: specifically around the AI model. Along with the Gpu discussion there was a a small drafting or type

148
00:35:22.420 --> 00:35:39.460
Philip Sellers: establishment that was a very small set of engineers that were planning growth without necessarily knowing what they're planning for. And in that they were specifically looking at Nutanix because the idea is

149
00:35:39.520 --> 00:35:57.910
Philip Sellers: with the solutioning in Nutanix. Obviously, as a software defined solution. They get the benefit of these 7.0 improvements. And that's really where they saw an opportunity of all right. How do we get in now with a solid foundation from Nutanix, knowing that

150
00:35:58.050 --> 00:36:11.440
Philip Sellers: we're going to need Gpus and either everything, or have to look at something, or should look at something more focused around. Hey? Can we focus on those better cpus? And maybe this being a small site.

151
00:36:11.620 --> 00:36:21.479
Philip Sellers: then this makes sense to focus on Nutanix. And that's where I think that it's a. It's a balance. Obviously, power consumption is something that's it's

152
00:36:21.670 --> 00:36:28.170
Philip Sellers: it's becoming more relevant. But I wanted to give that example, because to me that makes sense, because

153
00:36:28.310 --> 00:36:55.539
Philip Sellers: customers are having to decide, where do we invest? Because legacy hardware is not going to do it. They've maintained that as long as they could. And now it's time to forward plan and look at the next 10 years. What's the technology that's going to run my stack, and I know that Jara has mentioned this before. It's Nutanix is here for today's workloads, but also for the next 10 years, and that's the foundation.

154
00:36:55.540 --> 00:36:55.959
Jirah Cox: I think that.

155
00:36:55.960 --> 00:37:16.280
Philip Sellers: I think, is a little interesting, too, because their primary use case was Auc today. But they were really interested in in where things are going in cloud, native Kubernetes and AI for their future, too. So they. They loved the fact that they were able to get a platform

156
00:37:16.570 --> 00:37:26.529
Philip Sellers: that somewhat future proof them because they they are already trying to make strategic sort of decisions around how that affects their business. Moving forward.

157
00:37:27.850 --> 00:37:41.310
Jirah Cox: I think that mobility is so critical, right? We've had that for for years. It's now as important as it ever was right. We had customers over 10 years ago. Right? Who bought clusters random for 5 years. Let's say they started with 8 nodes. They refresh down to 6 nodes.

158
00:37:41.560 --> 00:37:46.210
Jirah Cox: but by doing a simple act of like cluster, expand, and then ejecting the old nodes.

159
00:37:46.220 --> 00:38:14.790
Jirah Cox: all the vms, but also all of the data, all the configurations for that platform all flowed over to the new hardware. We have customers that have actually done this twice now, right? The cluster itself. Log was over 10 years old. The hardware is is is much newer than that right. But the customer got over to it with no forklift migrations, no complicated planning and is able to swap that. And you know, basically change up those tires on the bus or change up the wings on the airplane as we're flying it. And I think that ability. When you're a customer who's who's in that

160
00:38:14.790 --> 00:38:23.760
Jirah Cox: camp of I am constrained in some way, shape or form, whether it's you know, thermals or power or rack space that ability to to

161
00:38:23.840 --> 00:38:37.520
Jirah Cox: to be on nutanix in some iteration, but then move to a faster, denser, smaller configuration has super benefits there right and and able to get there so easily. When the requirements maybe do get solidified in the future. Down the road.

162
00:38:39.570 --> 00:39:07.820
Philip Sellers: It's also interesting to me, because we've had a couple of customers come to us with some accelerated, and I'm being generous when I say, accelerated. Some really fast exits from a data center, because either one condition or another, and they had to get out quickly. The nice thing here is the platform has the ability to to help you with that portability of your data to another location. When that happens as well.

163
00:39:08.328 --> 00:39:30.080
Philip Sellers: Yeah, it's not something I would have thought like moving data centers is not something that generally comes to mind as a often, you know, running kind of initiative. It's kind of a big, painful thing but I've run into too many customers here in the last year that they're faced with that, and the timelines are short.

164
00:39:32.235 --> 00:39:46.990
Philip Sellers: Ben, I know you're you're focusing a lot in on AI and things right now. It was news to me as we were kind of talking before the podcast about Amx and kind of moving some of this towards the cpus. You.

