Docker and the Blue Fairy

TechReckoning Dispatch Vol. 4, №2.

Hi friends,

I told you it was going to get worse before it gets better. I had no idea how much the current political situation was in part caused by, and continues to be influenced by, social media. (Recent: 48M Twitter accounts may be bots.)

I got a lot of reaction to my last letter. It turns out a lot of us are taking a good hard look at social media and how it is affecting our psyches and how we spend our time. Some people are opting out completely, especially off Facebook. Some people are using tools to block out distractions during the day. I just pulled Twitter off my phone (although if you want to fall down the rabbit hole, you're welcome to refresh my Apocalypse Watch Twitter list of US journalists and a few liberal activists throughout the day.)

Too often -- and this has happened to me in 2017 -- too much information becomes noise that leads to paralysis. My brother, who has been a pastor and social worker, points to the stress of all the emoting without the benefit of slowing down for ritual, without coming together. Yelling is cathartic but does not sustain us.

Readers of this newsletter fall across the political spectrum and across the world. I think those of us in the US, no matter our political leanings, all realize that something profound has shifted. On the liberal side, people are starting to gather in their home towns more than before. I also see people controlling their media diet more. Newspaper and magazine subscriptions are up. Anything that curates or puts an editor between the raw noise and creates some sort of perspective is welcome.

A healthy America requires engaged citizens. We're still a divided country, but I think people are looking around. So wake up, slow down, establish some rituals, and connect with your community.

Docker grows up to be a real boy

In the story of Pinocchio, the Blue Fairy makes the puppet a real boy after he has proven himself brave, truthful, and unselfish. In the story of open source software, a company becomes an real company when it can start charging customers. Docker has had paying customers, but what Docker was charging for was a little buried, as was the relationship between Docker, the open source project, and Docker, the company. Earlier this month, Docker the company announced a new product line packaging. It's splitting Docker the platform itself into a Community Edition and an Enterprise Edition. And similar to VMware vSphere, Docker EE now comes in three bundled tiers at three different price points.

1. This splits the the product formerly known as Docker into two versions: an unsupported (CE) and a supported version (EE), and bundles in formerly separate products Datacenter and Security Scanning. This does not seem to me to be an "Open Core" strategy, in which some features are held back for the paid product in a sort of freemium model. In Docker's case, the base platform is fully functional and the higher level functionality was never part of the core engine. The two editions also have two (fairly) distinct audiences: developers (CE) and enterprise (EE).

2. One of the complicating (and clarifying) details is that Docker has split out its container runtime, containerd, as a separate project, and has donated it to the Cloud Native Computing Foundation. CoreOS donated its container engine, rkt, to the CNCF as well, although it was not as well-covered. (Rkt is sliced a little differently than containerd and has a different architectural philosophy.) I'm sure there's all sorts of politics behind this and I just probably need to get enough people drunk at DockerCon to get educated.

3. Once the dust has settled, the container base layer at CNCF will make everything cleaner: all the vendors can choose an open source container engine, which is not where the money is going to be made. Now other projects like Kubernetes, Cloud Foundry, and OpenShift can use just this core container runtime instead of pulling in all of Docker, The complaints that Docker (the platform) incorporates Swarm (the orchestrator) now go away.

4. This positions Docker clearly as a platform software company and makes an easier comparison to other platforms. One of the interesting things in the space is that people haven't settled on what a container-based platform looks like yet: Red Hat, Google, CoreOS, VMware, Platform9, Rancher, Google -- heck, practically all the vendors exhibiting at DockerCon -- all have a slightly different vision of the future and how users, containers, and apps will be managed.

5. This differentiates Docker The Project vs Docker The Company, but some of the open source community are worried that CE will be left behind. In the Hacker News thread, Docker founder & CTO Solomon Hykes displayed his typical empathetic style in explaining the change in positioning.

So I'm sorry that you have a different definition of Docker than the people who invented Docker. But just like Google decides what Google is - Docker decides what Docker is.

He's not wrong, but Docker the Company could be more sympathetic to people still sorting this out, because the positioning has been tangled until now.

