All Things Pivotal Podcast Episode #22: James Watters on Multi-Cloud Support and Pivotal Cloud Foundry 1.4—Pivotal Conversation

April 1, 2015 Coté

featured-pivotal-podcastThis week, we check in with James Watters on the Pivotal Cloud Foundry 1.4 release, including support for Amazon Web Services and free hosting, and what customers have been doing with Pivotal Cloud Foundry. We also discuss the wide multi-cloud options available for Pivotal Cloud Foundry now, and talk about some of the up and coming features and customer use cases we’re seeing over the horizon.

 

PLAY EPISODE #22

 

Resources

Transcript

Coté:
Hey James. How’s it going?

James Watters:
Hey.

Coté:
You’ve been busy talking with press and talking to all sorts of people about our upcoming 1.4 release.

James Watters:
I’ve gotten really good at saying the same thing over and over to different people.

Coté:
Exactly. I always find the issue is you have to vary up the jokes that you make because they get really boring after a while.

James Watters:
I just realized I didn’t ever tell any jokes. Man, I didn’t even try.

Coté:
I’m sure there’s more briefings so next time you hit a trough of repetition, you can lift yourself up with some little jokes.

James Watters:
I need jokes. I get so addicted to the strategy part of this. I’m like, “Isn’t this so exciting to you?” That’s more of my approach.

Coté:
Without going in it too much, one thing I was reflecting on is: you were recently promoted to GM, General Manager of Pivotal Cloud Foundry, if I remember correctly. You actually did strategy work a long time ago, and it always picks up in my mind when someone from strategy moves over into the business function, which I think is an interesting path. Anyways. What are you doing nowadays as GM of Pivotal Cloud Foundry? What’s that look like day to day? We tend to have a pretty unique way we go about doing things, so I’m curious how that’s been panning out for you.

James Watters:
Yeah, I think the key thing in my job is really helping to connect the resource investments we’re making which are pretty significant. We’ve got around … I am responsible for the cloud platform group here at Pivotal, which is Cloud Foundry, and some of the Spring R&D as well. We have $70 to $80 million dollars a year in R&D to put to work. It sounds like a medium sized number and then you realize that’s a lot of people waking up in the morning saying, “What can we go do? What kind of R&D can we go do to help customers?” My job essentially comes down to staying very close to the market, and staying very close to customers, and then helping all of the really incredibly bright people on the team align with the trends in the industry, and what customers need. We’ve had thankfully a really great year out the gate last year. According to Matt Asay, who I trust, no one had ever had that fast of a rise in open source software in their first year zeroed around $40 million dollars in first year sales.

A lot of it comes down to thinking like, “Hey, how do we attract the right talent to continue that march?” You’ve seen a guy named Michael Coté came to the team. We were pretty excited about that. Other folks, Benjamin Black, Josh McKenty, my very dear friend Andrew Shafer. We’ve had some other folks join the team recently that we haven’t announced yet that are really big hires. It comes down to how do we stay close to the market, attract great talent. That’s my job.

Coté:
Yeah. That makes a lot of sense. One of the things I was interested this episode to talk about you with is, like you say, you talk with customers all the time and prospects as well, and I think you always have a unique big picture of what our customers are up to and what they want from Pivotal Cloud Foundry. I’m curious, in thinking about … We’ve got a whole passel of features in 1.4. A lot of them are building out the developer related functionality, making it more operational. Essentially, making it a true multi-cloud platform that basically runs on whatever type of cloud you’re interested in-

James Watters:
Yeah.

Coté:
… Maybe except weird little clouds here and there, but all the great clouds. It’ll run on those.

James Watters:
Yeah. The top hits.

Coté:
That’s right. I’m curious when you think about what we’ve done with 1.4, how would you categorize what our customers are up to that have been driving those features? What types of applications and things are they building and using Pivotal Cloud Foundry for?

James Watters:
I think last year we had an incredibly successful year largely selling to people who want to do things on premise by and large. We have some customers … You could do a command line install of Pivotal Cloud Foundry last year, which was a little bit more intense, so a lot of people opted for the simple … I like to call it the dim sum menu install which is, “I want this many nodes of this, and this many nodes of that,” and, “Where’s the cart?”, and “It’s delicious.” We’re trying to take that dim sum style ordering menu. Two new things which is the Amazon platform so you can now do the simple install based on an AMI on Amazon. The reason that’s important is that you asked a great question which is what kind of apps are driving this? A lot of apps that drive Cloud Foundry are new business initiatives. We talk to a lot people in the IoT space and they’re like, “We’ve got to go build all of our APIs to do data ingest and device management for IoT.”

