Eric Chou, Author of Kafka Up and Running for Network DevOps: Set Your Network Data in Motion

Len: Hi I’m Len Epp from Leanpub, and in this episode of the Frontmatter podcast I’ll be interviewing Eric Chou.

Based in the greater Seattle area, Eric is a network technologist who was worked on some of the biggest networks at some of the biggest companies in the industry, including Amazon AWS and Microsoft Azure.

You can follow him on Twitter @ericchou and check out his website at networkautomationnerds.com, and you can find his Network Automation Nerds podcast at podcast.networkautomationnerds.com.

Eric is the author of the Leanpub book Kafka Up and Running for Network DevOps: Set Your Network Data in Motion.

In the book, Eric shows you how to get the Apache Kafka data message system up and running, helping you to expand your network engineering toolbox.

In this interview, we’re going to talk about Eric’s background and career, professional interests, his book, and at the end we’ll talk about his experience as an author.

So, thank you Eric for being on the Leanpub Frontmatter Podcast.

Eric: Hey, thanks for having me. I’m excited to be here. I believe in the platform, so it’s very nice to meet you over - virtually.

Len: Yeah.

Eric: And it’s good to be here.

Len: Thank you very much for being on the podcast. I always like to start these interviews by asking people for their origin story. So, I was wondering if you could talk a little bit about where you grew up, and the curious way that you found your way into a career in network engineering?

Eric: Yeah, for sure. I grew up in Taiwan. I immigrated to the US when I was 13. So that’s roughly like 7th grade, 8th grade. I got my education. That was in Southern California. And so - technically, I started ESL, right? For people who don’t know, it’s English As Second language, typically for new immigrants coming in. That’s where they’ll start.

So I never thought I would be writing books in English. That was quite a jump. Even though I’d been in the US longer than I was in Taiwan.

But it’s stil - language is something that you’re emotionally connected to. And especially when you’re younger - if you just kind of ingrain yourself, your brain - that’s how your brain functions. Especially for such different languages, like Chinese and English.

So, for a long time, I had to translate something I heard into Chinese in my mind, and translate back, and then respond. Gradually you graduate into a state where you just respond naturally and intuitively.

On the flip side, that means your Chinese is progressively decreasing in efficiency. I’m sorry my peanut-sized brain just can’t take two languages at the same time.

And, but anyways - yeah - so I grew up in Taiwan. I immigrated to the US. I studied business and finance back in college. So, really far away from technology. Although, because I was in college during the later part of the dotcom bubble - I don’t know if you remember, but it was circa 1996, 2000 - where everything was just crazily over-valuated.

And so, I decided to say, “Okay, if this finance thing doesn’t work out, then I’ll do a minor in a computer-related field” - not knowing anything about it. I picked a field that is obsolete right now, which is like multimedia web design. If you know this company, Macromedia - or FrontPage, that sort of stuff - that’s what it was. That quickly got out of fashion, out of date. But it did give me the opportunity to go intern at a local ISP - an internet service provider - at the time, and that’s what got me into -

I was in support, so that got me into a broad view of customer service, how people connected to the internet.

One thing led to another. I moved to a bigger ISP, launched some of these markets in - for cable broadband. And then gradually into hardware. And as you mentioned, I got into some of the top cloud providers in the early days, which was really eye-opening. And here I am, 20 years later, still in the - I still haven’t worked a day in finance. So that’s kind of in a nutshell how it came about.

Len: Just so people get a little bit of a sense of the kind of - because your career, what you do in your job changes so much over time throughout your career.

Eric: Yeah.

Len: Because you’ve been there for this really interesting - I mean, pick any 20 years in the development of these kind of telephony and kind of data communication networks, and you’ll find an interesting 20 years. But you’ve been in the industry for a very interesting time. And you got into it at a very interesting time as well.

I just wanted to mention that - it’s sort of like - people from California who get into kind of network engineering and stuff like that, from different paths - this has come up on this podcast a few different times with guests. One guy, speaking of multimedia - he ended up sitting on top of the Space Jam website.

Eric: Oh wow.

