17. Main Takeaways from Dockercon 2018

Rich: In episode 17 of Mobycast, Chris provides his main takeaways from DockerCon 2018. Welcome to Mobycast, a weekly conversation about containerization, Docker, and modern software deployment. Let’s jump right in.

Jon: Welcome Chris and Rich.

Chris: Hey guys, how’s it going?

Jon: Good, time for another Mobycast.

Chris: What number we are we on now?

Rich: Seventeen.

Jon: So now we’re committed to releasing this in order, but it’s good because this one should come up soon anyway. Chris was just recently at DockerCon, so today’s Mobycast will be all about his impressions and takeaways from DockerCon 2018. We only have 20 minutes and DockerCon took three days. Rather than find out what we did over the past week, obviously aside from Chris who was at DockerCon, let’s just jump right in. Chris, tell us about DockerCon.

Chris: This was my third DockerCon in a row, it started up in 2016 in the Seattle one, that one I believe was about 2000 attendees, 2017 was in Austin, and I believe it grew to about 3500 attendees and this recent one was in San Francisco where Docker’s headquartered and they were recording numbers of between 4000 to 5000 attendees I think.

Jon: I read about that, 4000 attendees at DockerCon, how will they ever “contain” them.

Chris: It’s still a T2 micro conference compared to re:Invent, so I think they’re okay for awhile. The point is it’s definitely growing and getting more popular, and the thing that just blows my mind in every single one of these conferences is, they always throw the stat that about half the people there are brand new to Docker.

Docker has been out there now for five or six years easily, and you still does have a lot of people just fresh folks looking at Docker. It still has a lot of room to grow, and a lot of market to take down. Some of the numbers they threw out were over the past year, they’ve had 1 million new user/downloads of Docker Desktop, that’s the client bits for either Windows or Mac which was kind of interesting, that’s definitely a pretty large pretty large number for developer reach.

On the image side of things, they were quoting numbers in the billions of polls and in a matter of two weeks, so 1 billion image downloads every two weeks is where they’re standing right now on Docker Hub. Pretty big numbers, pretty amazing. Docker is growing but, I think the big thing there is that there’s still lots of room to grow. It’s feels like we’ve been working with it for awhile now and kind of feel saturated. It’s always going to these things, it’s always realizations that it’s not saturated, there’s still plenty of room there for growth.

I think this one was really interesting too because I was in the first conference since some management changes there at Dockers, some pretty big one. At the last one in Austin back in April of 2017, both co-founders of Docker Ben the CEO and Solomon the CTO were still in place. Since then, both have left the company and Docker’s gotten a new CEO with bringing onboard Steve Singh, who founded Concur which even sold to SAP. I think the personality is definitely changing a bit. I think Docker definitely has some very clear business goals and objectives now.

It shifted a bit with that management change and it’s definitely intentional. The Docker has been around for five or six years now, it has raised quite a bit of money, it’s definitely one of the core key technologies out there today for anyone building a software especially in the Cloud, yet I think they’ve obviously struggled from the business model perspective. A great case in point is, we’ve been using Docker for quite some time. The honest truth is, we’ve never given Docker a single penny.

We’re just using the core Docker engine right, they don’t make any money out of that. That is by far the common scenario, that’s how everyone is using it. It’s kind of like, “Hey, here’s this really great free software to go build your modern software stack on.” And basically Docker, there’s just no money in there, in that business model. Definitely as a company, if they want to continue on, they need to find that business model, and find a way to bring money in, and to reach profitability, and bringing Steve Singh onboard is definitely loud and clear, that is his number one focus. That was felt throughout the conference this time, it was definitely obvious that, this is basically the number one priority, and that really came across with their business model now, the way that they’re going to make money is all through their enterprise edition of Docker. That’s how they actually get revenue.

Docker EE was the recurring chant, and the recurring theme across the entire conference this time, much more so than it has in previous editions. In previous editions, it was definitely more about technology, innovation, and new features, new products, or new technologies, and this one, very little on that front, instead, really focused on Docker

Jon: That’s sort of sad, I mean, imagine in 1998 or 1999, and you have been going to a Linux conference every year, and then you go to one in 1999 or 2000, and it’s still called the Linux conference, and all you hear is Red Hat Enterprise, and you’re like, this is not the conference I was planning on going to.

Chris: Yeah, absolutely and there’s definitely some of those feelings, the personality, the feel of the conference was markedly different than previous ones. Definitely not as much excitement but again, I realized, and I recognized, we want Docker to survive and continue, and they do have to find a way to make to make money. It’s just a necessary consequence I think.

I wish it would be more of a balance, because it used to be like especially with Solomon kind of taking the role of being the chief product guy, being a tech guy, being the CTO, co-founder, founder of the company really. I think you know his day-to-day perspective was all about just building cool technology, and innovation, and just and just really focusing on that, and maybe even to a fault where it was kind of building technology for technology’s sake.

And to say like, “We can build technology just as good as someone else.” in particular, something like Swarm to me as a bit of a head scratcher, especially at the point in time when they’re doing that with Kubernetes already had such a firm foothold as kind of like the de facto orchestrator to go and build your own competing one. There’s definitely some questions there, is that really the right thing to do?