That’s often less about staying close to the datacenter, and it’s more around staying closer to a cloud scale out pattern. What we’ve done now with the dim sum of 1.4 is we said, “Hey you can get started on Amazon today even if you’ve got this new business initiative. Let’s go get started today on Amazon together.” I don’t know. I see a lot of net new businesses and functionality being built upon Pivotal Cloud Foundry over the last year. That’s been the number one driver.

Coté:
Yeah, and especially as I’ve gone out on the road more and talked with people. More and more what I find people using Pivotal Cloud Foundry for is, “Here’s the new platform we need for writing all these wacky, fun new applications that we’re doing.” Everyone seems to have the notion of, “I know I should be writing cloud native applications, and I should be running them on a cloud thing, so I need that.” It’s a responsibility when it comes to architecting the new stuff that folks are working on.

James Watters:
That is who we are in the market. I think this is the easiest part. When I do executive briefings around PCF I say, “Look. We’re not here to sell you orchestration snake oil. You can’t go and slather us in your old datacenter and it’s going to be better run than you had before. We’re not an ITIL tool. We’re none of these old school things that’s your mess for less. It’s not what we do.” I think the trust we’ve built with customers is because we are comfortable with what we are and what we’re not, then when they come to an affirmative case of, “Yeah, I do need to build cloud native applications, and those will shape how my business is going, then yeah, you guys seem like a really honest, great, world leading choice.”

Coté:
To that end we briefly touched on this a little earlier. One of the things we have in 1.4 is we have the brilliantly named “license entitlement” for it, which gets to the point of what it is. Licensing stuff always sounds thrilling. It really means that if you have a Pivotal Cloud Foundry license to run instances on your own you can now include it in that pricing, no change in fee. Run it on our Pivotal Web Services instances. We were talking with some analysts about this earlier today. They raised a good point, which was fun to discuss with them which was, “So does this mean you’re all about public cloud, or private cloud?” They were trying to nail us down on that specifically. I wonder looking several years out, and especially from what you’re hearing from customers. How are you seeing them sorting through, “I’m all private cloud versus public cloud.” Are those distinctions clear in the conversations you’re having? How are people thinking through that?

James Watters:
One of our early customers, and they’ve been very vocal awesome supporters of us; we’ve been successful with them, is CoreLogic. I was down there in L.A. with them last week. I’ve been briefing every customer I meet around this coming entitlement, the hybrid skew as I like to call it. I said, “Hey. I’m here to tell you about a new product thing. The good news is that you already own it.” What I said is what we’ve done now is we’ve included Amazon capacity. Maybe you’ll only use it for dev test, maybe you’ll only use it for new business applications that aren’t as connected for legacy, but for free, and the software you’ve already bought for us, you can go run that on Amazon. No other budget process. You don’t need to hire people to operate it. The operations and the capacity are all included in the software you already own.

I really felt like Santa Claus, because he turned to me and he said, “James. I want you to know you validated why we bought Cloud Foundry in the first place, which is you’re about multi-cloud choice.” This lexical riddle that the analysts are trying to pin you down to is over, which is there is no public cloud versus private cloud. There’s really generation 2 style apps, and there’s cloud native apps. People are choosing a new platform for cloud native, and once they chose that platform, of course they assume it runs on all the Fab 5 top clouds. They assume they get that out of the box in this cloud native platform. I think that we’re lessening the distinction between public and private because it’s an artificial one. It’s really around cloud native patterns, or not.

Coté:
I think that’s the discussion we had with them, and I think that’s exactly it. It’s the point of if you’re an enterprise architect type, or whoever it is you’re sitting there and part of your job is thinking about the future sustainability of your technology choices like, “I’m going to write this application. I want to make sure it’s future proofed as much as possible.” What’s nice about this seemingly boring license thing is it brings together one of the main vision points that attracted me to Cloud Foundry and Pivotal Cloud Foundry which was there aren’t really specs anymore, but here’s as standard as a standard as you can get, and then it allows you to move around to various types of cloud architecture. You don’t necessarily lock yourself into 1 type of infrastructure.