Len: For Warner Brothers. And that was because he was in LA or whatever.

Eric: Right.

Len: One of the artifacts of the sort of history of, “What do you think of the Web?” is like, how big and how important California was. And often we say “Silicon Valley,” but it was like - this was a California thing, partly. And it was an experience that was shared by a lot of people. And to be - I mean, particularly - there’d be a lot of people who actually maybe got their degree in the specific area that they wanted, that was hot. And then the dotcom bubble burst, and they were like, “Oh my God, now I’ve got to switch to something that I didn’t study anyway.”

So that leads me - before we get into the details of what you maybe did at your first job, and then talk about how that’s changed over time, I was wondering if - this is a version of a question that comes up all the time on the podcast - if you were starting out now, with the intention of having the career in the same area that you’ve had your career in - would you do a four-year formal university degree, or not?

Eric: Wow, that’s a good question. I have two young daughters, and that’s something that me and my wife talked about, right? Does college education still - place as much value as it does, right? Because in my opinion - Let’s just go back and say, “What’s the value of a college education?” Right? Because that’s essentially what the question is. What kind of value does a four-year degree give you?

So I think first, is branding - for sure. Like if you go to an Ivy League school, right? You a have a Harvard MBA, Harvard degree - then that’s branding, right? If somebody has evaluated you and made sure you meet all the criteria for that. So that’s branding. You’re associated with the good brand.

And the second is connections, right? You’ve met the people that, perhaps, you were associated with later on in life.

And the third is, the technical knowledge that it gives you. Whatever major you are. And that goes along with like life skills. How to manage your time, how not to oversleep and not get drunk the night before a midterm, right? So those are, I think, three things that give you those value.

What I would say to that, is - knowing what I do now - I think, I will look back and think about whether I acquired that knowledge. I didn’t go to a Ivy League college, right? So first, branding - that’s out. I didn’t really associate myself with people whom I ended up working with. And that’s kind of iffy, right? That didn’t pan out for me. So the third thing really is the technical knowledge. And that, I didn’t even do that, as we mentioned before.

I think that’s a 50/50 chance. And because I didn’t grow up with a lot of money in the family, it took my mom basically all she had, to just put me and my sister through college. So I probably would not have gone to college, doing what I do now, knowing that I could still be disciplined enough to get to where I am. So most of what I - it was web design, but I ended up in networking. In order to get to a proficiency level, I have to do a lot of self-studying.

Once I acquired enough funding and financial backing, I actually went back to a certificate program to study Python. That is where the coding part comes into play. That is why the first book that came out, and so on.

So I think, to circle back to your question, I probably would not have, if I knew that I had the discipline to acquire that knowledge myself, on my own. But that is a big “iffy,” right?

Then the question becomes, “How do you know?” You just don’t, unless you start go and do it. So - yeah, for my own personal background, I probably wouldn’t done it.

But having said that, I still put aside money in my 529 for the two kids. If they decide to take a gap year, fine - right? But at the end of the day, I want to give them that opportunity. They don’t have to take it, but they will have the option to go to college. That is my approach now. Ask me two years later, I might say something different. It’s like, “Yeah, sorry Junior - I had good night in Vegas, and it’s all gone. I guess you just have to get some scholarship.”

Len: Thanks very much for sharing that. That’s really great, and it’s a sort of spontaneously structured answer, which are the best answers - in my view.

Eric: Yes, glad you liked it.

Len: Very much. I can say - just speaking from my own experience, having gone both to a non-brand name, and to a brand name university - it has a powerful effect on you. Not just on other people’s perceptions of you, but on you yourself, to go to a brand name university.

In my case, it was Oxford, for my doctorate. One of the things about it that it’s very - you can’t help but think you deserve the special treatment.

Eric: Oh wow, really? Okay, that’s interesting.

Len: I mean, I would say that’s true. I definitely confess that’s true of me. I think it’s true of basically everyone I know. That you just kind of - it’s like anything - you think you deserve it once you’ve got it. It’s very easy to fall into that mental trap. So although going to like an Ivy League university or other brand name university - it has this profound value that stays with you your life, at the same time, you need to - if you want to be a good person and realistic about yourself - you need to have an inner voice going, “This was largely accidental,” and don’t fall for the hierarchical stuff like too much.