165
00:39:47.240 --> 00:39:48.229
Philip Sellers: I guess

166
00:39:48.680 --> 00:39:54.800
Philip Sellers: how much of this are are you kind of seeing or talking with people from an AI perspective.

167
00:39:55.230 --> 00:40:06.650
Ben Rogers: Well as we were talking. AI on the edge is is, you know, a big thing. They want to be able to have their manufacturing lines have some kind of AI process into it. So that's become really big

168
00:40:06.650 --> 00:40:31.340
Ben Rogers: where I've also seen, you know. And this is kind of we were talking about this before the podcast was this idea of AI agents that'll run locally in the environment. And so I can see a day where your laptop has a enhanced CPU slash gpu that can do some some AI processing on it locally. And so, as you see, the AI market change, I think Jaira mentioned that

169
00:40:31.450 --> 00:40:32.400
Ben Rogers: you know.

170
00:40:32.830 --> 00:40:55.170
Ben Rogers: in some of the training that we've gotten, Gpus is the thing today. But that might not be the thing tomorrow. So this whole AI environment is going to continue to change. It's going to continue to put pressure on hardware vendors, software vendors to keep up. And so the thing that I like most about this with the Intel Amx capabilities is we continue to partner with our hardware vendors.

171
00:40:55.170 --> 00:41:03.130
Ben Rogers: and we continue to look at what's coming down the pipe, and how the AI landscape is changing, and it's changing daily, if not hourly.

172
00:41:03.710 --> 00:41:28.589
Ben Rogers: And so I think we're in a good position as a company to one understand what's going on. And we've got great partnerships with people like Intel and Nvidia and Amd that can really make these things happen. And so I consider us a player in the AI market. I think we're going to be a force in the AI market, and we've got customers that are leveraging our clusters to help deliver the AI initiatives very rapidly, as well.

173
00:41:29.820 --> 00:41:37.280
Philip Sellers: I agree. And and you know, in a rapid, changing sort of ecosystem industry.

174
00:41:37.772 --> 00:41:40.829
Philip Sellers: I mean, this is faster than normal, for

175
00:41:41.180 --> 00:41:48.630
Philip Sellers: for even the other types of innovation I've seen. So you're right that it's it's turning upside down on

176
00:41:49.373 --> 00:41:55.889
Philip Sellers: it feels like on a daily basis. I know that's probably overstating it. But this is moving quickly.

177
00:41:56.270 --> 00:42:02.030
Ben Rogers: What's what's wild about Nutanix to me right now? And I think what's gonna make the gts here just

178
00:42:02.100 --> 00:42:04.200
Ben Rogers: phenomenal and exciting is.

179
00:42:04.220 --> 00:42:18.269
Ben Rogers: you know we're dealing with the broadcom thing. So we're helping customers evaluate broadcom. Do they want to stay? Do they want to leave? How can they enhance their data centers. We're also looking at customers are going. I'm no longer able to support legacy vms. I got to move into a containerized environment.

180
00:42:18.270 --> 00:42:45.699
Ben Rogers: So we got solutions for that. And then you've got this whole AI thing that's brewing that, you know. Man really puts us at the center of 3 big pillars that are going on right now that, in my opinion, and I hope everybody on the call would agree, and I'm sure our CEO and executives would agree. We're perfectly positioned to handle those 3 pillars. So it's according to what our customers come to us, with what problems they come to us with. We probably got a solution that can handle most of what they're dealing with.

181
00:42:47.510 --> 00:42:50.309
Philip Sellers: I agree with that? Yeah, I mean it. It plays in

182
00:42:51.200 --> 00:42:55.670
Philip Sellers: so many of the use cases small to large

183
00:42:56.060 --> 00:43:09.040
Philip Sellers: on Prem, today, you know cloud anywhere in between. Your choice of vendors. There, there's just so much packaged in to to help you gain value for your business.

184
00:43:10.660 --> 00:43:16.839
Philip Sellers: We. We covered this at the top of the podcast, but I do want to come back to it because there are a few more

185
00:43:17.433 --> 00:43:25.309
Philip Sellers: points around the evolved release cycle for Nci or Nutanix cloud infrastructure, starting with 7 point. Oh.