Then things like the licensing change clears up the flexibility of that. Makes it a lot easier to … Nothing’s ever completely future proof, but it definitely adds a lot more, as we used to say, architectural runway to it. You’re not locked into whatever cloud platform you’re running on.

James Watters:
Yeah, think of budgeting processes that they have to run like, “Oh, I’ve got a budget for this, budget for that.” Now they can come to us and they know to the extent they need to run it on-prim. They’re covered, and if they need capacity on Amazon it’s actually included. The thing that to me is interesting I’ve raised with some journalists when I talked to about it’s like, “You should be asking me how I can afford to do this.”

Coté:
Right.

James Watters:
You should be looking at me like, ‘Is this guy crazy? Is this a plan to go out of business? Is this some sort of mattress sale with the inflatable arms thing? What is this?’ It’s really about containerization, and very high scale operational principles built into the platform. Our Linux container stuff slices up the memory space such that when you start an app you only use exactly what your app needs and no overhead, which is unlike the Amazon Virtual Machine t-shirt sizes. Then BOSH is so efficient at deploying, and operating, and the automated health checks built into that it doesn’t cost us that much more when people bring that new applications to the platform to run it for them. Then it’s almost like you turn that on them and we’re like, “We can do that for us. Imagine what you could do internally with this tool.” I think that inspires my own imagination is that hyper-efficiency. When you drive down the cost of something it can do amazing things.

Coté:
Yeah, it’s sort of insert Jevon’s Paradox here, right? As things get cheaper in all senses of the word you gain more power because you can use more it.

James Watters:
Yeah.

Coté:
Hopefully for better. Sometimes for worse, but usually it’s to actually do things that are not wasteful, and extend your capability.

James Watters:
Is it an evil application?

Coté:
Exactly.

James Watters:
That’s where my mind went.

Coté:
Speaking of looking toward the future. You mentioned Ben Black and he started up our Internet of Things lab to look at … I don’t know, new technology driven things that all these companies we’re working with are doing. When you look at 1.5, or 2.0, or what we’ll be doing over the next couple of years. What are some of the things in addition to IoT that you think are more interesting than not?

James Watters:
Yeah, Ben came in … The story there is there’s a friend circle here. I think the same way you came in because you knew Shafer, and Ben, Shafer, and Josh and I knew each other. I was talking to them on Skype about some of these challenging IoT style applications that people are asking us to go build. That’s honestly how it started. He was like, “Oh, that’s interesting,” and then eventually he’s like, “Hey, I could help you guys work on that, right?” I said, “Yeah, you’d be hired tomorrow. Do you want to do that?” And we did it. There was no kind of … That’s what I love about it is that customers saw the basic benefits of the cloud native platform, and then they came and challenged us. I’ll give you an example that Ben came and worked on immediately that’s not pure IoT, but it’s very high throughput.

Anti-money laundering is a big deal in financial services right now. The challenge is you need to bring together, not just the swift transaction of the money moved, but everything about all the people involved on that transaction, correlate that in real time, and do very high throughput processing.

Coté:
Right.

James Watters:
The first thing that Ben was working on was how do we go make that as a service more effective on Cloud Foundry for these customers? It came down to not the basic runtime app runtime there, but also these supporting services that are next generation for that, and then we’re going to probably standardize that pattern. I think that’s what he’s going to help us do in IoT. The thesis of him building out a team up there is that there’s going to be other services you need that are IoT specific that remove the need to write custom code, and custom services yourself.

Coté:
Yeah, I know. I think that’s a good analogy for one of the things that as I get to know and think about Pivotal Cloud Foundry more that’s exciting is, what we really have is this platform for making services like that, that operate at cloud scale. Once you pull back from, “Oh, it’s just a Platform as a Service,” and you realize it’s more of a services platform, that gets to be very interesting. Whether it’s IoT, or whatever, whatever kind of thing comes up that you need to run as a service, the platform is built for that, and it brings all that easiness of cloud to it as opposed to weird proprietary things that you have to stand up nowadays.