Eric: I think you should give yourself more credit than that, of course - right? I’m pretty sure there’s some sort of - little bit of entitlement, but also - you do deserve it, right? It is a hard thing to achieve. Just like anything else, you put a check mark next to it - and then you move on.

But I think you brought up a good point, this time value of education, right? If you had your education in your twenties, that means your payback period - your compounding period, is a lot longer than somebody who got it in say their forties, right?

Nothing wrong with the time spent, it’s just like - these are all finite resources that you decide to spend your time on. Your opportunity cost is different. Like I said, I think the biggest issue is your compounding, right? If you’re starting - if you’ve gone to this particular university and you could get 20% or 30% more in your base salary - stretch that over 20 years, it’s very different than stretch out over 10 years or 20 or 15 years. So that’s another good point, where you brought up about investment.

Len: Yeah. It’s very interesting, that point about compounding. Because one of the things I like to say about education, true education, is that– the thing where you’ve spent years out of your life, focusing on reading and learning and things like that, is that it gives you a power that has to be possessed in order to be perceived. This is why the idea of like - I use the word again - a hierarchy of people who are educated, and people who aren’t - is so ugly and ancient, right? It’s just because you have to have spent those years doing it. To finally go, “Oh man, I see the world completely differently now.”

Eric: Right.

Len: You really do, if you’ve got the privilege to have the time and the resources to spend years in your youth in study - it stays with you forever. Do you know the movie Braveheart?

Eric: That’s one of my favorite movie, man, “Freedom.”

Len: One of my favorite scenes in that is when Uncle Argyle is going to take young William Wallace off. William Wallace, who’s now been orphaned - right?

He sees his Uncle Argyle’s awesome sword, and obviously he wants to learn how to fight. Argyle pokes him in the forehead, and says, “First I’m going to teach you how to use this,” meaning, “your mind.”

“Then,” he holds up the sword, “I’ll teach you how to use this.”

Eric: Right.

Len: It might seem that learning, spending years with your nose in books is not a very martial way to spend your time. But actually it is. So I guess we’re getting maybe a little bit derailed here. But talking about education -

Eric: No, no it’s–

Len: It’s important stuff.

Eric: It’s related, right? Because what you’re doing with Leanpub really is about democratizing an education and bringing that meritocracy up. Just because you’re eliminating the gatekeeper for publishing, which is awesome.

Of course - and I didn’t know that about, you liking Braveheart. I used to play around with my friends, and just try to mimic - like an Asian guy trying to mimic like a Scottish accent, so that’s pretty funny. But it is one of my favorite movies. Just because I think - I admire people who believe in something so firmly that they’re willing to give their lives for it. Not just their lives, their everything, right? So I think the movie came out, what? 1995 or -?

Len: That sounds right, yeah.

Eric: Yeah. 1995 or so, I think they won “Best Picture,” or at least were nominated for it.

So at a young impressionable age for me, looking at that movie - I’m just like, “Oh my God, I wish I could find something that I believe so much, that I could dedicate my life and my whole being to it.”

I’m still searching for that. But I admire people who have found that calling.

Then back to what you were saying about having the right mindset, having the right discipline - that accumulates over time. That is a very good value for it. Hopefully you start developing that while at a young age, so you compound it further. Because that’s a macro skill that applies to everywhere.

Len: Yeah, totally true. That’s also actually funny, that people often describe their experience going out of high school into the military that way as well.

Eric: Yeah.

Len: I learned how to be disciplined. It’s often like - we often think of the sergeant barking at you, which I’m sure there’s a lot of. But you’re making a years-long commitment when you do that, in a much more serious way than sort of going to college. Which you can drop out anytime if you just feel like it.

And that learning - in addition to what I was talking about, in my romantic way of talking about the power of education. The power of discipline, and spending years disciplining yourself when you’re young, can have this compounding effect throughout your life as well. So those are twin, very important things to keep in mind, when you’re making those decisions, when you’re starting out.

