Nutanix Weekly: Elevating hybrid cloud for AI, databases and general-purpose workloads at Microsoft Ignite

Hybrid clouds are gaining popularity and maturity as organizations juggle workloads with different performance, governance and efficiency requirements. Nutanix has long held the view that most organizations will require both on-premises and public cloud presence.

As a result, Nutanix has invested in building a platform that provides the best way to create a unified hybrid cloud architecture. At Microsoft Ignite, Nutanix to shared best practices for running hybrid workloads and simplifying your hybrid cloud deployments.

Blog Post: https://www.nutanix.com/blog/elevating-hybrid-cloud-for-ai-databases-at-microsoft-ignite

Host: Philip Sellers
Co-Host: Andy Whiteside
Co-Host: Jirah Cox
Co-Host: Ben Rogers

Transcript

WEBVTT

1
00:00:02.740 --> 00:00:09.549
Philip Sellers: Hey, and welcome again to another episode of Nutanix Weekly. One of the many podcasts here at Zintegra.

2
00:00:09.630 --> 00:00:29.060
Philip Sellers: content with context is what we like to call them. And we're happy to have you back on listening with us for another episode. I'm Phil Sellers. I am the practice manager for modern Data center here at Integra, and I'm joined with some awesome guests, including

3
00:00:29.160 --> 00:00:35.889
Philip Sellers: none other than our chief enthusiasm officer Andy Whiteside Andy, how you doing.

4
00:00:35.890 --> 00:00:39.400
Andy Whiteside: I'm doing good, Phil, glad, glad to be on for an episode here.

5
00:00:39.400 --> 00:00:41.424
Philip Sellers: I know long time.

6
00:00:42.324 --> 00:00:50.099
Philip Sellers: for a long time you were the the primary host for this podcast. And handed over the reins to me not long ago, and

7
00:00:50.160 --> 00:00:53.279
Philip Sellers: it's good to have you back to join us in this afternoon.

8
00:00:53.280 --> 00:00:55.410
Andy Whiteside: I'll I'll try to contribute where I can.

9
00:00:55.934 --> 00:01:00.134
Philip Sellers: We've also got Jaira Cox. Jaira is

10
00:01:00.720 --> 00:01:07.760
Philip Sellers: chief putin tate from Nutanix. I don't know what title I should give you. I always title right.

11
00:01:07.760 --> 00:01:13.000
Jirah Cox: I've been called the Grand Poobah of the East men's bathroom, working on the west.

12
00:01:14.530 --> 00:01:16.638
Philip Sellers: Working on the West. Yeah.

13
00:01:17.060 --> 00:01:18.259
Jirah Cox: Got to aspire to something.

14
00:01:18.500 --> 00:01:26.730
Philip Sellers: Happy to have you here on the podcast we're also joined by Ben Rogers. Enterprise se here in the Carolinas. Ben, how are you doing.

15
00:01:26.730 --> 00:01:38.059
Ben Rogers: I'm doing great today, man holiday week. So today and tomorrow will be kind of kind of busy. But hopefully, Wednesday, the plane will start to land. Enjoy some family on Thursday and Friday, so I'm in all good spirits this week.

16
00:01:38.210 --> 00:01:56.945
Philip Sellers: That's right. Yeah. And it is the holiday week. So wherever you're at and listening, especially if you're here in the United States. Happy Thanksgiving, and we hope that you spend some time with family, and enjoy some turkey or some tofurky, if you're so inclined.

17
00:01:57.320 --> 00:02:00.570
Jirah Cox: Phil, are you a light light tofu, or dark tofu kind of guy.

18
00:02:00.570 --> 00:02:07.060
Philip Sellers: I am a full, full fledged turkey guy. So yeah, I I will stick to the

19
00:02:07.060 --> 00:02:09.630
Philip Sellers: the anal protein-based anal protein.

20
00:02:09.639 --> 00:02:20.695
Philip Sellers: the animal protein. That's it. Yes, 100%. So yeah, on behalf of everybody. Thanks for listening. You know, as always, we've got a blog post up

21
00:02:21.089 --> 00:02:45.199
Philip Sellers: this past week Andy and I were both out at Microsoft Ignite in Chicago. And so we we had a lot of chance to interact with customers. We brought about 30 customers. It's 1 of the value, adds we. We try to put back into the ecosystem is taking our customers with us to industry events, and so we had a bunch of people that we talked with hosted dinner, for at Microsoft ignite, and.

22
00:02:45.199 --> 00:03:04.549
Philip Sellers: as luck would have it, one of our great partners, Nutanix was also on site at Microsoft Ignite. And you guys were talking about AI databases and all sorts of other things. And so we've got a blog post today called elevating hybrid cloud for AI databases and general purpose workloads at Microsoft Ignite.

23
00:03:04.719 --> 00:03:12.995
Philip Sellers: We'll talk about all the different things that we were unpacking at the event, and dig into those in a little more detail.

24
00:03:13.519 --> 00:03:19.324
Philip Sellers: you know it. It's good to see Nutanix out in the marketplace, not only at their own events. But

25
00:03:19.669 --> 00:03:31.929
Philip Sellers: you know, at bigger industry conferences like ignite 1st time back post covid in person. It's been virtual for a few years. Had about 10,000 people I heard. It seemed busy.

26
00:03:32.109 --> 00:03:36.329
Philip Sellers: Andy, I mean, kind of what was your impression? Just of the week in general.

27
00:03:37.330 --> 00:03:58.519
Andy Whiteside: All things, Microsoft. All things Microsoft partners, which includes partners like us, includes partners like Nutanix. You know the entire, the entire eco ecosystem like the whole ecosystem. Was there? You know a lot of people that we haven't been around a long time where it was good to be back around. I know the pandemic's been over for a little while, but this was the 1st conference where it felt like

28
00:03:58.520 --> 00:04:07.410
Andy Whiteside: all of the vendors and people that we've worked around for all these years were there and active, and everybody was front and center.

29
00:04:07.510 --> 00:04:17.499
Andy Whiteside: Of course, a lot of talk around AI and copilot almost to exhaustion. But I think it's so real that you we have to be talking about it.