6. To enterprise folks, the new support lifecycle is 🤔. Docker EE comes out every quarter and has one year of support. So unlike something like Ubuntu and VMware, which release platforms with 5 years of support, Docker EE customers have one year to get their software upgraded. This now pushes old-school "War Room Enterprise," where upgrades are fraught with peril and done as infrequently as possible, out of the market for Docker. Your organization must be able to absorb quarterly releases. Only the new "CI/CD Enterprise" should apply.

7. The platform licensing is also interesting. Docker EE licenses are per-platform (AWS, Azure, RHEL, CentOS, SUSE, Ubuntu, Oracle Linux, and Windows Server) and are not transferable between platforms. If you get them through the Docker Store, they are month-to-month subscriptions. You can also get them via VARs and OEMs (including HPE, IBM, Microsoft, Canonical, Canonical, Alibaba Cloud) on a yearly basis.

8. Note you can get Docker EE on the AWS and Azure Marketplaces. Docker Store is its own Marketplace and will offer Docker Certified containers.

9. Notice Google Cloud is mentioned nowhere in this picture.

10. Expect to see a lot of stories about enterprise apps and their journey to the cloud at DockerCon in April.

Research from the Influence Marketing Council

We've started some new research as a part of the Influence Marketing Council.

The Top 50 Overall VMware Influencers. This is an algorithmic analysis of who follows whom in the VMware influencer community; not a list that I personally curated. Given the current state of Twitter, it's important to note that while the VMware Twitter community formed over 2010-2013, activity has tailed off in recent years, and once somebody follows you, they usually don't bother to unfollow. Therefore we're looking at a historical cow paths rather than current conversation. However, if you were a marketer (and that's who this list was really aimed at), most of these folks indeed have an audience in the VMware community.

All that was a long preamble to one of the obvious observations on that list: no women in the top 50, and only 4 in the top 100. Those are some terrible numbers. (The analysis is sensitive to the seed -- I ran it again starting with women with similar results.) We're working on a top women to follow in the VMware community list.

The Top 50 Google Cloud Influencers is interesting in that it's very Google heavy, although there are some external folks & Google Developer Experts in there. I discuss in the article that it's not quite fair, since Google Cloud is actually made up of lots of other offerings, all of which have their own influencers.

One of those components is Kubernetes, of course, which powers GKE, the Google Container Engine. We'll publish that map on Tuesday ahead of CloudNativeCon Berlin, It shows a very diverse, very non-Googley, view of the ecosystem.

If you're a marketer, we've got more research coming at the IMC. On our upcoming podcast, Kat and I will be talking with innovators and practitioners from tech companies to talk about how they work directly with technologists, influencers and advocates to help their companies understand their customers, and to help their customers understand their companies.

If your company might want to join the IMC, please reach out.

Always Read The Comments

I got a lot of private comments, but here are two of the public ones from my last note in January. Michael Poore:

One of my biggest issues with social media platforms is around the distraction and loss of focus that usually comes with them. Sometimes I have to reign in my usage of Twitter, Facebook etc for a little while and gradually build it back up to a comfortable level.

Recently I had a brief conversation with Ray Heffer and Grant Orchard (that took place some what ironically on Twitter) about social media usage and a book that both Ray and Grant recommended called "Deep Work". I'm still reading it, but they both suggested that it might make me re-think how much I use social media platforms.

For me though, things like Twitter are still a valid way to engage with people across the IT industry. In some cases they could be potential customers so I can't ignore Twitter altogether. However that does mean that I ought to behave! And that brings me to the other issue that I have with social media; some folk who are nice in person all of a sudden turn in to badly behaved and often unpleasant people for some reason.

Roger Weeks:

Below is a relevant extract from author Warren Ellis's last newsletter, about social media. I have done very similar things: I have deleted my entire tweet archive, walked away from Facebook, and deleted my LinkedIn account. I don’t see the benefit of these services any more, and I think they do more societal harm than good. In my own life they are a horrible distraction and I won’t be returning to them in 2017.

This may prove to be a year without social media for me. I have a private Instagram account and I'm bored even with that. This may be the year I drop off entirely. Which is tough for an old nethead like me. But it's way past time to step out of the room.

I already got rid of my LinkedIn account - too many journalists were saying that they were being asked to give up their account URL at US customs. LinkedIn, of all things. The day after the US election, I nuked my entire Twitter archive. A month later, the US instated a demand for one's Twitter ID for entry to the US. Currently "optional." Won't be for long.