Uou started out, your first job that I see here on LinkedIn was for Time Warner Cable, starting in 2001. I was wondering if you could - you said you did a lot of learning, you had to do a lot of learning on your own, I imagine on the job, and things like that. What was the work you were doing as a network engineer in the early 2000s? What did you do in a typical day?

Eric: The first job I started was actually through a local ISP. But unfortunately, just like a lot of ISPs back then, they were belly up. They went bankrupt and no longer exist. I think I had a hard time adding that to LinkedIn. But in actuality, I started there, I want to give them credit. It’s called GUS Networks. They gave me a start as a young person who really didn’t know what they were doing. I was able to contribute and make a significant, meaningful impact to their business. That gave me a tremendous amount of confidence, to a young engineer, especially someone who did not study Computer Science, or didn’t come up the traditional way. Really, I would say, it made a lot of difference working for GUS Networks.

One of the things - I know you had other guests who went through a Cisco route, in particular Nick Russo, and some of the books that he published on CCIE, CCDE.

I went through the same route. Cisco Systems provides this great certification path. Back then, Cisco was the way to know. I think at that time, they had something like a 70%, 80% market share for enterprise. For service providers, a little lower. But still significant.

As long as you know Cisco, then you’re probably guaranteed to get a shot at a job, right? But it may not be the highest position available. It may not be the job that you want. But that is their path. It was an education, but it was like a Cisco - through a Cisco lens. You get certified through that, right?

While working at the ISP, I got through the entry level, which was CCNA, like associate level. Then I had to go for the professional level, which was like four additional tests. Which lends me enough credibility to land a job at Time Warner Cable.

Time Warner was also one of the big companies that gave me a large scale exposure. I don’t know if you remember, but in 2001, that was still a dial-up world. Broadband was just rolling out. DSL and cable modem were - they’re the two competing technologies that they were rolling out. They were actually head-to-head. They’re really competing with each other to do the rollout and get enough coverage to cover United States or England, or other developing areas.

So, Time Warner Cable - at the time, they were trying to convert their cable plans into cable broadband ready. They were looking for young engineers to relocate to other areas that might not be as desirable. Even though I was in California - California is an interesting state. I guess some of you will probably be - I think last counted, it was like the eighth-largest economy in the world - even by itself, right? If you just separate it out. Very diverse, what you were taught.

Even relate it back to what you were talking about earlier. You have Northern California, which is very tech-focused. Very, I don’t know? Hip-ish, if you will? Then Southern California - which is a lot of media companies, a lot of diverse, different industries. Then the middle part, which is very much agricultural and farming - that stuff.

Anyways, so in California, Time Warner Cable was looking for people to go to the the Palm Springs desert, to launch like the cable broadband. Not everybody wants to go live in the desert. But at the time, I was brave enough to raise my hand - even without ever being there. I just went out. I got my large scale networking exposure.

For people who don’t know what network engineers do on a day-to-day basis, basically you gain your knowledge by doing more and more - more complex work, on a larger scale. Build out networks.

That’s how note your street cred, right? Of course, you study your books and you get your protocols and you get your exchanges - you’ll get your book knowledge. But the way - at least at the time, without virtualization - the way that you gain your credibility and apply those book knowledge, is working on more complex and larger networks. That stayed with me, to go to Time Warner Cable.

Then eventually, I was working on large scale data networks for Amazon in the early days. You’d get the hockey stick growth for Microsoft Azure. Sorry, you hear my Amazon assistant going in the background. But, anyways. Those are the things that I got into, and just stay with me.

Len: For example, when you were working for Time Warner and they got you this job in this less desirable area - were you going to a big building, and racking and stacking servers together, and things like that? People might’ve seen in the movies these big rows of columns of basically computing machines.

Eric: It was even more basic than that. You see these servers. But these servers need to be connected somehow, right? What I was doing, was actually connecting residential areas for people. Instead of dial-up - I mean, it sounds so ancient at the time. But it was actually at the time where people - when they needed to get on line, they need to plug in their phone line and dial into a service provider via your RJ11. You see these modem like, “Do, do, do, do, do.” These were like the last generation.