30
00:04:19.060 --> 00:04:36.730
Philip Sellers: I I agree. Copilot was a a big, big letter, and I don't just mean the giant letters you could stand in front of and take photos in front of on the show floor. But we saw copilot everywhere. The logo I mean it was it was pervasive. The thing I heard back was

31
00:04:36.790 --> 00:04:43.880
Philip Sellers: how much it's improving, how fast it's improving and getting better at helping people do things. And so

32
00:04:45.300 --> 00:04:50.760
Philip Sellers: there's a lot of different places. You can plug it in. And so that was fun to hear about

33
00:04:51.210 --> 00:05:03.619
Philip Sellers: use cases. How? How are people tangibly getting benefits from that and since we're talking about copilot. It makes sense for us to start off with the 1st topic, which is AI.

34
00:05:03.840 --> 00:05:15.020
Philip Sellers: And as the blog post said, the workload, everyone's talking about just like at Ignite Jaira. Do you want to kind of lead us off? What? What's the 1st big news that you guys were sharing at Microsoft? Ignite.

35
00:05:15.020 --> 00:05:29.279
Jirah Cox: Yeah, so and then, for all of our listeners, right, you can recap this on this blog post. Which serves for us as a an ignite, recap blog post for nutanix, right around what we cover. So you'll find that on our blog for elevating hybrid cloud.

36
00:05:29.290 --> 00:05:34.170
Jirah Cox: The 1st one, of course. Yeah, definitely is AI and helps our customers.

37
00:05:34.190 --> 00:05:43.519
Jirah Cox: 1st off the I don't want to bury the lead. The big announcement. Our product, Nutanix enterprise AI usable, of course, in azure. Right? So now we can offer you

38
00:05:43.790 --> 00:05:53.780
Jirah Cox: that sort of command and control layer to govern your AI deployments on Prem, but also in azure as well. So one consistent identical look and feel

39
00:05:53.880 --> 00:05:57.250
Jirah Cox: for provisioning, for entitling, for operating

40
00:05:57.270 --> 00:06:05.640
Jirah Cox: the Llms. The Gpts that you want to use to power. Your business now, really, truly, anywhere, right within your 4 walls, within azure or both.

41
00:06:06.420 --> 00:06:14.640
Jirah Cox: that has a lot of benefits right? Because the way that I can then do inference in the cloud inference at the edge inference in my data center with my own private data.

42
00:06:15.750 --> 00:06:23.800
Jirah Cox: There's another blog post I saw recently that talked about the larger and more regulated a customer is, the more they want to

43
00:06:23.810 --> 00:06:32.189
Jirah Cox: probably use AI in a private fashion, right? Which doesn't mean they're for their 4 wall data center. It just mean anywhere that is considered within their cloud platform, right? Which

44
00:06:32.230 --> 00:06:35.309
Jirah Cox: part of what we do at Nutanix is, help your cloud. Go anywhere.

45
00:06:35.400 --> 00:06:43.570
Jirah Cox: run anything, run anywhere, including your cloud, can contain public cloud. Right? You can have availability zones in your cloud that are public cloud.

46
00:06:43.660 --> 00:06:55.740
Jirah Cox: so that ability to say, I can now do AI one way across all of this. But I own all of my data. I control all of the access to it. It's all mine. Massive shot in the arm for privacy and really, even data sovereignty. Right?

47
00:06:55.820 --> 00:07:01.930
Jirah Cox: Thinking back a couple weeks. Actually, Phil, to the blog post we covered about Nutanix unified storage

48
00:07:01.940 --> 00:07:09.889
Jirah Cox: for AI training. Remember, we were in that 2 ways, one in our data centers, one on Nc. 2 in public cloud, right? So that ability to

49
00:07:09.960 --> 00:07:21.899
Jirah Cox: host data with full governance, with full sovereignty and privacy over that. But now give that to maybe, Gpus, that I want to rent in the cloud lets that new Phoenix AI fabric really do very powerful things for me.

50
00:07:22.620 --> 00:07:28.399
Philip Sellers: When the interesting thing is the the week after we, we did that, podcast talking about it, I was out

51
00:07:28.690 --> 00:07:48.200
Philip Sellers: with other nutanix technology champions at the San Jose headquarters. We did a deep dive, and they were super super stoked over this performance benchmark we talked about. So if you haven't heard that episode definitely click back a few and listen to the Ml perf podcast episode. It's really great stuff

52
00:07:48.590 --> 00:08:11.380
Philip Sellers: that also gave me the ability to preview what we're talking about here. The enterprise. AI so nutanix enterprise AI or Nai for short, is the next iteration of Gpt in a box. It's the productized version of AI. That's consumable for everyone. And and that's my elevator pitch for it. Right? Is

53
00:08:11.710 --> 00:08:33.929
Philip Sellers: no matter where you run this, it makes AI, which is incredibly complicated, accessible. You get a management plane where infrastructure folks can have a ui, and you can go in and you can add your different large language models. You can expose those endpoints, you can train it. You can

54
00:08:33.990 --> 00:08:39.950
Philip Sellers: kind of control everything in the life cycle of an AI large language model

55
00:08:40.169 --> 00:08:53.600
Philip Sellers: in one place, and so I haven't seen that from anybody else. You know you've got vendor partners like hugging face and stuff like that that are part of this, but nobody's thought about it from an end to end perspective until I saw this.

56
00:08:53.870 --> 00:09:16.689
Jirah Cox: I like the way you put that I'll plant a little flag here. We'll come back to it around, making the difficult more accessible right. And and that is 100%. A great way to marketing might steal that. By the way, just heads up, Phil, that is really what it does right is, it makes things that are very complicated and nebulous, very, very accessible, like, click, click. Next done. Okay. Now, your AI is in production. What did you want it to do again? Let's go start.

57
00:09:16.690 --> 00:09:17.070
Ben Rogers: Right.

58
00:09:18.150 --> 00:09:30.269
Ben Rogers: So what's what's really cool about it for me being out on the sales and seeing how customers are absorbing. This is the simplicity of it also the ability to stand this up fast. So day one's really quick.

59
00:09:30.270 --> 00:09:30.630
Philip Sellers: But yeah.

