AWS needs a new CEO now that Andy Jassy has been tapped to run the Amazon mothership. Does the cloud giant need a high-powered, Type-A successor to drive AWS’s growth, or is it now so dominant that a ham sandwich could run the company?
That’s the topic for discussion on today’s Day Two Cloud podcast. Co-host Ned Bellavance moderates an intramural debate between two Packet Pushers: Ethan Banks and Drew Conry-Murray.
Ethan takes the position that AWS’s domination is not at all assured, which means the succession matters a great deal. Drew argues for the ham sandwich.
Let the conversation (polite disagreement? verbal knife fight?) commence.
Sponsor: Onix
As an award-winning cloud solutions provider, Onix provides consulting services for cloud infrastructure, collaboration, devices, enterprise search and geospatial technology. For a limited time, Onix is offering your organization a FREE 6 Hour Cloud Data Strategy Workshop (normally valued at over $2,000). For more information on this special offer, visit onixnet.com/packetpushers.
Tech Bytes: ThousandEyes
Stay tuned for a Tech Bytes conversation with sponsor ThousandEyes on VPN monitoring for remote workforces. Our guest is Alex Cruz Farmer, Principal Product Manager at ThousandEyes. To learn more about the solution, check out the on-demand webinar “VPN Monitoring for a Remote Workforce” or sign up for a free trial at thousandeyes.com.
Show Links:
‘A managerial Mephistopheles’: inside the mind of Jeff Bezos – The Guardian
The South African origins of Andy Jassy’s profitable Amazon division – Quartz Africa
Battle For Cloud Dominance – Who Comes After Jassy At AWS? – Forbes
What Andy Jassy’s promotion to Amazon CEO could mean for AWS – TechCrunch
5 Things To Know About Andy Jassy’s Transition To Amazon CEO – CRN
Show Transcript:
[00:00:01.030] – Ethan[AD] Day Two Cloud sponsor Onix is a premier Google cloud partner and an AWS advanced consulting partner they’re also an 11 time Google Cloud Award winner and was recently recognized as tech titans number one for best cloud consulting services in 2020. For a limited time. Onix is offering a free six hour cloud data strategy workshop. That’s kind of a big deal. That’s normally a two thousand dollar service. For more information and to find out more about the six hour Cloud Data Strategy Workshop, visit Onixnet.com/PacketPushers. [/AD] [00:00:40.380] – Ned
Welcome to Day Two Cloud, the topic for today’s episode is going to be, hey, AWS just lost their CEO. What’s going to happen there? I mean, he isn’t gone but he is moving on and they need to replace that void with something else. Does it need to be another high powered type person an A type of personality, or could they just swap it out with a ham sandwich? And to answer that question, I’d like to bring onto the show tWo humans I mostly admire. [00:01:08.610] We’ve got Ethan Banks and Drew Conry-Murray from the Packet pushers, both joining me to talk through this issue. So we’re going to do that. Stay tuned. After the episode, we are going to be also talking to Thousand Eyes about remote VPN monitoring in a 15 minute tech bite extravaganza. So that will be after this segment. So let’s get into it. Drew, you brought this topic to my attention at first, so why don’t you lay out your thought process and overall thesis? I’m personally undecided. I have no idea what’s going to happen. [00:01:42.630] – Drew
Yeah, so I did bring it to you guys. So if this podcast goes sideways, it’s entirely my fault. And I apologize to the listeners, but there’s a lot of hoopla around Andy Jassy’s ascension from AWS to being CEO of the overall Amazon business. And of course, that means there is a succession question in AWS. And so my thought was AWS seems like it’s been set up pretty solidly. And that pretty much as you said, Ned, a ham sandwich could step into the CEO role and things would generally be fine. And now I’m here to prove that point or at least try to prove it. [00:02:12.030] – Ned
Ethan, are you thinking something along the same lines or do you have a contrary opinion to what Drew was positing? [00:02:18.900] – Ethan
Somewhat contrary. I think I understand that perspective from Drew, but my take is that AWS is in a situation where they stand to lose. They are in a market position of of leadership for sure. Right. Huge market share. They’re the gorilla in public cloud. But if you look at what’s been happening as enterprises especially have been adopting public cloud, I think they are losing customers. And despite their many billions of revenue each year, it’s not just a ham sandwich can walk in and things are just going to continue on as they have been the next five to 10 years for Amazon are going to require a leader that gets what enterprises are going through and can figure out clever ways to capture their revenue for the long haul. [00:03:02.550] – Ned
So you’re saying they need more of a ham sandwich with Brie? That would really, and maybe some pears? That would really level the playing field. Obviously, that’s preposterous. I’m glad you have a contrary opinion. Otherwise, this would be a very boring episode of us just agreeing with each other. So, Drew, what do you think AWS has going for it now? That could continue on regardless of who’s sitting in the CEO chair. [00:03:27.330] – Drew
First thing, is AWS really they had first mover advantage. They essentially created the IaaS market. They built the basic constructs EC2, S3, Route53, all those key elements to spin up and use compute in the cloud. [00:03:40.920] We’ve seen competitors emerge, but AWS being the first mover, gets them out in front, sort of helps them define the market, attract the first customers and so on. [00:03:48.750] They’re also they designed a company to be sticky in several key ways. First, they captured the developers. Right. They make it easy for developers to write applications on the cloud. You don’t even have to talk to a boss. Just use a credit card. You’ve got to compute instance. Bam, you’re up and running. Once you start building an application in the cloud and data starts hitting that cloud, then the data gravity takes over and that sort of turns into a perpetual motion machine where the more data you’re putting into that cloud, the stickier the cloud becomes. [00:04:13.500] And then it’s just self reinforcing for AWS. [00:04:16.110] – Ethan