In order to have it always on, Time Warner Cable would have this HF [high frequency] signal, right? Which is - the long haul is through fiber, which was very clean. They spent probably 10 to 15 years laying down these fibers. On the last mile, they would have these coax cables that come into people’s home, mainly for a TV, because the coax could actually carry a lot of frequencies, and so on. They tried to dedicate a frequency for your downstream network. What they did was slowly, gradually converting these networks, to have an upstream bandwidth.

That is why when you’re talking about both DSL and cable modem, you’re talking about asynchronous. But your download speed usually is a little bit different than your upload speed. That’s the difference.

Most of these, I would say - I’m not as familiar with DSL. But I would say, both the DSL and the cable network were meant for downstream traffic, to carry over signals that way. Then, because of the need for the internet, they start upgrading the upstream.

What I was doing was gradually converting these, node by node, what they call “node by node.” Each node covers like 500 homes, and so on. You just go in and there’s some different calculation.

It’s a little different than that. We’re doing these for the residents. We would go to these nodes, and we’ll certify them and say, “Now it’s internet-ready.” People could go out and sign up for broadband for their $29.99 per month on the cable broadband. We’re doing that node by node.

As a young engineer, it was really great exposure. Because when I arrived at the market I was at, I think there was like - I don’t know? 40 people who were like internal employees, who were just trialing this network. But by the time I left two and a half years later, I think it was like 24,000 users, or something like that.

You get that exposure, and it’s something that you don’t really get to do very often - is to launch a market, and see that growth period going up. So it’s much lower than the data center of - a bit different than what I’ve done later on. That is exactly what you were talking about. Those data centers are just rows and rows of servers. You’re worrying about connecting them most efficiently, and how to do - the question becomes, does the server talk to other servers most efficiently? As well as, how does the server talk to the end client in a more, I guess, cost-efficient way?

Len: So you were you like the cable guy’s cable guy?

Eric: Yeah, that’s a very good way to put it. The cable guy would probably - in terms of connecting you, I was in the back end, connecting all the cord together. But in order to do my job, you don’t necessarily have to - but I do go out with the cable installers to see how they actually install the wires. How they clamp it.

What are some of the interferences? That was the biggest thing for us, is these interferences - when, in the traditional network - when it’s all just from your head-end, going down to each of your customers. You’re able to control the source to be really, really clean - without interference into each of the residential customers.

But when you start talking about like upstream, the signal comes from the customer, back into the headend. It’s like a reverse funnel. All these signals will be coming back. If you turn on the microwave, those are some interference that comes back - if you turn on your vacuum even, right? Those are interference.

If you have a piece of cable or fiber, for example, that is damaged, that actually aggregates back. Because you’re over-subscribing a bunch of customer to one node - that one node, when it goes down, it impacts all 500 customers. If you trace back large enough, it’s going to impact a lot more people.

Anyways, I guess my point is that it’s an interesting world that I was in. It gave me the exposure that I would not otherwise have.

Of course, it wasn’t all unicorns and rainbows - it was very harsh. People were banging on the door trying to get internet, and you just have to turn them away. Which was, I guess, a good problem to have? But at the same time, it was a problem, nonetheless. We couldn’t build it fast enough for our customers.

Len: Speaking of things being hard and stressful - and then, once they’re hooked up, it better work 100% all the time, or else they’re going to be banging down your door, saying, “why is it down?’

Eric: Oh, 100%, right? Also, the triple play was huge at the time, which was - you get your TV, internet and phone services from your cable provider. In order to do that, you get into the whole other territory. Your 911 calls, your emergency battery backup.

What if the power goes out, right? The phone - for those people who doesn’t know, the phone is supposed to still work. Not the wireless one that you require it plugged into power. But the ones that are - those old handhelds, even if you don’t have power, you’re supposed to be able to plug in your phone jack, and you should still be able to dial 911, regardless of whether you have a phone service or not, right? You could be in an abandoned building, and you should still be able to dial 911.