60
00:09:30.630 --> 00:09:39.710
Ben Rogers: Day, 2 day, 3. Day, 4. The operations piece of it, and the maintenance piece of it. We're also empowering, you know, existing Nutanix engineers

61
00:09:39.710 --> 00:10:00.959
Ben Rogers: to AI now, and you know we're not really changing the game on them if they know Nutanix, and they understand how Nutanix is from a bread and butter standpoint. They can walk in in AI very easily, and so that's been another win for us is for existing customers. They now have, and they have had the platform that they're now going to run their artificial intelligence on.

62
00:10:00.960 --> 00:10:14.329
Ben Rogers: And then the second thing of this is really cool to me is, you know AI has been kind of thought of as a cloud thing, and we've quickly learned that security is is making that where people are wanting to pull that back on Prem.

63
00:10:14.330 --> 00:10:33.200
Ben Rogers: Well, this just gives the best of both worlds. Companies need to run AI on Prem, specifically, in manufacturing and healthcare. It's got to be next to the workload, but they also need to be able to have these engines running in cloud where they can disperse them and scale very easily. And so again, we're bridging that gap between these 2 environments.

64
00:10:33.950 --> 00:10:44.771
Philip Sellers: Yeah. And and it's not always just private data. You know, it's talked about use cases, right? You know, we're we're delving into more and more. How do we make this beneficial for folks?

65
00:10:45.908 --> 00:10:48.840
Philip Sellers: You're a support organization, right? You have

66
00:10:48.890 --> 00:10:54.380
Philip Sellers: tons and tons of cases and notes and things like that you also have knowledge based articles and things.

67
00:10:54.410 --> 00:10:58.930
Philip Sellers: You know you. You turned it loose on your own data. And

68
00:10:59.577 --> 00:11:10.549
Philip Sellers: you know, that's helpful to your support folks to surface that information. We're seeing that applied to other organizations and their knowledge bases.

69
00:11:11.310 --> 00:11:16.969
Philip Sellers: You know, organizations that are based around research. They're private research research repositories.

70
00:11:17.398 --> 00:11:24.849
Philip Sellers: Yeah, I think we're we're far cry from days of it, solving health conditions for us, because there's a lot of other

71
00:11:24.910 --> 00:11:47.772
Philip Sellers: implications to malpractice and other things like that. But even in the healthcare space you could put it at, maybe billing or coding problems in your billing system. And you know, there's there's all sorts of other things that. You can ask it, and and start to draw out conclusions or information and point to information. So

72
00:11:48.230 --> 00:11:52.539
Philip Sellers: again having it close to those workloads is a a huge benefit.

73
00:11:52.540 --> 00:11:57.249
Andy Whiteside: So, Philip, I'm the more the layman at this point in this conversation with you guys. But

74
00:11:57.490 --> 00:12:05.189
Andy Whiteside: this seems like the workload of all workloads that Nutanix was built to satisfy, to solve for.

75
00:12:06.220 --> 00:12:13.016
Philip Sellers: I think so, because a again, our point has been for a long time right? Nutanix is great for many workloads.

76
00:12:13.630 --> 00:12:17.330
Philip Sellers: and in a lot of ways. It's great for consolidating

77
00:12:17.630 --> 00:12:27.049
Philip Sellers: all of these different types of workloads into a standardized way of operating. And so this is another great example of bringing that all together, and putting a bow on top.

78
00:12:27.490 --> 00:12:33.529
Jirah Cox: Totally agree. Yeah, I mean, it's it's a workload powered by data that requires great speed. That

79
00:12:33.600 --> 00:12:40.700
Jirah Cox: needs some general hardware right? For running containers need some, some specific discrete hardware, for, like Gpus for inferring

80
00:12:42.080 --> 00:13:02.620
Jirah Cox: and for the most part for most companies, you don't build it once you want to be able to rebuild it 10 times a week, 10 times a day. Right? You need that automation, that Cicd framework around it for saying, Oh, there's a new model of Llm blah, blah, that runs my application. Let's put it in a dev environment. Put it in test environment. Put it in prod environment and roll it through that whole lifecycle fully autobatable.

81
00:13:02.720 --> 00:13:16.010
Jirah Cox: and go back to what we talked about, Phil, with our that in us for inference episode, I think, I said, like none of us would use Chat Gpt, if it took 3 min to answer your question right like speed matters right?

82
00:13:16.913 --> 00:13:24.030
Jirah Cox: So there really is no such thing as too fast. When it comes to that, when we're like having a model just ransack terabytes of data.

83
00:13:24.060 --> 00:13:27.570
Jirah Cox: the faster the better to get better, more high quality answers.

84
00:13:28.360 --> 00:13:29.863
Andy Whiteside: It's amazing how we

85
00:13:30.710 --> 00:13:48.519
Andy Whiteside: I don't know how we, our expectations change. I, for whatever reason, I just used copilot to look at my emails from the day that part changed my my world last week, not because they they taught me anything. They just got me thinking outside the box on what I could use them, for which one of which was, Hey, summarize my calendar for the day

86
00:13:48.520 --> 00:14:04.023
Andy Whiteside: like Get out of bed, or like at the end of the day. Hey, what was my email today? What was the important stuff? And you know, I still need to look at it. But there was. It really did a good job of giving me a synopsis of that I did a minute ago, and it took 5 seconds, and I literally was like, what in the world is wrong with this thing. It's taking forever.

87
00:14:05.635 --> 00:14:06.710
Philip Sellers: Yeah.

88
00:14:06.710 --> 00:14:14.080
Ben Rogers: You know, Andy, that's funny, because I'm gonna age myself. Here. Do y'all remember Clippy? In the Microsoft office.

89
00:14:14.080 --> 00:14:14.400
Philip Sellers: Oh, yeah.

90
00:14:14.400 --> 00:14:24.489
Andy Whiteside: Yes. So I listened to a podcast yesterday Mark Benioff, the the salesforce. CEO founder, was talking about co-pilot from Microsoft is just another clippy, and I'm like none

91
00:14:24.960 --> 00:14:26.496
Andy Whiteside: more valuable than that.

92
00:14:26.880 --> 00:14:40.440
Ben Rogers: So, my boss, at the time, who was a visionary, he said, Ben, can you imagine when you get up and you look in the mirror. Clippy comes up in the corner and goes, hey, this is what you got going on today. This is what's important. This is what you need to pay attention to

93
00:14:40.440 --> 00:15:05.359
Ben Rogers: when you said you had it. Recap your calendar. I can just see the little clippy coming up going, hey, Andy, it's bedtime. Here's what's important for you today. Here's what's going to be important for you tomorrow. So that's how I kind of see AI really getting into our lifestyle where it is truly a digital assistant, and we'll remind you and bug you and let you know that these are the things that you need to be paying attention to in your

94
00:15:05.360 --> 00:15:06.290
Ben Rogers: daily life.