James Watters:
Yeah, I think platforms as a service was a word we used because it was close to what we were doing, but from the day we started I talked to Scott Yara about it I remember back in 2012 and we decided we’d continue to use that word, because people at least knew that was the next evolution in the cloud stack above infrastructure. I think that the word got overly contaminated with a very overly opinionated, overly limited runtime manager. That was it. I think what was lost in that word in some of the common parlance was the scalable data services that were just as much part of it. To me if DynamoDB if I’m going to compliment Amazon, is a really nice cloud platform service. I think we’re going to end up calling these cloud services, scalable cloud services, to get away from the limited runtime history that the word PaaS got.

Coté:
Right. We’ll definitely have to avoid having “services as a service.” That would very confusing. See how many “services” we can get in a word.

James Watters:
That’s the funny story. The only joke I told actually today. I did tell maybe 1 joke. They asked, “What’s the name Pivotal Web Services? Where’s that from?” I was like, “Well there’s 2 things in the industry you can call it. Google calls it Google Cloud, and Amazon calls it Web Services, the hosted edition.” I said, “Well, it was called Cloud Foundry and I wasn’t going to call it the Cloud Foundry Cloud.”

Coté:
Right.

James Watters:
That would … I would have a bad time if we went with the common industry use of web services to denote a hosted version of it. The world will never get to meet the Cloud Foundry Cloud Cloud Cloud.

Coté:
Exactly. It’s like Cloud cubed.

James Watters:
Cloud in Foundry Cloud in Foundry Cloud.

Coté:
Well great. I think this is a nice overview of what we’ve been up to in Pivotal Cloud Foundry-land for a while. Is there anything else you want to add, or throw out there before we wrap up?

James Watters:
I think my honest question, strategic question to you Michael is will anyone follow us? Is there going to be … Is someone at Oracle going to go, “Hey. We noticed that when we buy WebLogic it doesn’t come with a cloud service of it integrated. It’s not cloud installable and scalable natively.” I wonder how long we’ll be alone in this, and as the price of cloud infrastructure falls how much more middleware out there will start to include capacity in it right out of the box? That’s an interesting strategic question I think. I’d love to get your thoughts sometime on that.

Coté:
No, I’ve been thinking about that quite a bit as well. I always half jokingly say, “Oh, you know it’s just middleware,” but you highlighted why it’s … I don’t know what half it is, but why it’s half inaccurate as a sense of it’s not just middleware like a bunch of bits that developers get a hold of and do something with. There’s this whole operational nature to it. It’s not completely different, but it’s significantly different than middleware in the past. I think to your point I’m glad that it’s well established there’s good first mover advantage for things. One, when you have this sort of epiphany, they’re like, “Oh, it’s a platform for running services.”

James Watters:
Yep.

Coté:
That’s a major step. Then two, what you were talking about when you have this epiphany of, “So middleware’s cool, but being able to run it is even cooler.” Then it’s all soft of obvious, and it’s a race to do it, and get it implemented. Hopefully it’ll turn out that it’s been nice that we’ve been doing it for so long, because those two shifts are important for recasting this stack of software that people use and provide.

James Watters:
To close, the customer I was visiting in L.A., they were previously mostly an Oracle shop, so imagine after I just came down Santa Claus style and said, “And now you get a cloud, and now you get a cloud.” See, I’m trying to tell more jokes, Micheal.”

Coté:
Yeah. I see. It’ll make your energy last longer. Just a little jokes here and there.

James Watters:
Imagine being the Oracle rep to walk in with really nothing new to offer other than the same old, same old to that then excited CIO. I honestly … I think it’ll be a real conversation starter.

Coté:
We’ll have to do another check in around 1.5 time or so, or some before then and see how you’re off season Santa Claus-ing is going.

James Watters:
Yeah.

Coté:
I think it’ll be a good update.

James Watters:
Traveling the country giving away clouds.

Coté:
Exactly. All right. I’ll let you get to developing more jokes to talk about with people, and-

James Watters:
That’s great.

Coté:
… And we’ll see everyone next time.

About the Author

Biography

Previous
EMC Turns Open Source Sensei, Stages First Cloud Foundry “Dojo”
EMC Turns Open Source Sensei, Stages First Cloud Foundry “Dojo”

In this guest post from Brian Gallagher, President, EMC Cloud Management Division, and second degree black ...

Next
The New Flo for Spring XD
The New Flo for Spring XD

Flo for Spring XD is an incredibly powerful tool with a graphical canvas and DSL access. This first product...

×

Subscribe to our Newsletter

!
Thank you!
Error - something went wrong!