Sort of. So my take on that is sort of as in everything you said is exactly true. But so one of the big one of the big buts, I couldn’t phrase that worse. [00:04:26.970] Anyway, one of the big buts is that we there are things that are showing up now in common IT speak that belie that. That’s everything that’s going on. One of them is cloud repatriation. I put something in the cloud. It’s so bloody expensive that I’m now taking it back out of the cloud and I’m going to host it either somewhere else or just bring it back in internally. Let’s just focus on that for a minute. That being a thing is due to the fact that enterprises are realizing the lift and shift approach isn’t working out. [00:04:59.790] If I take my workload and just basically replicate up in the cloud what I’ve been doing in my data center for the last decades, that doesn’t work. It’s not a cost efficient spend. We know this. We’ve had this conversation Ned many times on Day Two Cloud with different folks. The whole idea is don’t think about picking up a workload or just dumping it in the cloud, save money. That doesn’t work. Instead, you need to rethink about how that application is delivered, what that delivery system is. [00:05:24.090] Re-architect that application and that’s the hard part. Re-architect that application to take advantage of cloud native. Some combination of services that are up there in the cloud that allow you to deliver that application cost effectively and. If you do that, then public cloud can give you an economic win, some might argue enterprises don’t care how much they’re spending, but I think a lot of them actually do because the cost of running your workloads in the public cloud is a major topic of discussion, how to be more economical about it, how to be smart about it. [00:05:56.920] We’ve had Corey Quinn on the show and he’s built a business out of that, helping companies optimize their cloud spends. So my take is, yes, it’s easy to say yes, it’s developer friendly and so on. [00:06:10.090] But the costs of doing it and doing it right means it’s not as easy as swiping a credit card. It means having to rearchitect your applications. And that’s that’s a tough thing and something that I feel new Jassy needs to get his or her head around. [00:06:24.790] – Ned
Right. One of the themes of Reinvent last year was really focused on actually reinventing your applications to run on the cloud. That was the primary focus of everything they were doing is saying it’s just an acknowledgment that they can’t lift and shift, enterprises can’t do that and expect magic to happen. They have to reinvent the way they do things. And, oh, here’s all our Whiz-Bang services that allow you to reinvent yourself in the way you do things. I like the messaging. [00:06:53.710] I don’t I’m not convinced enterprises can move in that sort of agile fashion a lot of the time. On the cost side of things, if you think about data centers and the sunk costs you already have in those you don’t see that on an annual basis or a monthly basis, let’s say. But you do see your AWS bill come in on a monthly basis, single wahaha, and you probably haven’t successfully shut down your data center. So now you’re kind of like paying for both. [00:07:18.940] Drew, do you think that they can just coast by on the messaging they already have out there or do they have to get in front of that big truck that’s coming through? [00:07:26.410] – Drew
Well, I think we’re in the stage where the cloud companies are working through that big truck because cloud I remember being originally positioned as low cost because, look, you can get this instance for essentially pennies or fractions of a pennies and wow, OK, that’s amazing. And I don’t have to bear all the costs of running the hardware itself. That’s a win for me. That’s an ROI win for me. But then once you actually start using the services, that’s where the bill shows up. [00:07:47.740] So I think we’ve gotten past the notion that cloud is somehow cheaper and we’ve worked through the disappointment of enterprises realizing that that’s not the case. Ethan to your point about cloud repatriation, that’s not just an AWS problem. That’s a problem for anybody who’s providing cloud services. So Azure, Google, IBM, Oracle, whatever, they’re going to face that same issue. The third point is that AWS is opportunity is now that enterprises have made the mistakes, let’s just try to duplicate what we do on Prem up in the cloud. [00:08:15.250] Is that AWS has rolled out the digital transformation roadmap, which every organization says they want. They’re saying here’s the constructs you need, like a sort of a container and micro services architecture. Just spin up the services you need, find a way to connect them. Here’s the path to go down. Here’s the services we used to do that. Maybe someday you’ll get to the Shangrila of serverless. AWS is showing customers this is the way you’re supposed to do cloud. [00:08:38.200] This is the way cloud should have been done from the beginning. But you came at it with the wrong model. I think we’ve hopefully gone through that pain now. And so maybe we can be looking forward to the next generation with a growing cloud native. So we’ve had those growing pains. [00:08:50.380] – Ethan
I want to argue that that’s a step in the right direction. Lots of documentation, lots of things you can read on the AWS website, whether it’s just simple blog posts, whether from white papers, there’s procedural documents that kind of show you a lot of this stuff, but it needs to go more than that. It needs to go beyond their certification programs. I think they need to have a consulting organization that they build that really gets good at handholding the enterprises and taking them, not just here’s some documentation, here’s some training, figure it out. [00:09:21.220] But really, does that work for them to make that customer sticky for the long term? I know that there’s a zillion AWS partners out there. Right. But those are going to be hit or miss. What your success is with those folks I used to work for not an AWS partner, but a partner for various vendors. And I can tell you the way those things work is, again, hit or miss. It comes down to the ability of the engineer within those partner organizations and what they have experienced with what their training level is, how many other projects they’re working on at the same time that they’re working on yours to help you do that transformation. [00:09:57.010] So my logic is, if AWS builds a consulting firm in House that is going to really hold the hand of those enterprises and get them moved into those architectures, do that digital transformation, Drew, like you were talking about that roadmap that they laid out, what they end up with is a successful cloud migration where that application that they picked up and moved doesn’t get repatriated, doesn’t fail. [00:10:24.160] It gets re-architected on the front end like it needs to, gets moved into the cloud. It’s now an economically viable model. And because they’re using magical AWS special services, then it’s a sticky customer because moving them back out at that point, really. It is hard not just for data gravity reasons, but because of operational reasons and API reliance and things like that, that would just be like move it, it’s cost effective. So I don’t have a reason to move it. [00:10:50.150] It’s fine. And now they’re a customer for life and you’ve got that revenue. And then the next project that comes along, where’s it going? It’s going to AWS. It’s not going multi-cloud. It’s not hybrid cloud. It’s going to be more and more and more AWS. [00:11:03.770] – Ned