95
00:15:06.290 --> 00:15:06.940
Andy Whiteside: Yeah.

96
00:15:07.370 --> 00:15:10.996
Philip Sellers: He sees Ben reach for the comb and goes. Oh, it's a customer meeting day, is it?

97
00:15:11.496 --> 00:15:17.229
Philip Sellers: You know, if it could get me into the right t-shirt for the right vendor for the right day.

98
00:15:17.230 --> 00:15:19.060
Philip Sellers: Right help, too. Right, you know. So.

99
00:15:19.060 --> 00:15:22.520
Jirah Cox: Oh, totally! I could walk in my closet and go like internal calls or customer calls today

100
00:15:23.250 --> 00:15:24.230
Jirah Cox: like the polo shirt.

101
00:15:24.230 --> 00:15:24.890
Ben Rogers: Right? That's right.

102
00:15:24.890 --> 00:15:49.237
Philip Sellers: But I mean, you know, this is the ultimate assistive technology, right? We, we have a lot of things built into our modern operating systems or phones to help with accessibility. But this is ex assistance. This is true digital assistance at its core. And you know, AI comes in all different places, but it really is around making us more productive.

103
00:15:49.650 --> 00:15:55.637
Philip Sellers: And I'm weird the way I said that more productive at the end of the day

104
00:15:55.990 --> 00:16:19.700
Ben Rogers: There is a back back end platform that this AI engine has to run on, and that's where Nutanix has hit the nail on the head of being able to provide that platform. And in a lot of cases with existing customers, with things they have today. And so that's what I'm excited about. Is this AI thing you know the front end of? It's cool. But, man, how you deliver that as a platform is even cooler to me sometimes.

105
00:16:19.700 --> 00:16:38.781
Philip Sellers: Well, well, let's think about that. I mean, it's a great call out to Ben, like, what is the front end for this in a lot of places. It may be somewhere you already interact in your organization. Maybe it's in Microsoft teams. Maybe it's in slack that may ultimately be the front end talking to an Api that's hosted back in the Nutanix.

106
00:16:39.080 --> 00:16:51.610
Philip Sellers: We may not be talking about going to a chat gpt like web interface. We may be using this in our daily productivity tools. I think that's even more powerful when you can bring something like this to the

107
00:16:51.660 --> 00:16:54.889
Philip Sellers: day to day of what someone's already consuming.

108
00:16:55.300 --> 00:17:01.009
Ben Rogers: Man. We have a customer that is re-engineering their production floors.

109
00:17:01.020 --> 00:17:25.220
Ben Rogers: and they've actually gone and looked at the footprint of their production devices. And they're able in this software to like move the whole floor around. So before they even move a lick of equipment, they've got the square footage of the building all the components in that building. And they're actually using an AI model to see, where is it most efficient to put this stuff in this line building. So, man, it's happening today. Dude, it's cool stuff.

110
00:17:25.220 --> 00:17:31.710
Ben Rogers: really cool stuff. I've heard AI be considered to what electricity has done for the human man.

111
00:17:31.710 --> 00:17:40.459
Ben Rogers: You know, human being, that this is the next. You know, big wave of technology, that face and the things I've seen. And I'm I'm not a doubter at this point.

112
00:17:41.640 --> 00:17:45.510
Philip Sellers: I guess ultimately it's gonna be a time will tell sort of story.

113
00:17:46.040 --> 00:18:10.231
Philip Sellers: Yeah, there, there is more, though, that we have from Nutanix or from Microsoft ignite and so I don't wanna shortchange. The other 2 big announcements that we were talking about next up is is databases, and I love this topic because Ndb, I think, doesn't get nearly enough spotlight. So I love that we're talking more and more about nutanix database

114
00:18:11.200 --> 00:18:34.739
Philip Sellers: for a lot of people listening. Nutanix database, even though we've talked about it before, may still seem a little weird. It's not a core infrastructure. It's not like vmware we're talking about databases. We're talking about middleware places where data resides. And so I get the feeling that a lot of people find it odd that Nutanix is playing in the data space. But, Jaira, why is that.

115
00:18:35.300 --> 00:18:41.200
Jirah Cox: It's because we're a cloud platform and cloud platforms have advanced services right? Like it shouldn't just be a matter of

116
00:18:41.360 --> 00:18:49.910
Jirah Cox: humans. Go build vms. And then other humans put database engines on those manually forever. Right? Faster time of value, more automation services.

117
00:18:50.336 --> 00:19:16.919
Jirah Cox: Is the value of cloud, right? Or of any kind of mature cloud platform, including ours. So yeah, so ndb, which is our nutanix database service in its core, right? Helps our platform owners to to create inventory power outage over here. But ups is going strong. They helps our platform owners to offer their customers right database as a service anywhere that they want to operate.

118
00:19:17.170 --> 00:19:22.420
Jirah Cox: and with ndp. My usual high level summary right is like

119
00:19:22.510 --> 00:19:35.700
Jirah Cox: helps you provision faster, protect those databases faster, patch those databases faster, and recover whenever you need to recover from like a human oopsie, or from a kind of data, loss, event, or malware, or anything right?

120
00:19:35.820 --> 00:20:05.099
Jirah Cox: So all the kind of things that are like. I said I'd call back to more complex operations that we make more accessible right? I can click a button and patch my database without being an expert in doing that, because I can just have my database. Follow my organization's guidelines and blueprints and recipes for hey, we're moving to this version of SQL. In this patch release click here to have your database align with that declarative standard right? And it's just easier for me as now. Maybe an app owner

121
00:20:05.100 --> 00:20:30.099
Jirah Cox: to get in line with that right without needing to go bug a dba for something like patching right? So the cool part here now is that Ndb, our database service can now provision onto Nc. 2 and azure. So when you have your hybrid cloud stretching into azure powered by Nutanix, that now also can be a super valid deployment target for Ndb. And can also attach to flow virtual networking. So the same way I can stretch networks now

122
00:20:30.100 --> 00:20:44.080
Jirah Cox: from on-prem a lot of boot camp. For this last week we took customers, and in about 90 min we went from on-prem static IP address to moving those vms across to another cluster that didn't have that same Vlan available to it.