I strongly urge you to take half an hour this week to take a look at your social media footprint and decide what you don't need and what you can live without.

If you're on Windows, then Windows Archive Eraser will blitz almost everything on one go.

Otherwise, I suggest you consider becoming the last glowbloggers on Tumblr until it gets shut down. Glowbloggers only ever said "#glow." You'll be fine.

Atemporality and the great pause is over, folks. This is the acceleration, and it's going to make the Nineties look like The Enlightenment in comparison. Things are in the saddle now.

Worth a Click

Going with the topic of Docker, this is a fascinating post on the internals of From dotCloud to Docker by Jérôme Petazzoni on his blog. If this kind of thing rocks your world, check it out and Docker's architecture may become more clear to you.

In theory, any good CS grad student will tell you that this seems like a perfectly good case to use some bin packing algorithm.

In practice, anybody who has worn a pager long enough knows that network latency and packet loss are both non-zero quantities, and that therefore, we are facing a distributed systems problem (aka potential nuclear waste dumpster fire).

Most “standard” algorithms assume that you know the full state of the cluster when taking a scheduling decision. But in our scenario, you don’t know the state of the cluster. You have to query each machine. The request has to go over the network, and then the machine has to read the state of all its containers before replying. Both operations (network round-trip and gathering container state) can and will take some time. Using aggressive timeouts (to avoid waiting forever for unreachable nodes) gets problematic when a host is very loaded (and takes a while to gather container state).

The million dollar engineering problem by Achille Roussel and Rick Branson on the Segment blog. Here's why I'm linking to this. A lot of you are IT people. You may think to yourself, "Oh, woe is me! Those San Francisco startup hipsters have green juice and MongoDB and skinny jeans (they really do) and all I have is this steaming pile of old applications and comfortable cargo shorts. They are special and I am not!" Fear not, brave reader. Read through this post and ask yourself, "Do I have exactly the kinds of troubleshooting and infrastructure experience that, with a little bit of AWS training, could have totally helped these kids figure out how to save a million bucks on AWS costs?" If the answer is yes, get some AWS training. You don't even have to join a startup.

I'm not a Woman Engineer, I'm an Engineer, by Anne Jeanette on A Cloud Guru's medium site. Her experience is great and all, but see all the extra work she had to do to get the respect of the men she was working with? And kick ass at her job at the same time? Stop being such man-babies, fellow men.

In time, they began to lend their credibility privilege in conversations with other techs who might be anxious about having me in their future rotation. The techs now invited me to lunch and included me in diagnostic conversations. They vouched for me and my work not because I was a woman, but because my value was indisputable.

Why Is Silicon Valley So Awful to Women by Liza Mundy in the Atlantic

Such undermining is one reason women today hold only about a quarter of U.S. computing and mathematical jobs—a fraction that has actually fallen slightly over the past 15 years, even as women have made big strides in other fields. Women not only are hired in lower numbers than men are; they also leave tech at more than twice the rate men do. It’s not hard to see why. Studies show that women who work in tech are interrupted in meetings more often than men. They are evaluated on their personality in a way that men are not. They are less likely to get funding from venture capitalists, who, studies also show, find pitches delivered by men—especially handsome men—more persuasive. And in a particularly cruel irony, women’s contributions to open-source software are accepted more often than men’s are, but only if their gender is unknown.

Let's Catch Up

Kat and I will be at DockerCon this April. I'd love to talk with you -- let us know if you'll be there. We'll also be at OpenStack Summit in Boston in May.

• Geek Whisperers #128: Engineering your time at conferences with Aaron Delp. Aaron approaches a conference like a geeky general.
• Geek Whisperers #129: Trials and tribulations of a technical generalist for hire with Dave Henry. Where do you try to land in the org chart if you're a jack of all trades?
• Geek Whisperers #130: Experiences with side gigs with Mike Olson. When side gigs are right and when they're wrong.

Subscribe to the TechReckoning Dispatch. If you like this newsletter, forward it to someone who needs good ass-kicking. For more groovy links, check out TechReckoning on Twitter, Facebook, or LinkedIn. Photo credits: John Troyer & some jokey tweets about microservices.