I’m curious because one of the things that I was reading when you sent me some of the links that you’d been looking at was that the total market for IT spend is far larger than the spend that’s currently on AWS. You think AWS has now hit a 50 billion dollar annual run rate, which is staggering? That’s that’s pretty amazing. But in the world of tech, it’s actually not that big. It’s about the same size as Cisco. So just let that sink in for a second. [00:11:32.750] And then the other number I saw was, I believe three point six trillion dollars is the pool of IT money that’s out there. So they’re just a fraction of a fraction of what’s available. The pushback on AWS’s growth recently has been their percentage of growth year over year continues to go down. Part of that is the law of large numbers. But part of that is that maybe their growth actually is decelerating. Do they need a new strategy to chop off a bigger portion of that IT spend? [00:12:01.500] – Ethan
[AD] Ned are stopping the show for a moment to talk about today’s sponsor, Onix. Onix is a cloud solutions provider, and they’ve got the credentials and awards that tell you they know what they’re doing for me or Google Cloud partner AWS Advanced Consulting Partner one time Google Cloud Award winner, Tech Times number one for best cloud consulting services in 2020. You’re going. I don’t care, Ethan. What is Onix do? All right. Fair enough. Onix provides consulting services for cloud infrastructure, collaboration devices, enterprise search and geospatial technology. [00:12:31.490] Let’s drill into this. Think about one of the recurring themes of Day Two Cloud. We have said for in many episodes there are companies that do lift and shift of workloads to the cloud and they are sorry they did that. They pay dearly for that approach. They don’t automate well, the companies don’t leverage cloud native services. They don’t consider application delivery design like they should. And why do they have all these problems? Cloud brings a lot of new technologies into an I.T. shop. [00:12:57.290] It is hard to get it all right the first time, even with training, even though you’re smart, even though you’re a very capable engineer, even if someone gives you the luxury of time that most of you, frankly, are not given by your senior management, well, OK. This is where the real value of Onix comes in. Onix is going to help you figure out what you really got. That’s the discovery process. And it matters because you kind of think, you know, everything you’ve got in your infrastructure, that’s gotta move to cloud, but you probably don’t. Onix is then once the discovery is done, going to help you construct a plan up. [00:13:26.570] This is important. How do we move all the stuff we’ve got into the cloud? That plan is going to include a lot of the nitty gritty detail about getting that data move to the right place and stored in the right way and then visualizing what’s actually happening in the cloud and more even stuff that maybe you haven’t considered before. Like machine learning. They have experienced cloud based machine learning nerds on staff at Onix. [00:13:48.620] All right. Another Day Two Cloud topic that comes up a lot, automation. Onix can help you there, too. They have specialists with deep expertize in cloud toolset and automating that tedious stuff that you’ve been screwing up, trying to do it all by hand. Been there. I’m right there with you. Well, Onix can build you a data pipeline, integrate it with the existing tools and databases that you already have, and then bring new functionality to your current data solution. [00:14:12.770] You’re like, I don’t know, that sounds complicated. It maybe it is. But once they help you build all that fancy stuff, they’re going to train you on the solution. So it’s it’s good, right? You’re all good. Pretty cool stuff from Onix. And for a limited time, Onix is offering your organization a free six hour cloud, data strategy workshop that’s a service they would normally sell for like two thousand dollars or more. But it’s free for people who respond to this offer. [00:14:37.460] And if you want more information, visit Onixnet.com/packetpushers. One more time. That’s Onixnet.com/packetpushers. Let me spell Onix so you don’t get confused here. ONIXnet.com/packetpushers. For more information about their six hour cloud data strategy workshop. And now back to today’s episode. [/AD] [00:15:01.690] – Drew
So this goes back to my general position that the next CEO doesn’t really matter that much because AWS is moving from what I think of as an innovation phase to a stewardship phase. So, yes, there’s still things a new leader needs to do, like try to get more enterprise partnerships, build out more better relations, maybe even develop an enterprise sales force to start bringing those customers and getting more of that money into AWS. [00:15:25.600] That’s not an innovation move. That’s a stewardship move. The innovation, the operation model, the approach to technology, figuring out how to make a shared infrastructure, building what you can and buying what you need, and the company culture on how ferocious AWS is as a competitor that’s already in place. So the next CEO just needs to come in, maybe tweak the formula a little bit, but otherwise it’s just cranking that wheel, cranking that wheel. So getting back to your point Ned about there’s a ton of money for AWS out there to get, not to mention all the other cloud competitors. [00:15:53.530] The pie is huge. So AWS has the strategy. Now it’s just time to execute and start bringing those customers in. The pool of candidates you could find who could lead AWS to the next phase is much broader than the pool of people who could actually build AWS from the beginning. [00:16:08.470] – Ethan
But what is differentiating AWS at this point from the other clouds? Most of the other clouds, they’ve caught up in some way or another Azure lots of similar product offerings, at least for the core services, right? GCP the same thing, even Oracle Cloud, same thing. And then there’s all these other clouds or cloud like services that can give you a similar kind of result. So particularly if I’m early in my digital transformation journey, why do I necessarily go to AWS? [00:16:35.080] Just because the most recognizable brand why wouldn’t I look at Azure? Why isn’t Azure necessarily going to be the one that is going to grow the most here coming up? Because a lot of enterprises have that deep Microsoft relationship. Wouldn’t I start there? I think I would. And in fact, we know that many companies are doing that, at least for certain kinds of workloads. [00:16:55.840] – Ned
Google Cloud made a really smart decision when they hired Thomas Kurian to head up their cloud division because he had those deep enterprise roots. It’s the reason that he was in charge of Oracle Cloud. And if Larry Ellison really had gotten out of his way and allowed him to build that business the way he wanted to, I think Oracle Cloud would have been much more successful now than it currently is. And we’ve seen the massive growth that Google cloud has seen over just the short term that he’s been in charge. [00:17:24.730] And it makes me wonder, Microsoft has those deep enterprise relationships. Google historically has been terrible at building enterprise relationships, but now they have someone who knows how to do it. Where does AWS fall in that spectrum? Do they need someone in the leadership position whose sole focus is building those enterprise relationships? [00:17:42.880] – Drew