123
00:20:44.180 --> 00:20:53.559
Jirah Cox: Static IP just worked so like we could. We could move your vms today, if you wanted to, into cloud, get you there faster. But now also not lose that governance model with Ndb.

124
00:20:54.950 --> 00:21:01.320
Philip Sellers: Well, 2 things that I love about Ndb as a platform really come in play. When we're talking about app dev

125
00:21:01.400 --> 00:21:17.799
Philip Sellers: combined with Ndb, and it's copy data. So you can order up things through the Api or through a user interface. If you're an app developer and you need a copy of a database. Well, now, that's at fingertips. No tickets. It's no

126
00:21:17.910 --> 00:21:24.660
Philip Sellers: waiting. It's near instantaneous, because you're tapping into all the goodness of the underlying storage subsystem right?

127
00:21:24.842 --> 00:21:25.570
Ben Rogers: In the West baby.

128
00:21:25.570 --> 00:21:28.170
Philip Sellers: You know, Aos for the win.

129
00:21:28.370 --> 00:21:49.410
Philip Sellers: and then the other thing is is being able to then take that and extend that into automation. So app dev platforms, pipelines, automated testing. You know, data has to be a part of the total application landscape. Now, when you're testing those things out and doing things. You can do that automatically. With the

130
00:21:49.801 --> 00:22:00.759
Philip Sellers: pipeline integrations against Ndb. So I I love to call those things out because I came from a large at dev before joining integra. And so

131
00:22:00.760 --> 00:22:11.419
Philip Sellers: data is a tough layer to handle. You know, Ben, I know you get really excited talking to customers about Ndb. You know what what sorts of things

132
00:22:11.500 --> 00:22:13.600
Philip Sellers: resonate most with them.

133
00:22:13.910 --> 00:22:25.909
Ben Rogers: Well, so the one thing I would like to go to before I answer that is man. I think the way Jaira described Ndb was very eloquent. I'm a little more, you know, core, and

134
00:22:25.930 --> 00:22:40.819
Ben Rogers: our management of data is really kind of where Ndb bubbled out of. We understand the zeros and ones. So you're you're mentioning of, you know, test dev app deployment, being able to snapshot the database real quickly and then anonymize it.

135
00:22:40.820 --> 00:22:59.990
Ben Rogers: All of those things are really what bring that to the table. But they all come down to Aos, and our ability to manage the zeros and ones, and where the zeros and ones need to reside and where they need to be. And so that's where I see customers getting excited in 2 things, and I'll go back to. You know, our Nkp product.

136
00:22:59.990 --> 00:23:07.030
Ben Rogers: The reason that the Nkp people love us is that we brought a storage platform that they can now leverage.

137
00:23:07.030 --> 00:23:31.600
Ben Rogers: and that's where customers are really digging. Ndb. Is that they now can leverage the same goodness they get with Aos. But now they can do that in a database sense and the ability to patch the databases, the ability to look across your SQL landscape and go. Okay, I've got multiple versions of SQL out there. I need to make sure they're all patched at a level so that I can come back to a compliance standpoint

138
00:23:31.600 --> 00:23:49.750
Ben Rogers: and feedback to my company, that we're compliant for security reasons, Yada, Yada Yada. So again, man, understanding where the zeros and ones are being able to deliver those zeros and ones in the way our customers need, and then being able to manage. That is really, I think, what the power of the database engine is for us. Ndb.

139
00:23:49.940 --> 00:23:59.239
Jirah Cox: Well, and we're in the business of helping customers get where they want to go faster. Right? So in your stuff is, call it, you know, all in in one place, all within your 4 walls today.

140
00:23:59.300 --> 00:24:23.529
Jirah Cox: But you want to get to cloud as fast as reasonably possible. This helps you can get. This helps you get those databases there way faster, right like I can. Now take those databases running on Nutanix. Take a snapshot, make that instantly available over in azure or even use both. Right? I can have one data center. That's for test dev, one that's for production. Maybe I do production in the cloud test dev on prem, either. These are totally valid in my data. Landscape now stretches wherever I go.

141
00:24:24.570 --> 00:24:34.710
Philip Sellers: Yeah. And and to me, like, you know, when when I think about the nutanix platform, Lcm in particular comes out, you know lifecycle management. When Nutanix is in

142
00:24:34.760 --> 00:24:39.960
Philip Sellers: control, you're getting the best practices. You get the correct order of operation.

143
00:24:40.230 --> 00:25:07.580
Philip Sellers: You're making it accessible back to that flag we planted earlier in the podcast you're making difficult things accessible. And I think Ndb is another one of those. Now with a generalist team, not a team of highly skilled expert database administrators. You're able to get outcomes where their best practices they're going to pass audits things like that in addition. How accurate am I on that.

144
00:25:07.850 --> 00:25:14.520
Jirah Cox: Totally. Totally. I mean not just. I mean, we can keep it kind of sequel themed, because this is the ignite recap.

145
00:25:14.780 --> 00:25:29.149
Jirah Cox: But think about how many customers run 90% SQL. But a little bit of other things right? Some postgres, or some mongo or some oracle, and so that standardization of like look. The habit is, go to Ndb, click the patch button and know that it's going to do what you need to do in the back end, whether it's

146
00:25:29.240 --> 00:25:52.840
Jirah Cox: SQL. Single instance, or Ag. Or Fc. Or the Postgres, or the Maria, or the Oracle, or whatever like my team's in the habit of. I go to the dashboard click, patch right like that's pretty high value of like abstraction, not having to go like crack open the manual. And how do I do this one again just because it's a 1 off. I get that that simplicity of I do everything. Quote the same way because I do it all through Ndb

147
00:25:53.070 --> 00:26:05.589
Jirah Cox: or recover the same, or patch the same, or replicate the same. Because I can just move this data around wherever I choose to. So yeah, I think it's a pretty apt analogy to kind of compare it to Lcm of like, let it do the right thing for you the same way every time.

148
00:26:06.070 --> 00:26:06.630
Philip Sellers: Yeah.