Which was these requirements. As a cable company, it’s great to have the business. But at the same time, you’re under a different set of rules and regulations, in order for you to offer that third triple play service.

So what you talk about as uptime and people’s expectation - lucky for us at the time, they were so happy to have broadband internet - I was able to have a maintenance window, which I think for me at the time, it was between 3am to 6am.

Anything, any outage that happened within that window - it doesn’t count toward our service level agreement. In other words, we don’t have to pay refunds for customers.

But it’s a different world now. Can you imagine now you tell people that, “Hey, at any given time your internet may be out for three hours and we’re not going to do anything about it.” That’s just not going to fly. But back then, it was okay - people’s expectation were different back then.

Len: It’s so interesting, the idea of expectations and things being serious and difficult more generally. I imagine one of the things that must be frustrating sometimes being an engineer, is that like everybody respects doctors, because they know what doctors do. They have at least an idea of what doctors do, right? They don’t understand the details. But it’s like, “I’m hurt, you fix me.” Everyone knows what teachers do.

But particularly when it comes to things like computers and computer networks and things like that - people are like, “It just works.”

Eric: Yeah.

Len: It’s like, that duck is paddling really hard under the water there, and that duck really knows their stuff. I just want to bring it up in that context, to say - because you brought up the Cisco certifications and things like that earlier.

Eric: Yeah.

Len: Those things are incredibly hard. You have to learn a great deal, and you have to be tested very vigorously - precisely for the kinds of reasons that you’re bringing up. It’s not, as it were, just customer expectations, although at scale you’re talking about whole populations of people. It’s 911 networks that better work. If the power goes down, that better work.

The people who are setting these things up, designing them in the first place, the people who diagnose and maintain them, and the people who are brought in to like, “Well, how do I know where a fault is somewhere between point A and point B?” These kinds of things are not just really hard in themselves, but to do under pressure, and have to be hardened. That’s a lot of the work.

I just wanted to make sure, for everyone listening - it’s not like “just getting a Cisco certification.” This is very, very, very difficult thing to do.

Eric: I appreciate being appreciated. But back to like the episode with Nick, right? Nick comes from a military background. What you guys talked about, I feel like Nick should pay me some commission for - He’s a friend by the way -

Len: Oh okay.

Eric: For mentioning his podcast episode so much. But anyway, he came from a military background. What he talked about was, when you have nothing, you need to drop into the desert and provide connectivity for the soldiers to have secure communication back home - to both their family, as well as to their colleagues, and receiving intelligent information - not being intercepted by the enemies, and so on. Those are different areas - as you mentioned, right? That’s a different service level or security requirements, and all of that.

You’re right - as far as the Cisco certification, and it’s the same for a lot of vendor certifications. The degree of difficulty goes incrementally.

At the time when I was going for - they call, “the expert level exam,” it actually took me five years. I had to go learn a bunch of stuff. I had to go sit in a lab that they provided, and for eight hours, to prove that you’re able to solve anything they throw at you.

And then, at the end, they give you a number - and say, “You’re now–” If you pass, right? They give you a number.

My number was “22460.” Supposedly that number is never going to be given to anybody else. It goes to the grave with me. That’s the effort that you have to put in, in order to achieve that certification.

And, yeah, you probably use - like anything else, you use like a fraction of it in real life. But hopefully that fraction is something that you studied before. If not, then you learn on the job.

Len: Thanks very much for sharing that. It’s super interesting actually. It leads me to ask about the - next stage of your career, as it were - I guess?

Eric: Sure.

Len: Was also one where you experienced this great growth, which was at Amazon and AWS. You started working there in 2006, which is forever ago, in AWS terms.

Eric: Right.

Len: I was just wondering - for people who are familiar with this, and might want to hear a little bit about what it was like on the inside - can you talk a little bit about just what it was like being in that environment for those years?

Eric: It was very interesting. I want to say I was very lucky to be able to ride that first wave, which was the broadband wave. If you think back at 2001 time frame, during the dotcom bubble, where the bubble burst - I was still lucky enough to have a job, because I was in a growth area within Time Warner.