So, yeah, so here’s the thing. I think that sort of reinforces my point. Google turned to somebody who has enterprise and sales chops. They didn’t necessarily turn to a technologist for the phase of their cloud growth. I think probably that’s what we’ll see AWS do for its move because they, I think, do recognize they need to differentiate. And I guess I would say they are differentiating in that. To me, AWS is the velocity cloud. If you want new cutting edge features coming at you, one after another after another, because you’re at the cutting edge of digital transformation and you can’t get it fast enough, AWS is the one for you. [00:18:14.140] As I said, they’re blazing the digital transformation trail. I think that’s part of their culture, part of how Amazon itself does business. And that’s how they differentiate from Azure, which is kind of your steady, comfortable friend. And you know, you know what you’re getting from them. And Google is just sort of all over the place. And in six months, it may not even be there because everybody knows that Google kills products. So. [00:18:35.980] – Ethan
I don’t think Google’s killing GCP. But, you know, we’ve seen stranger things happen. But, Drew, just to react to the point about if they hire an enterprising salesperson, then, you know, that’s good AWS can build those relationships. But that’s not a ham sandwich, is it? [00:18:52.030] – Drew
OK, so I was exaggerating about ham sandwich. It has to be a sentient ham sandwich. [00:18:59.080] – Ned
Oh, that’s terrifying. I don’t need my ham sandwich talking to me. [00:19:04.060] – Ned
I think if I had to guess and based off of my reading where I would think AWS needs to invest their time and we’ve already seen this a little bit, is in growing the hybrid and on prem portions of their business. Yes. I mean, their cloud business, traditional public cloud will continue to grow at a decent pace. But if they really want to have explosive growth, it’s not going to be in those big regional data centers. I think it’s going to be at the edge in customers data centers and through sort of their local zones initiative. [00:19:33.430] Those are all technology things, though. Do they need a technologist overseeing that, or is that more of a sales type thing? And they can just let whatever groups in charge of that make sure the technology works? [00:19:44.770] – Ethan
It’s both to me because they have to displace VMware. It’s just as simple as that for for a lot of for a lot of the use cases they’ve got to displace VMware. And that’s going to be tough to to get VMware out of there and replace it over time with. Outposts and local zones and, you know, when all the rest of it. [00:20:01.890] – Ned
It’s interesting because I’ve never heard someone say that so clearly and it really just turned a light bulb on my brain. AWS needs to displace VMware because if you think about who’s running it, all those data centers, it’s VMware. You have this weird parallel where VMware also runs on AWS. But that’s probably you know, we’re not we don’t need to get into that. But, yeah, if I think about it, who are the enterprises spending the big bucks on? They’re spending it on VMware and then whatever hardware vendor du jour they want. [00:20:33.840] – Drew
I mean, actually, I think I’m going to push back on that. I don’t think AWS needs to replace VMware because AWS isn’t trying to be in the data center. They are a cloud company and they are trying to draw their customers into the cloud. And the products that they’ve released so far are about that. They’re just a stepping stone into the cloud. They don’t they don’t need to live in the data center. And if you look at how VMware is developing its products, it’s trying to develop to be a hybrid cloud or multi-cloud solution in that we’re giving you a kubernetes base layer where you can continue to run VM, but also develop your new applications and net new applications are probably going to happen in the cloud. And that’s what Amazon and the other cloud companies are betting on. [00:21:12.120] – Ethan
It’s a use case problem, though. I think the reason there’s still a hybrid cloud and so much interest in doing things at the edge are because there are specific use cases that are driving that right at the core of it being latency needing to be close, you know, from a latency perspective, close to the work that is being done, Amazon or AWS, I should say, with Outpost’s is making it easy for you to do AWS like work very close by so you can maintain your operational model that you’ve built up with AWS so that that’s a use case there. [00:21:46.980] Does that mean, you know, displace VMware? And I think that goes back to well, if you want that one handshake, then maybe you start looking at your relationships and go, OK, I’ve been with VMware. It cost me this much. I’m I’ve got all this other investment in Amazon and I’ve got this much. And goes back to my point about, you know, your question Ned, technologists or sales, who do they need to put in place to actually drive that adoption ahead? [00:22:10.590] And I think they need both. On the sales side, you need to convince the C Suite that consolidating around AWS, whether it’s in-house, you know, on premises or whether it’s up in the public cloud is the right thing. And then the technologists need to believe in that as well. Probably starts in the developer room. I’m going to guess. But look, you can have a consistent operational model and have all of the functionality and performance that you’re looking for and all this consistency, if you but it’s got to come, I think, at both levels to make that happen. If the use cases are there. [00:22:43.590] – Drew
I want to push on that a little bit in that I agree AWS needs to be at the edge because they can do, you know, have a solution for local processing. There’s a use case for that. But VMware, I don’t think, has a strong solution for the edge either. They are a data center company the way they’re currently constructed. Yes, they’re building out HCI platforms and stuff, but know they’re also fighting over with Nutanix and HPE and everybody else. So they’re in the same boat as AWS and trying to go to the edge. [00:23:07.740] – Ethan
We’re getting into semantics about what do we mean by data center versus the edge and all that stuff. It in my mind, Edge is a really overblown term where I think of if you’re doing it in-house, call it edge. [00:23:17.400] OK, close enough. That isn’t what that means in every context. To be fair. Certainly isn’t what that means for like a service provider where. [00:23:23.040] – Drew
Yeah that’s not what I’m meaning. But yeah, so. [00:23:24.900] – Ethan
But what do words mean? What do they mean? I don’t know. [00:23:27.870] – Ned