149
00:26:07.860 --> 00:26:21.229
Philip Sellers: you know, Andy, you you say you're the novice in the room today. But as we're talking about Ndb, I mean, like you've been around databases from the very early days, you know. Every single citrix app.

150
00:26:21.230 --> 00:26:21.920
Jirah Cox: Ouch!

151
00:26:21.920 --> 00:26:28.570
Philip Sellers: Installation. We've we've had databases as a persistent part of our world, right?

152
00:26:28.930 --> 00:26:34.839
Philip Sellers: Making that easier. I mean, how much does that sound like the amazing thing.

153
00:26:34.840 --> 00:26:55.429
Andy Whiteside: If you'd have told me before this AI thing happened. What's the best workload for this extensible in the cloud on Prem, in a partner's data center. You know this thing that could go anywhere you wanted to. Aka hyperconverged the thing that Nutanix, you know, created for all of us. I would have told you that a database.

154
00:26:55.430 --> 00:26:55.810
Philip Sellers: Isn't.

155
00:26:55.810 --> 00:26:58.450
Andy Whiteside: Other workload that makes the most sense.

156
00:26:58.450 --> 00:26:58.910
Philip Sellers: Yeah.

157
00:26:58.910 --> 00:27:08.030
Andy Whiteside: And that's coming from the Vdi guy that's like, hey? If you don't, if you run Vdi, and you don't run it on a hyper converged solution like Nutanix. Well, you're totally missing the mark. Database.

158
00:27:08.270 --> 00:27:13.020
Andy Whiteside: you know. Once that became a reality in the nutanix world is even a better workload.

159
00:27:14.010 --> 00:27:14.690
Philip Sellers: Yeah, yeah.

160
00:27:14.690 --> 00:27:24.759
Jirah Cox: I was in a briefing last week and learned that something like 90% of all customers actually run databases on our platform. It actually probably is eclipsing Vdi as our our most applied workload.

161
00:27:26.100 --> 00:27:31.382
Philip Sellers: And I can believe that I mean data rules the world, right? I mean, everyone has this need. And

162
00:27:31.730 --> 00:27:37.749
Philip Sellers: you know, we go there to prove out what's a high performant platform

163
00:27:38.194 --> 00:27:48.515
Philip Sellers: we have tons and tons of tests around it, and that's it's it's an easy way to prove out that you can you? You've got the real chops. You got the real deal for infrastructure.

164
00:27:49.060 --> 00:27:50.770
Philip Sellers: It's a great, great call out,

165
00:27:52.140 --> 00:28:02.620
Philip Sellers: But that's not all that we have. We also have some more. Nc, 2 news, and so we'll keep going close out the blog post from Microsoft ignite. But

166
00:28:03.130 --> 00:28:07.349
Philip Sellers: continuing innovation around Nc 2, Jaira, I'm gonna throw it to you again.

167
00:28:07.570 --> 00:28:26.700
Jirah Cox: So some good quality of life improvements here for Z. 2 and azure, which is how we can get bare metal nodes out of the azure cloud. Use them to build a new Tnx cluster, to run our customers, workloads. It runs just like any other new Phoenix cluster either in a service provider environment like integra, or in your own data center

168
00:28:26.860 --> 00:28:42.360
Jirah Cox: behind your own firewall. You can now run Nutanix, of course, and you've been able to for years run it in the azure environment. Now, with an enhancement to be able to support user defined routing. So now, like even more of those kind of customizable. I need this traffic to go to certain kind of ways of custom. Routing

169
00:28:43.070 --> 00:28:50.749
Jirah Cox: makes it more useful for virtual appliances as well. When I want certain things to get hair pinned, or certain kinds of of inspection to occur on that traffic.

170
00:28:50.810 --> 00:28:59.470
Jirah Cox: and then kind of the usual, what's new with Nc. 2. Now I can run it more places, right? So enhancing availability in Uae and also India Central

171
00:28:59.480 --> 00:29:07.299
Jirah Cox: as well. What's cool about Nc. 2, right is when you run nutanix on Prem, you basically buy servers, you buy licenses, you deploy it all

172
00:29:07.390 --> 00:29:11.289
Jirah Cox: in Nc. 2. You actually can blend some of that. Maybe I own

173
00:29:11.340 --> 00:29:16.320
Jirah Cox: a pilot light cluster. I do a multi-year commit and use part of my Mac and all that good stuff

174
00:29:16.350 --> 00:29:21.529
Jirah Cox: for building that environment and using it against my traditional committed funds.

175
00:29:21.690 --> 00:29:34.309
Jirah Cox: Maybe when I want to have a want, want is the wrong word. When I have a declared disaster for Dr. I can get additional nodes in more of a pay as you go by the hour model. Right? So I get to rent part of my Dr. Environment only when I need to

176
00:29:34.400 --> 00:29:40.450
Jirah Cox: and go back to minimally viable after the disaster is over. So I just have some replication targets set up there as well.

177
00:29:41.040 --> 00:29:42.779
Jirah Cox: So some really good flexibility. There.

178
00:29:43.680 --> 00:30:03.696
Philip Sellers: Yeah. And and I will add, I mean that Nc, 2 is a great place to run these workloads. And it's not the only place. There's also service providers like integra gotta throw a plug in there, too, right? And we we offer very similar things, and and it gives me a good opportunity to kind of share and brag on the team. We had a

179
00:30:04.100 --> 00:30:21.969
Philip Sellers: an unplanned outage like you said not something that you ever really want. But it came in as an email and email request came in and and I saw the email. And it says, this is not a test. And so you know it's bad when the the customers coming to you going, this is not a test. This is, this is not a drill.

180
00:30:22.950 --> 00:30:32.420
Philip Sellers: And in 20 min we had everything failed over and their users logging back into Citrix. The technology is extremely powerful

181
00:30:32.918 --> 00:30:41.160
Philip Sellers: being able to do that and map out and test full Dr. Recovery, whether that's Nc. 2 or managed service provider like us.

182
00:30:41.440 --> 00:30:58.219
Philip Sellers: the underlying Nutanix provides us with an incredible platform to orchestrate that, and to have a successful outcome. And so, you know, these are great use cases again, where you don't necessarily have to make that capital investment in a second location.