Definitely the cool technologies, and some great stuff, some people are using it but not sure really that’s the best use of the resources.

Jon: Let’s use that as a segue because, you just mentioned Swarm, and you mentioned Kubernetes, and it seems like maybe that’s where the excitement has gone. Docker was exciting on its own for the last two years. Now, Docker is a company, the company is trying to make money, and they’re just doing the things that companies do. I was at Gluecon a couple of months ago, and it’s like whoa, Kubernetes, and now it seems like, maybe all of the people that were fervently working on Docker open source projects are now, their attention has shifted to Kubernetes, and that’s maybe where the excitement is.

Chris: The truth is that people have always been using Kubernetes and it is the de facto standard, I think we talked about this in previous editions of the podcast, it’s somewhere around 70% of the folks running containers are using Kubernetes as their orchestrator, and then 30% is the rest, whether it be ECS, Swarm, or something else. People have been using it, I think what was really different with this one is that, Docker has just basically opened up its arms, and it’s giving just a huge bear hug to Kubernetes.

This is now the Docker acknowledging that Kubernetes is the de facto standard, and that they’re no longer going to compete against it, instead, they’re going to embrace it. I think the writing is on the wall there for Swarm. At this conference, whenever they talked about the orchestration, it’s always Kubernetes, and Swarm were kind of mentioned as just you can choose this one or that one, they always position against the two of them. Notably absent of course was in the Amazon technologies, ECS was not mentioned once throughout the entire conference which was also kind of interesting but not surprising.

Jon: That is surprising to me because it’s not like Docker is competing with Amazon. Docker would benefit if Amazon succeeds at using Docker and lots of its products. It doesn’t fully make sense to me. Turning their back on Amazon is like turning their back on their state of goal, and profit, and the growth.

Chris: Docker’s business owners, really all around Docker EE, and what is Docker EE? Docker EE is the infrastructure software for running basically a datacenter. It’s things like user management, and permissions, and security, image repost, it is the UIs in management of how do I deploy apps, and hook into orchestration, how do I manage clusters of machines and whatnot. All of that though is like, that’s exactly what AWS does.

You get into this situation where it’s like, and this is one of the reasons why Docker EE has not been too terribly interesting to us, in the work that we’re doing, we already get all of those features just about through the collective dashboard that Amazon provides. Especially with things like ECS, with its deep integration with all the other AWS technology such as scale groups, and Route 53, and load balancers and whatnot.

Especially ECS, it’s very much a competitor to them. They do give it a little bit of lip service to the fact that like, hey, if you got some easy tools that you’re running, not on Kubernetes or something like that, you can go ahead and manage that with Docker EE. If you’re all-in in Amazon, then it’s really hard to make a compelling reason for why you would use Docker EE. I think maybe the other thing or two is that, you can say that in a way they don’t compete in that Docker, Docker EE is definitely much more suited for enterprises that are mostly On-Prem or maybe hybrid. They’re not really going after the folks that are all in the public Cloud especially something like AWS. From that standpoint. They’re going after different customer segments.

Jon: It just feel so fraught with mistakes, it’s like let’s build a bunch of management software that loves being On-Prem and put our whole company behind it. Docker is like the best and easiest way to get off-Prem on to the Cloud. It just doesn’t sound right to me. It’s like building management software for managing application installations On-Prem is great. I’m glad that they’re doing that. But, betting their company on it and ignoring Amazon it’s like, they should figure out a way, for me, it just feels like they should figure out a way to have a pay for version of Docker where enterprises can feel comfortable that they’re not downloading some sort of scary open source software.

That they can give support on the Docker containers that they have in production on the Cloud or On-Prem, maybe it has just a little bit of features here and there just to sort of like get a little insight into what’s happening with your containers, and some sort of seal of approval, maybe they have certificates in them or something that you can only get from Docker to prove to yourself that these containers are real Docker EE containers, something like that, and then that’s it. Don’t go through this management software and put your whole company behind it. It just seems ridiculous to me given what Docker is for. That’s too strong of an opinion but it’s wild to me what they’re doing.

Chris: Yeah, I think this is definitely a work in progress for them, and I don’t think they’re really executing right now against a five-year plan, this is just more of what’s going to happen over the next years. With all that said, I’m sure they’ve had plenty of discussions and they thought through all this stuff, and they do have those longer term plans. I think they’re looking at this like look, this enterprise market that’s out there where folks have these legacy apps running on outdated systems that need to be modernized it’s such a huge opportunity, it’s not really been served so well, this is where they can focus, they can say like this is where we have a really strong message.

That’s phase one, they are given the promise that Docker is a great way for moving your stuff around and get enough On-Prem, or just run it anywhere whether that’s running On-Prem, Private Cloud, Public Cloud, it kind of feels like they’re still for right now on, Amazon is just really not part of those near term plans.

Jon: For those that have been keeping up with Silicon Valley, it reminds me of the box, like Docker, I’ve got it, let’s make a box that can put on to your data center shelves, let’s make a Docker box and you can literally plug it in.