That’s much longer podcast. Let’s not get into that. I think what you brought up in terms of VMware, it just it really it bears additional thinking because what VMware has and what it’s touted as its advantage for something like VMware and AWS is you get to keep your operational model right. You’re currently using the VMware. Here’s a way to use VMware in the cloud. Now if you take that from the opposite end of the spectrum, you’ve got your cloud team developing these cool new applications and you’ve got an operations team that’s helping them. [00:23:59.220] And what does that team comfortable with? What are they using? They’re using AWS. They’re using infrastructure as code. They are comfortable with that toolset. And they look at what you have running in your data center and they hold their nose and go, no, it’s like a rotten sandwich. You don’t want that. They’re not interested in that. And AWS comes along and goes, Hey, buddy, guess what? You know, when your hardware refresh comes around, why don’t you, why don’t you buy some of these Sweet Outpost’s and then you’ll have the exact same operational model in-house. We will manage it for you and you can pay for it monthly. How’s that sound? [00:24:33.190] – Ethan
Depends on what that bill looks like every month, I guess. Yeah. [00:24:36.060] – Drew
I mean, I also feel like we’re getting too hung up on bills because it’s not like VMware is cheap and it’s not like VMware and AWS is saving you any money. You’re spending twice as much for the whole caboodle. So enterprises are going to spend. [00:24:46.650] – Ned
Yeah, right. I think it’s more of a if you can rationalize your operational model to focus on one or the other. [00:24:52.260] – Drew
Yeah. [00:24:52.770] – Ned
That is a big advantage. That will be a cost savings because there’s less training. There’s less. Administrative overhead, possibly less headcount, because you can do more with less. [00:25:02.710] – Drew
VMware is doing what it can to keep people within its operational model. That’s why it’s making the things it’s doing, like with project Pacific and so on. [00:25:10.090] It understands that developers are going to use what developers want to use. And right now that’s moving toward cloud native stuff like Kubernetes and so on. And so they have to embrace it. But it also means slowly over time, they’re eroding their base in favor of the AWS model. And that gives AWS that linchpin with an outpost to come in and start getting some of that enterprise business. [00:25:28.150] – Ethan
You still think, Drew, that it doesn’t matter who Andy Jassy’s replacement is? You know, in this context, it feels like we’ve articulated it’s complicated for enterprises to make the decision about what the right way is to go. I mean, so I still take the tact here that it does matter because you need the right person to position AWS in all the right ways to encourage not just adoption, but then also stickiness. I can’t tell if you are agreeing with that kind of viewpoint or still disagreeing, Drew. [00:25:58.870] – Drew
No, I agree. But I guess I just feel like, you know, again, thinking that innovator’s stewardship model, we’re moving into a candidate pool that’s a lot broader of people who will have the chops to do what AWS needs to do. The hard part was building that model in the first place and laying out the strategy. And you can I mean, I feel like AWS has already anticipated all the things we’re talking about, like, OK, we build cloud. Now we have to get the enterprise. So we’re building outposts and we’re doing all this stuff to sort of suck them in. So the next person who comes in has the roadmap laid out for them. And so, yes, obviously you can’t hire a total knucklehead who’s just going to drive the business into the ground. But the candidate pool is so much broader. And are we going to remember who the second CEO of AWS was 10 years from now? [00:26:40.750] Probably not, because they’re just cranking the crank. [00:26:43.600] – Ethan
They’re just a ham sandwich with good hair who can crank the crack. [00:26:47.270] – Ned
And so you’re thinking this is sort of the Steve Jobs versus Tim Cook sort of thing. Steve Jobs was an innovator, building a company from the ground up. Tim Cook is a steward and he does have a direction. But a lot of the stuff that he promotes was already in flight when he took over. [00:27:06.640] – Drew
It was baked in. Yes. [00:27:08.380] – Ned
He just needs to guide the ship successfully. [00:27:11.170] – Drew
Exactly. [00:27:11.650] And the company culture, I think, which is probably one of the keys here, is baked in as well. Right. How AWS operates, what they expect of employees, how they do the business that that’s baked in and that’s been successful for them. And so all of that is set up for the next person to take over. [00:27:28.870] – Ethan
So I’ll still argue against that. Drew, in this context, even though AWS is the market leader right now for public cloud, for sure, nobody can dispute that. I don’t think everyone is nipping at their heels now. And I’ll I’ll put the John Chambers to Chuck Robbins at Cisco Systems transition out there, who replaces John Chambers. You know, it mattered who and Chuck’s been fighting it for however long it’s been. Now, I’m not sure how many years he’s been in that in his role. [00:27:54.790] But, you know, Cisco had a major transformation to, to undergo a little different than what we’re talking about with AWS here. But there are parallels to be drawn. I feel that AWS because they’ve got that target on their back, stands to lose bits of market share to a ton of different companies that can do what they do or can do some portion of what they do. [00:28:17.200] That might be good enough for a lot of enterprises where enterprises aren’t they don’t have a reason to go, oh, AWS, obviously I’ll start there. And if there’s some reason I can’t do it there, I guess I’ll look outside. I don’t think there’s a reason to think that way anymore. So I think that the market dominance and, you know, dollars that they’re, they’re cranking through, notwithstanding, there’s so many other companies that I as an enterprise could go to do the same thing or very similar thing that did AWS does. Why? Why wouldn’t I necessarily if there’s reasons why I like some other option. [00:28:53.080] – Drew
And that’s the same argument, though, you could make about Cisco, like, why don’t I buy Juniper? Why don’t I buy Arista? Why don’t I buy extreme? [00:28:59.860] – Ethan
Which has been happening? There’s Cisco was not as dominant as they were, say, five or certainly ten years ago, if you look at market share in certain categories. [00:29:09.610] – Drew
But that’s because they’re fighting over a smaller pie. The pie for cloud services is a trillion dollars now. Right, because there’s so much enterprise spend that could transition from on prem to cloud. So the market, yes, AWS could competitors could start to catch up. But AWS can still keep growing into that market. [00:29:26.860] – Ned
Right. Right. It’s not like it’s a winner take all game. [00:29:28.910] – Drew
No, no at this point. [00:29:30.520] – Ned
They’re not they’re not fighting for scraps here. [00:29:32.920] – Drew
Exactly. [00:29:34.660] – Ned