183
00:30:58.410 --> 00:31:21.840
Jirah Cox: I mean, it's to call back. It's that 3rd area of helping people get a complex outcome done more simply and faster. I won't say fastly, I'm going to put that out there because it's like I want to use cloud for Dr. I want to move my production to cloud faster. And then we're like, let's go right like we can move your vms today if you want to. We can help you use Cloud as a Dr. Target, which is

184
00:31:21.840 --> 00:31:38.359
Jirah Cox: one of the most common desires I see customers starting with, for, like what I want to do with Cloud, let's use it as a dear target. Get rid of that second data center, move that into a cloud environment and we're here for it. Right? We help make that happen more fastly. I'll own that, and and more simply as well.

185
00:31:39.230 --> 00:31:46.100
Ben Rogers: See more customers leveraging Nc. 2 for Robo. So I mean, I'm working with a customer in Asia

186
00:31:46.100 --> 00:32:11.859
Ben Rogers: that they're looking at moving their remote sites that are in the Asia pack into in C 2 clusters in azure, because we're building out more in that environment. And it makes sense. And you know, they're they're reducing their real estate. They're going to more of a subscription type model, which is where the company obviously want to go, and we're able to deliver next to where the workloads are and their users are. So that's all. Good wins for the customer and for Nutanix and for azure.

187
00:32:12.420 --> 00:32:15.513
Philip Sellers: Well, you know, I having past experience offshoring

188
00:32:16.210 --> 00:32:23.090
Philip Sellers: it. It may be a very viable use case where you will never have hands in Asia pack.

189
00:32:23.110 --> 00:32:34.240
Philip Sellers: You will never have people on the ground. But you have constituents, consultants, 3rd party contractors. You're delivering services to. You need to be in their region.

190
00:32:34.290 --> 00:32:40.840
Philip Sellers: It is an extremely valid cloud. Use case to extend your geography.

191
00:32:41.310 --> 00:32:48.329
Philip Sellers: Let someone else handle everything below that. Nc 2 layer, and then you handle the things that matter on top

192
00:32:49.006 --> 00:32:54.580
Philip Sellers: where you can arguably provide more distinguished value to your company.

193
00:32:54.840 --> 00:33:14.479
Jirah Cox: I think I've said it at some point on the on the podcast, right, like, in my opinion, running nutanix on a hardware as a service platform really like completes the initial vision of Nutanix. Right? We already assume that hardware in a long enough timeframe is gonna fail. We don't tightly couple to anyone hardware node, we can fail forward, we can rebuild.

194
00:33:14.640 --> 00:33:38.370
Jirah Cox: we do software, defined self healing in place. So when I can get a new node as a service. Right? I don't care about that, failed dim that failed. Power supply that that downed motherboard. I give that node back to the Cloud provider. Now it's their problem. I get a new healthy one, and that's now part of my cluster, and I'm back to normal right? I don't go get a new tire. A new tire just appears for me. I put it on the car and keep on driving.

195
00:33:38.370 --> 00:33:58.890
Jirah Cox: The analogy breaks down. Sorry, but same thing with patching. Right? I don't do bios updates in the cloud. Right? I go. Request a new prepatched node from the cloud provider that already has their new. Whatever new recipe for hardware success prepatched at the time, cluster and inject an older one that I don't need to go, patch. It's not mine anymore. So it's really a really cool paradigm in a way that

196
00:33:59.250 --> 00:34:06.810
Jirah Cox: you know, owning nutanix on-prem running tennis service provider doesn't quite get me as abstracted from a look. It's just hardware as a service.

197
00:34:07.610 --> 00:34:14.899
Ben Rogers: So I'm gonna give a call out to my boss here, Tom Powell, he drills into our heads that we're a software company.

198
00:34:14.900 --> 00:34:39.559
Ben Rogers: We're not a hardware company. And sometimes this, he's look back at him. Go, yeah, we're a software company that's dependent upon hardware. With the Nc. 2 story. He is absolutely correct. This is us at software at its finest, with infrastructure, hardware being provided to it that we really might have a knowledge of or might not have a knowledge of, but again to his point. And if he's listening to this, I hope he has a grin. This does define Nutanix as truly a software company.

199
00:34:42.260 --> 00:34:57.460
Philip Sellers: Yeah, well, and and you're breaking that dependency right? Everything else we've talked about, you know, lives directly on the Nutanix platform. But there are lots of things you guys are working on, you know. Files now can run natively in cloud.

200
00:34:57.480 --> 00:35:06.349
Philip Sellers: You get all the same. Goodness. So you're starting to see a paradigm shift within the company, too, that you're getting more and more native services that are closer to cloud

201
00:35:06.906 --> 00:35:20.099
Philip Sellers: but arguably, Aos for the win right. And we said it earlier. A lot of this innovation is powered by that core storage subsystem. So yes, you are really a storage company under the covers. But

202
00:35:20.270 --> 00:35:29.210
Philip Sellers: as a cloud company to Jairus Point earlier. It's all about the services that you can deliver, and you're delivering the services that people need. Whether that's

203
00:35:29.270 --> 00:35:31.349
Philip Sellers: nutanix enterprise. AI,

204
00:35:31.640 --> 00:35:45.359
Philip Sellers: whether that's databases disaster recovery as a service across the board. You're giving lots of choice. Nkp, I heard a shout out for earlier. You've got all the high level services that a customer is looking for

205
00:35:47.300 --> 00:35:50.470
Philip Sellers: any closing remarks before we wrap it up today.

206
00:35:52.060 --> 00:35:54.490
Ben Rogers: It's a good time to work for Nutanix. Man, we're busy.

207
00:35:54.730 --> 00:36:00.129
Ben Rogers: The the market is is is doing good for us. So, man, we're excited about the future. Here.

208
00:36:01.350 --> 00:36:08.749
Jirah Cox: If we missed you at ignite. Sorry about that. Call your intense account team. Call yours integra account team for a personalized recap.

209
00:36:08.750 --> 00:36:09.570
Ben Rogers: Yeah, no doubt.

210
00:36:09.570 --> 00:36:11.970
Philip Sellers: I'm I'm just gonna highlight that.

211
00:36:12.400 --> 00:36:16.889
Andy Whiteside: You know, Nutanix having what they have, being where they are at this moment in time.