186
00:43:25.827 --> 00:43:34.520
Philip Sellers: you know, for me, the new release model moving away from long term service releases and short term service releases is that

187
00:43:34.770 --> 00:43:42.460
Philip Sellers: these new features get into the hands of customers faster in a mainstream long term supported release. And I think

188
00:43:42.720 --> 00:44:00.669
Philip Sellers: that's a a big advantage for us. You know. I I guess I'll ask the question, Chris, have you seen customers kind of holding back, staying on the Lts and missing out on some of the features and stuff that maybe this new release cycle is gonna help

189
00:44:01.120 --> 00:44:18.210
Philip Sellers: definitely in in the enterprise space where I covered previously. There are customers that would go through one cycle of nutanix installs a year just because they had over 500 nodes. So that's not a small

190
00:44:18.270 --> 00:44:39.750
Philip Sellers: task to undertake, and knowing that it takes a month to get on a new release. Now they see the advantages of the stability of this particular 7.0, and the future models will have that sustainability so that they can plan better, and knowing that they

191
00:44:40.120 --> 00:44:57.179
Philip Sellers: are not missing out, and they will have a more predictable cycle. That's 1 of the biggest things that I tried to help plan and look ahead with them specifically, so that they wouldn't be on the back end of a new update in like

192
00:44:57.300 --> 00:45:08.789
Philip Sellers: just finishing an upgrade so they can stay ahead of that. And I think that that is a again, a focus from Nutanix on. How can we better help our customers?

193
00:45:09.379 --> 00:45:18.099
Philip Sellers: In today's age, and and really, truly listen to what's best for those customers, because you want those

194
00:45:18.520 --> 00:45:31.240
Philip Sellers: improvements. But you've got to balance that out with day to day workloads as far as responsibilities, and I think this lends itself to better structure and better planning to help those customers.

195
00:45:32.520 --> 00:45:38.779
Philip Sellers: So, Jara, you said 2 releases a year. What are the other details in terms of support ability for the new model.

196
00:45:39.710 --> 00:45:53.711
Jirah Cox: Sure. So the the article calls out, How again, this is within our nci product family, right? So that's ahv aos you know, prism central that manages all that the flow networking stack

197
00:45:54.090 --> 00:46:21.159
Jirah Cox: and then, of course, lcm, for all the above. So each release getting enjoying 24 months of lifecycle. So 15 months in the limelight of maintenance and updates 9 months of additional support in that second year for break, fix support and we're still, of course, supporting your your full installation there. So so 15 months of updates covering you for more than a year if you were on annual updates and then also additional

198
00:46:21.423 --> 00:46:30.129
Jirah Cox: runway, if you ever had to delay first, st you know, things like, you know, updating every 18 months, or even, you know, at the at the tail end. Here every 24 months.

199
00:46:31.910 --> 00:46:32.860
Philip Sellers: That's awesome.

200
00:46:35.370 --> 00:46:52.640
Philip Sellers: there's a lot here, and I'm sure we're only touching a small percentage. I've seen other slides and presentations. I know there's a lot coming out in Version 7. So we're gonna have some more things to talk about, things that are layered solutions like in us.

201
00:46:52.800 --> 00:47:08.699
Philip Sellers: self service in upcoming episodes. So I look forward to delving into more like I said, you guys are keeping us on our toes. There's a lot going on a lot of new releases. So I'm really looking forward to exploring more together.

202
00:47:09.222 --> 00:47:29.019
Philip Sellers: and for everyone listening, I would say, stay tuned. We've got a lot more happening. And I'm sure this week the 4 of us are going to get inundated with new information. As we're here at Global Tech Summit, surrounded by all the technical folks at Nutanix, so

203
00:47:29.667 --> 00:47:35.030
Philip Sellers: on behalf of our crew, Chris, Jaira and Ben and myself.

204
00:47:35.250 --> 00:47:44.140
Philip Sellers: Thanks for listening and spending a little time with us, and we will definitely be here again very soon and look forward to catching up.

205
00:47:44.270 --> 00:47:47.450
Philip Sellers: Thanks for listening, and we'll see you soon.