I think I would like to bring up and this is probably the last point to bring up a replacement CEO doesn’t necessarily need to reinvent the culture or change the strategy or direction of the company, but they can easily ruin the strategy or the culture or the technology of a company by making bad decisions. I think we could take a look at Microsoft as a as a parallel. What happened when Steve Ballmer was in charge and then what happened when Satya Nadella took over and course corrected for what Ballmer was doing. That was transformational. [00:30:07.930] What we don’t want, or at least what AWS doesn’t want, I won’t say we I don’t want to I don’t want to assume what our motivations are. What AWS doesn’t want is a Ballmer type situation where somebody bombastic and big comes in, a knucklehead, if you will, and ruins the culture. And what has catapulted AWS to the position they are now? [00:30:29.770] – Drew
I think that’s a great example. But I would say the counter to that is Ballmer came in still believing that the operating system was the value proposition and the market had moved on from that. I don’t know that the market has moved on in a way that the next CEO is going to try to have to anticipate. We’re still trying to digest what public cloud means. They don’t have to transition away from that. The business model needs to be expanded as opposed to just transition. [00:30:52.120] So I take what you mean. The next CEO could screw things up. But it’s also that Steve Ballmer came into a position where he just wasn’t suited to rethink where the market was going, the way that Satya Nadella understood and then did a fantastic job transitioning to. AWS isn’t in a transition space there and a drive adoption and grow its base. [00:31:10.900] I would also say that the next CEO is going to be like five feet away from Andy Jassy. He’s still there. It’s not like John Chambers left right. [00:31:20.740] John Chambers left. And let Chuck Robbins run it. Andy Jassy still there. He’s running Amazon. But the other thing to keep in mind is that AWS contributes nearly half of Amazon’s profits. Right, the net income that they take home. So AWS is key to that business. And Andy Jassy, I assume, is going to have at least one eye on the person who’s taking over. So if that person does start to screw up, then, you know, we’ll see a quick transition. [00:31:44.590] – Ned
OK, I like that. So final thoughts, Ethan, why don’t you go first, have you been convinced? Do you think a ham sandwich, a sentient Ham sandwich could take over at AWS and run the ship? [00:31:56.800] – Ethan
I think the AWS growth curve is not predictable at this point. And you need a particular person at the helm to drive it in a specific way because the competitive landscape of public cloud is not anything like what it was when AWS started. It’s a whole different world right now and it’s a careful, cautious place that AWS is in. I don’t think it’s could simply crank the handle. I think that I do still think the person matters. [00:32:21.700] – Ned
And Drew, your final thoughts? [00:32:23.380] – Drew
I think the there’s a broader a much broader pool of people who could step into AWS now and keep it successful because all of the key components have been set up and it’s a self-perpetuating motion machine and they just got to not screw up. [00:32:37.760] – Ned
I think I tend to fall more towards Drew’s side in the sense that they could, in theory bring someone in and as long as they don’t screw up, AWS will be fine for the foreseeable future. However, they could bring in someone who would further accelerate that. So making a good choice could have a lot of value. Ten years from now. [00:32:57.800] – Ethan
So it was just one more comment I’ll make then, Drew, is I actually agree with both you from a standpoint of there’s a huge assumption in this place that I’ve been coming from in this show, the other folks, the competitors of AWS, have to execute well to take advantage of AWS not executing well. And that is far from a given. [00:33:16.100] – Ned
Well, Drew, Ethan, thank you for jumping on and talking this through and give me a lot to chew on. And I’m sure it’s given the listeners a lot to chew on. Coming up next on the show, we have a special sponsored Tech Bytes Conversation with Thousand Eyes. We’re going to be talking about remote monitoring of VPNs and why it’s a challenge that you actually probably care about right now and how thousand eyes is solving that for you. So that is coming up after this. [00:33:49.650] – Ethan
Welcome to the Tech Bytes portion of today’s Day Two Cloud episode. And joining us is Alex Cruz Farmer from Thousand Eyes. And we are going to get into the topic of VPN monitoring for a remote workforce because because we know y’all are out there with your remote workforces VPNing in from everywhere that is life these days. Isn’t that true, Alex? [00:34:11.040] – Alex
Absolutely. Absolutely. It’s been it’s been a tremendous migration for people to move to working from home. And I think it’s been a it’s been a huge struggle for even the most technical of us to really get up to speed with the types of ways that we’ve got to work differently with our teams, things like collaboration. And the last thing that we really want is, you know, those connectivity problems that are not our fault somewhere in between us and us and the application itself on that network. [00:34:37.620] – Ethan
So VPN connection specifically lets this dig into this, me being one of these people that has had to deal with a lot of VPN inbounds over the course years. And I’ve had VPN concentrators and firewalls and so on that I’ve had to manage. The performance question is always a huge one, I’m trying to connect the dots, and so I’m on the VPN and the applications just crawling. And I think a major point here is that you guys, at Thousand Eyes can actually help me solve that problem. [00:35:06.340] – Alex
Yeah, I mean, the big thing is whenever you have an application crawling and you’re connected to the VPN. The immediate thing you’re going to do is point at the VPN because there’s always going to be that bone of contention. No one likes logging into a VPN it’s hard work. You’ve got to put 2FA in and you got to put a password in. You’ve got to wait for it to connect. You lose local access to your network. Sometimes it can be really, really painful. [00:35:27.160] So immediately everybody will point at the VPN for all of the problems that happen when they’re trying to reach an application. What actually we’ve seen is that when users are connecting to an application via a VPN, what can often happen is that the routing between them and the VPN gateway and the VPN concentrator could be congested. It could be a particular issue. That could be some weird routing issue that they’re upstream ISP has. [00:35:53.590] – Ethan
Now you’re talking about like where the VPN tunnel is going across the Internet. Let’s say somewhere between that end point and the VPN concentrator. There’s something happening there. [00:36:03.490] – Alex
Yes, yeah, exactly that. And I mean, imagine you work for an organization that’s that’s based out of the US and you’re based in the UK and you’re having to connect to it to a US-based VPN or you accidentally connect to a US-based VPN. Sometimes it’s hard to even see that. And it’s hard to understand whether you’ve made a mistake with the connection that you’ve that you’ve made out to that to that VPN gateway. And that could be the root of all your problems. And it’s really hard for for an IT engineer to to identify where the problem really is. [00:36:31.540] – Ned