212
00:36:17.150 --> 00:36:19.939
Andy Whiteside: is, you know, a lot of dedication, hard work

213
00:36:19.990 --> 00:36:38.340
Andy Whiteside: fortuitous timing with the idea that the guys over at Vmware we're not doing this in the hyper converged motion in the one foot in the cloud, one foot into the data center, one foot in the semi-private data center, or wherever you want to put it with the software driven approach. I mean, it's all just kind of working out

214
00:36:38.490 --> 00:36:42.280
Andy Whiteside: to be at the right time, right place, the right motions, right investments.

215
00:36:42.280 --> 00:36:53.580
Philip Sellers: Yeah. And and I like to point out as we're doing lunch and learns and stuff around the country right now, there's differentiated value here, you know, there there are places Nutanix playing that those other guys aren't

216
00:36:53.640 --> 00:36:57.539
Philip Sellers: vmware has been really happy in the infrastructure stack. But

217
00:36:57.580 --> 00:37:04.380
Philip Sellers: things like database and you know some of the the differentiated value directing cloud.

218
00:37:04.752 --> 00:37:17.340
Philip Sellers: They're not playing in that space today, and I I don't see it anywhere in the roadmap right now. So there are places where you guys are ahead, too. So that that's also cool to see. And it's 1 of the reasons why

219
00:37:17.460 --> 00:37:20.589
Philip Sellers: we we like to go to market with our Nutanix partner.

220
00:37:20.820 --> 00:37:29.599
Jirah Cox: I think it's only gonna get going to get more fun. I mean, thanks to you guys, thanks to our customers, you guys have helped us get where we are not just financially, but also really through feedback right? Like half of our

221
00:37:29.610 --> 00:37:36.959
Jirah Cox: half our product announcements come from customer direct requests and and feedback. So it's all super important to us. But

222
00:37:37.000 --> 00:37:42.449
Jirah Cox: if we're all dusting off our crystal balls, I think in the next, I don't know. Call it. 1218, 24 months.

223
00:37:42.940 --> 00:37:57.730
Jirah Cox: We're comparing it now in this theme of like, basically a 2 horse race. I think it's gonna get absolutely silly. Right? It's gonna be a 10 horse race like everyone, and their uncle is launching a hypervisor. What? Standby for an announcement for a new gyro visor or something like that like it's they're everywhere, right?

224
00:37:57.964 --> 00:38:00.305
Jirah Cox: Wanna buy into that one. Can I buy into that?

225
00:38:01.350 --> 00:38:05.519
Jirah Cox: The I don't know no features. It just comes with barbecue sauce, or something.

226
00:38:05.520 --> 00:38:07.989
Andy Whiteside: I just want to see the logo.

227
00:38:08.210 --> 00:38:15.659
Jirah Cox: The chaos. I think that we're going to see in the industry around. There's so much of this fracturing ecosystem.

228
00:38:15.740 --> 00:38:26.410
Jirah Cox: I think that the light of like a valuable differentiated solution is only only going to get brighter right? It's gonna be even becoming more of a standout around. Lots of other people like.

229
00:38:26.640 --> 00:38:31.539
Jirah Cox: you know, we're here talking about a platform here that has roots that go back 10 to 15 years.

230
00:38:31.600 --> 00:38:36.019
Jirah Cox: If we're starting a new platform today in 2024, and we're like, look, it's a new hypervisor.

231
00:38:36.900 --> 00:38:39.190
Jirah Cox: Good luck! Best of luck!

232
00:38:39.320 --> 00:38:57.829
Ben Rogers: Well, and I also think that man Kubernetes and containers, the doctor containers, and all that that's going to change. You know, companies are going to have to embrace that. Some companies are already on that model. Some companies are resisting that a little bit. But AI is going to push that container environment. And you're already starting to see that happen.

233
00:38:58.350 --> 00:39:02.059
Philip Sellers: Well, and and let's go back to the flag that gyra planted at the beginning.

234
00:39:02.110 --> 00:39:09.910
Philip Sellers: It needs to be more accessible. That's the thing with Kubernetes. It's inaccessible for most places

235
00:39:10.260 --> 00:39:21.769
Philip Sellers: in the packaged formats that it's existed to date. And so it's good to have you guys partnering to make it more accessible, because that's a very needed valid need in the marketplace.

236
00:39:22.463 --> 00:39:25.029
Philip Sellers: You! I've tried it with

237
00:39:25.940 --> 00:39:35.230
Philip Sellers: 6 different platforms, hands on trying to figure out, how do we? We make it happen, and there are arguably

238
00:39:35.250 --> 00:39:59.629
Philip Sellers: 2 to 3 that have a complete vision when it comes to to making the full Kubernetes stack work. But it's incredibly different. Taking a team through this, it takes years of retraining, getting used to the new way and paradigms of deploying and managing the infrastructure. It's not easy, and so I never go into a customer conversation and understate

239
00:40:00.020 --> 00:40:02.499
Philip Sellers: how difficult it is to turn that curve.

240
00:40:02.500 --> 00:40:11.730
Ben Rogers: Yeah, like, I'll give you an example. I just went through some of our in AI training, and words mean different things in different platforms like even cluster.

241
00:40:11.830 --> 00:40:32.739
Ben Rogers: Are we talking about a new traditional nutanix cluster? Are we talking about an Nkp cluster? And so some words in the environment can mean different things. So you're absolutely correct. But that is one thing that Nutanix does good is, it makes sense of these difficult topics and difficult configurations. And so I think that's where we're going to win. And customers are starting to really see that.

242
00:40:33.140 --> 00:40:36.540
Philip Sellers: I I thought you were. Gonna say, cluster meant something else.

243
00:40:37.260 --> 00:40:39.520
Philip Sellers: Amazing namespace pollution.

244
00:40:39.842 --> 00:41:06.590
Philip Sellers: Well, on behalf of this integra team, and, thanks to Jira and Ben joining us from Nutanix. I wanna say, thanks so much for joining us on this episode of the podcast you know, if you've got a partner out there that's not adding value to your world. Give us a chance, give us a call, and we would love to work with you and partner with you, and and see what we can do to add value to your world

245
00:41:06.800 --> 00:41:15.849
Philip Sellers: on behalf of Andy, and everyone else want to say thanks for joining us, giving us your time, and we will catch you on the next episode of and excuse me, Nutanix weekly.