Chris: I have not been keeping up on Silicon Valley.

Jon: But you can imagine the box, you can imagine the physical piece of hardware, and you can imagine making a ridiculous decision to build a physical piece of hardware when you have this awesome software.

Chris: Yeah, and by the way, that whole thing did happen back in the early 2000’s, back then instead of a box, it was called an appliance. Maybe coming back around like this whole thing, because it really was true, like that was the way that you sold your software into a way of putting a tangible value on it, is you sell it as a box. Someone combines them and just rack and stack it in the datacenter.

Jon: Okay, so we’ve had the Kubernetes discussion, we’ve had the mysteriously absent AWS discussion. AWS by the way was there, and they have a huge booth and the AWS umbrellas were tweeting up a storm the whole time. It wasn’t that they weren’t there, but I guess they weren’t mentioned in any big keynotes, or other types of breakout sessions that much.

Chris: And very little presence on the speaking circuit. There was the talk by Abby Fuller, she did the same talk in Austin as well about making efficient Docker images that I think when she did—she’s recently at Amazon, I don’t think she’s been there all that long. I forgot where she was before that. In previous congresses, they actually had big announcements with AWS, and partnerships, and like they actually had Docker for AWS was one of their product pushes two years ago. All that stuff just is gone. It’s no longer there.

Jon: And there’s rampant speculation around the internet that Docker will be the next to go, and they’ll get bought by Microsoft, or Google, it would be interesting to see if Microsoft or Google kind of got into a war trying to get Docker. The CEO there, based on their attempts to clean everything up, and get focused on getting cash flow positive, that aligns with that speculation.

Chris: Yeah, it brings in a whole bunch of like—so Steve’s onboard now, and his whole thing is get it to cash flow positive. They’re rallying around Docker EE and it’s basically all about sales now. That’s going to dictate whether or not they can get there. I think it’s very much a big question mark on whether or not EE can get enough attraction to actually be a viable business model that sustains the current evaluation that they’ve raised at. I think it’s a hard road for them to go, it’s kind of hard to see down the road of them going, having been strong enough financially to warrant an IPO.

I think definitely acquisition makes a lot more sense here and definitely much more reasonable evaluations and much easier acquisition to make versus a get up type of thing. I think it’s just like a slam dunk for Microsoft to come in and just do it, it’s a one, two punch that kind of get up, and Docker is all about going back to the Steve Bomber on the stage, developers.

Jon: I would say that Docker really is, if they were to look at which one, be it Google or be it Microsoft, which one are they courting themselves to, it’s pretty obvious it’s Microsoft right? Because they’re saying, let’s make this easy for legacy systems. Let’s get legacy systems on here, let’s build On-Prem tools whereas if they were cozy enough to Google, they would probably think about what’s exciting for startups, what’s exciting for big data companies, and that would be more interesting to Google I think in the Google Cloud mindset.

Chris: Yeah, absolutely spot on. Microsoft, their core customer base is definitely the enterprise, that’s what they’ve always been super strong in. that fits really well here, and then the other thing is that Docker and Microsoft have been really tight over the last few years as they work to bring Docker to Windows and then get deeper integration. There’s been a lot of collaboration between these companies much more so than with Google.

They have a long standing, just about every single one of these shows, they always have some Microsoft person up on stage, and that was true again this year. They had one of the SPPs from Windows I believe it was as part of the keynote with Steve Singh.

Jon: Microsoft is just working its way backwards towards Docker just as fervently. Gluecon, Brandon was on stage showing Visual Studio and he had a plug-in that was only available on his machine, so I right click here and I click the button that says Kubernetes and staging, voila. See, now I’ve got this thing running right here in Visual Studio in that Kubernetes cluster and I can see it in play with that right here, and there’s my code still on the screen.

I certainly hope that they get that done quickly because while Docker makes itself more boring as a company, let’s just get this part over with so that they can become part of Microsoft and then if there is a DockerCon 2019, it’ll be a big Microsoft funded event and maybe have some more interesting things on your innovations, because certainly, any other thing as Microsoft is moving quickly, they’re not looking sluggish at making new technology, and new stuff, and their development sets a development tool. DockerCon 2019 if Docker is part of Microsoft, can be pretty fun.

Chris: Absolutely, it’s full speed ahead, so with Amazon being such a fierce competitor with Microsoft, and Google, and just Amazon having such a huge fleet definitely can see Microsoft, “Okay, what’s next?” that’s the average strategy, let’s execute against it and get back to those developer roots and getting people really coming back to Microsoft as the great platform for building, because Amazon has kind of taken that with Cloud being a platform now.

Jon: Sure, well thank you Chris, and thank you Rich for another fun week, and maybe next week we can talk a little bit more about some of the technical stuff, the few bits and pieces that did come out of DockerCon that were interesting.

Chris: You bet. All right, thanks guys.

Well, dear listener, you made it to the end. We appreciate your time and invite you to continue the conversation with us online. This episode, along with show notes and other valuable resources, is available at mobycast.fm/17. If you have any questions or additional insights, we encourage you to leave us a comment there. Thank you and we’ll see you again next week.