Right. We’re really dealing with two different sets of traffic. We’re dealing with the traffic that’s traversing that tunnel, and then we’re dealing with the traffic of however that tunnel is being constructed across the wider Internet. And there’s so many different areas where that ball could get dropped. And I feel like, at least in my experience with the VPNs, I don’t have any visibility into that underlay where the tunnel is going. All I have is visibility into my connection back to that concentrator and then whatever is happening with that tunnel. [00:37:02.410] So how would you go about constructing that view of what’s happening in the underlay that’s supporting the tunnel? [00:37:09.860] – Alex
Yeah, that’s really interesting. So when when you’re when you’re looking at the path from from a user’s connected to a VPN, you can never really see the underlay, it’s completely hidden. So what we do at Thousand Eyes is we we we engineer a way that we can then create a path and give you visibility into the path to the VPN gateway. And then what we do is then we merge together the path from the VPN gateway then to the applications. [00:37:33.610] You’ve got a full end to end view. So if the issue is potentially a couple of hops away on your on your upstream provider from your, from your broadband, then we’ll be able to identify it. But if there is actually a problem behind the VPN, we will also be able to help you identify that too, because we’ll have full visibility. That overlay. [00:37:51.430] – Ethan
Oh, that is behind the VPN as in I’ve actually made it to the other side of the tunnel, out the firewall, the concentrator. And there’s something even beyond that. That’s my problem. [00:38:00.640] – Alex
Yeah, potentially. I mean, if you if you think that the VPN concentrator could be congested because there are lots of people connecting to it at the same time, that could be where that bottleneck is. And because we can give you the underlay, we can rule out your connectivity as being the problem and then move that visibility and focus onto the VPN concentrator. [00:38:18.670] – Ned
I think we saw a lot of that issue when the pandemic first started and everybody started hitting the concentrator. And I know, like whenever I sized a concentrator for a deployment, the assumption was you’d only have a small percentage of your workforce, even your mobile workforce, connecting at any given time because they were going to connect and get the thing they needed and get back out. It wasn’t going to be connected all day for eight hours and 90 percent of your workforce. [00:38:44.030] So that’s the problem that we had to solve for a year ago. What sort of problems are cropping up now beyond just oversubscribed appliances? [00:38:53.470] – Alex
So it’s interesting today what we’re seeing is a lot more organizations, are wanting to move a lot of the VPC concentrators away from hardware based solutions into the cloud. Look at things like SDWAN and WAN transformation projects because a lot of organizations don’t see this remote workforce coming all back to the office. I think, you know, as we’re moving forward, a lot of us are going to spend multiple days at home. I’m hopefully going to be spending more days at home, which is great because I have a seven month old daughter that I want to spend more time with. [00:39:25.660] So for me, I know I’ll be spending a few more days at home that I would be pre-pandemic where I was five days in the office. [00:39:33.430] – Ethan
So the point you’re making here is that the model of VPN is changing. That is, it’s not old school. We’ve been doing this right into the firewalls in the concentrator and stuff. And everybody comes into headquarters and we apply our policy and we do things. But now, because our resources are distributed everywhere, it’s up in the cloud. It’s wherever it is. Right. We’ll build an SDWAN fabric that includes those original headquarters locations, perhaps, but also has a remote office locations and also has our cloud presences all tied into that fabric. So when we’re doing VPN, we’re VPNing into perhaps a third party service, perhaps into a concentrator that’s up in the cloud. [00:40:05.740] There’s a variety of ways that we could do it now to put our VPN users closer to the resource they’re trying to consume, which makes monitoring and troubleshooting all that much all that more complicated. [00:40:17.380] – Alex
Yeah, I mean, if you if you consider the you know, we’ve got people within Thousand Eyes who are based out of Portugal, we’ve got them in the US and we got them in the U.K., we’ve got we’ve got teams out in APAC. You know, we’ve got to support all of these different disparate locations. [00:40:33.820] So not having that visibility into knowing where they’re connecting to why they’re connecting to a specific VPN gateway, what the latency is, is that potential congestion there, all of these different problems. It’s really, really challenging to try and identify these without that in depth visibility that you get from Thousand Eyes. [00:40:52.270] – Ethan
So let’s get into the magic of how Thousand Eyes is doing what it’s doing here. In this use case, I’m a former Thousand Eyes customer, Alex, going back a few years. And the model that I remember was there was an agent that would sit in a virtual machine and that I could run tests against some some resource that was out there, perhaps against a remote Thousand Eyes agent. [00:41:13.450] And that gave me all kinds of insights into what’s going on in the way and what’s going on my MPLS cloud, what’s going on across the Internet. And I could see hop by hop where the problems were you telling me I’m getting the same kind of a thing with this VPN monitoring use case. But how are the tests running? Is that is there something sitting on the the endpoint where that user’s VPN client is running? [00:41:35.800] – Alex
Yeah, great question, Ethan. I mean, one of the things that Thousand Eyes is known for is the enterprise agent and our cloud agents. What we actually also have is something called our endpoint agent, which is something that sits on the user’s device. So this is something that can be extremely nimble, extremely mobile. You know, we even found some testing recently and found that our agent runs less, uses less CPU and less RAM than the notepad does. [00:41:59.830] So it’s a very, very lightweight agent that can sit there behind the scenes. And it really, really gives gives that edge to our customers to get the monitoring going. So, you know, looking at the endpoint agent itself, we can do a lot of the same things that the enterprise agent does. We can run scheduled testing so we can have 10 scheduled tasks which allow you to probe out to multiple different applications. We have a browser extension which is added to Chrome and Internet Explorer 11 and Edge, and the browser extension, essentially pulls metadata in and kind of gives you that chrome waterfall style feel in Thousand Eyes dashboard to help you identify where there are problematic documents that may be loading on a particular Web application [00:42:44.920] – Ethan
But wait, you just said thousands dashboard. Does that. Is that a centralized thing that I as an I.T. staff person I’m looking at, or is that something that the end user might look at? [00:42:53.500] – Alex
So the Thousand Eyes dashboard is what the IT team would look at. So this is a centralized view where where it will consolidate all of the endpoint data that we’re collecting for all of your users into one central place. And you can use that and build a thirty thousand foot view to understand the whole landscape of all of your users, but then also then drill down to really, really specific windows down to a five minute window, look at the performance and see how things are looking for that user. [00:43:18.400] – Ethan
Does that mean if I have some high touch customers within my organization, people that are really critical and I want to make sure that their VPN performance is absolutely excellent, I could zoom in on them specifically. [00:43:31.210] – Alex
Absolutely. So the way our agents work is if you go ahead and you can deploy the agents silently through your SCCM or some kind of global deployment, or you could just download it from a website and install on your device, once you’ve got it installed, we automatically will detect that you’re running a Cisco AnyConnect VPN. And we also detect Palo Alto Global Protect, Pulse Secure, and also zScaler Internet access. And what we what we can do from that is that when when schedule test is set up or if you’re using just the browser extension, we’ll start collecting all that path data and giving you that network visibility straight away. [00:44:05.110] – Ned
Alex, where were you ten years ago? When I needed you. [00:44:10.500] – Alex
That’s a great question. [00:44:12.710] – Ned
Horror stories about previous CEOs and such that I’ve had to support with their VPN. And this would have been amazing at the time. It sounds pretty cool now. And I just want to drill to a point that you mentioned multiple different VPN vendors. So this is not a VPN vendor specific solution that you’re supporting multiple vendors with this solution, correct? [00:44:33.130] – Alex
Yeah, absolutely. I mean, while Thousand Eyes did get acquired by Cisco and Cisco was one of the first VPN providers that we did support. And that was way before the acquisition. It was really, really important for us to also support other VPNs as well. There are lots of organizations out there who are looking at doing web transformations, moving from say Palo Alto, over to Cisco AnyConnect or potentially they’re looking at comparing two different VPN vendors or cloud VPN providers together. [00:45:00.530] So we need to give that kind of agnostic approach to it. Otherwise, they’re not going to be able to get that true visibility. [00:45:06.140] – Ethan
So that supporting those VPN clients directly means what exactly? You’re hooking into some information that’s peculiar and specific to those VPNs and being able to trickle that up into the Thousand Eyes dashboard? [00:45:17.450] – Alex
Yeah, exactly that. I mean, we essentially look to see if you’re connected to a VPN or not. And then if we if we know you’re connected to a VPN, what we can then do is, is essentially take them and plumb that into the way that we present the path trace data to you. So that’s how we know and can give you the underlay and overlay data. If you’re not connected to a VPN or using split tunnel, for example, then we don’t put any VPN overlay in that because there isn’t any VPN you’ll be traversing at that point, [00:45:44.750] – Ethan
Right. No path trace data, you just said the magic words there, because when we were prepping for the show and you showed us exactly what you can see, what you can visualize from the endpoint, that person running the VPN client all the way to wherever their VPN tunnel terminates on the other end and all the underlay hops in between, including Alex. And I think this is really cool, like they’re on Wi-Fi at home and it’s terrible. You can actually see that. [00:46:07.380] – Alex
Yeah. I mean, it’s amazing if you think that a lot of the problems that we’re having a, you know, when you’re at home is, you know, you may have like a son or a daughter who’s like playing the Xbox in the next door and downloading stuff and saturating the Wi-Fi. Or it could be just simply that you’ve connected to the wrong Wi-Fi access point. [00:46:24.530] There are some very, very basic things that IT are not able to see when you’re at home without something like Thousand Eyes on your device, because this is a type of telemetry that just isn’t isn’t usually available. So with Thousand Eyes in that, we will be able to pull things like your signal quality and we can make sure we’re providing that SSID. So we know well, the IT team knows that you’re connected to the right access point and then we can kind of trace that back and make sure that things like your local gateways not got packet loss or not high latency. [00:46:54.030] There’s a whole load of step by step things that you can do all the way from the from that device, through your Wi-Fi, through your local gateway, across the network, all the way out to the application. [00:47:03.950] – Ethan
Hmm. Great stuff, Alex. If people want to know more about this, they want to dig into the solution. Maybe they want to give it a trial. Where would they go? [00:47:11.780] – Alex
Yeah, I mean, had ThousandEyes.com. We provide a free trial for 14 days and you can try out a cloud agent, enterprise agent. And you can also obviously get our end point agent as well and try it out for for 14 days. And our support team is is stellar and they’ll be more than happy to help you along the way if you get stuck. [00:47:28.010] – Ethan
Alex, thank you very much for being on our Tech Bytes portion of Day Two Cloud today. That was Alex Cruz Farmer, chatting with us about VPN remote monitoring. And that brings us to the end Ned. [00:47:39.590] It’s very sad. That brings us to the end of our Day Two Cloud today. So virtual high fives to you for listening. You’re pretty awesome out there. It’s still a pandemic going on. We know we’re all sick of it. We know. But you are still awesome. We just want you to know that’s a big deal. [00:47:53.150] If you’d like to reach out to Ned and I about a topic you want us to cover by all means, Twitter at Day Two Cloud show or go to the forum on Ned’s very fancy website, nedinthecloud.com. [00:48:05.480] Anything you’ve got that’s cloud related or career related, like we did a show on IT Consulting recently that we got some good feedback on a lot of people really liked that show. Hey, we’ll talk about that career stuff too, if that’s what you want. Send in your questions and we’ll we’ll find some guests and and have a show. We will do that again, those virtual high fives to you. Just remind yourself how awesome you are because you really are. Last but not least, remember, cloud is what happens while it is making other plans.
Podcast: Play in new window | Download