Again - in 2006 - I stayed at AWS and Amazon during the financial crisis, right? That was the second crisis, and I was lucky - so lucky to have a job, because I was in such a high growth market. It’s no merit by myself. It was just basically be at the right place, at the right time.

But, yeah, you’re right, 2006 was when I joined Amazon. It was very much a retail business, right? You get your three percent margin. They’re very cost-conscious on that front. If I remember correctly, it was - they barely got out of the dotcom fiasco, basically - and I think that our fierce leader at the time - Jeff Bezos - he was hurt so much by that, he rarely gave interviews, after he was being named “Time Person Of The Year,” in, I think it was 2000? It was very much a retail, low-margin mindset.

Then they launched this small service, called S3 - back in 2006, like March of 2006.

I joined August of 2006. But it never really took off until the year after, which they start offering EC2 - which was like a virtual machine. Basically you render compute power. It just took off by then.

We knew we had something special. But we also knew it was so young, and so fragile - that if you don’t take care of it, it’s just going to die in its infancy. I think a lot of us back then were on this mission of bringing cloud computing - which was the term they later coined - but bring utility computing, cloud computing to the world. We were just - it was a very exciting time, but it was also working really hard.

None of us knew what we were doing. We were given a lot of leeway to do a lot of stuff that we haven’t tried before. For myself, I was part of the original team to launch CloudFront, which was a super great exposure to be able to survey around the world, and see which area shall we go next?

Again, growing a business from the ground up. That’s another area that gave me great exposure. Being a founder yourself - you know going from zero to one, it’s super difficult. Just as from one to scaling up. Len: Yeah, it’s really interesting. I was just - before we started this interview, I was listening to one of your podcasts. You talked about going to an all-hands meeting at Amazon around that time, when Jeff Bezos made this comment. I’m really glad you actually teed up this question, because I was trying to find a way to ask it.

But, about the vulnerability, right? That, and the sense of caring for something new, and the threat of low margins.

By the way, none of this is to praise Jeff Bezos or Amazon in any particular way. But that company, it is what it is now, because it went through a great deal. One of the things that Jeff Bezos said at this meeting that you attended, was, “Good intentions–” I think it was either - it was like something like, “Good intentions don’t -“

Eric: Yeah, “Good intentions don’t work.”

Len: Don’t work. Right.

I was actually going to say, “Good intentions don’t matter.” That was me being uncharitable. My instinct, which is why I was checking myself before.

But like, “Good intentions don’t work.” I mean, you might - how do you get through what Amazon went through - having a little bit of a background in finance myself, years of unprofitability - especially when you’re a public company doing an extraordinary thing, right? Like, “We’re going to lose money for years, trust me.”

Eric: That’s a hard thing, hard pill to swallow for investors - for sure.

Len: He really did have to prove it, right?

Eric: Yeah.

Len: And I mean, like the whole - everybody who worked there had to prove it, right? What does - in your view, what does “Good intentions don’t matter–?” Don’t matter, I did it again. “Good intentions don’t work,” mean?

Eric: Yeah. I think I was with - just to give a little context, for people who did not listen to that episode. I was in this group called “Tier One Services.” Our group’s main responsibility was to make sure the website stays up. When it goes down, we contact the right group to fix it. Just like a lot of companies, the services are classified into tier one, tier two, tier three. What classifies tier one, is usually things that, when they break, you don’t make money.

Like networking, for example. If the network breaks, you don’t make money. It’s very crucial. That’s a question that was always on our mind, as a team. It was a brand new team too. I think I was maybe one of the first network engineers that was hired into the group?

For background information, I was really thinking about, “What does it mean to be an American junior in the tier one service, and how do we monitor -?” Then they were, “How do we scale monitoring to a point where people still have a quality of life, a reasonable life - but at the same time, we could still keep a certain SLA going?” When I attended that meeting, those were the questions always in my mind.

When Jeff came out on the stage and said, “Good intentions don’t work,” he took his time to explain why good intentions don’t work.

People have good intentions. But when it’s 3 am in the morning - you’re tired, you’re not thinking right; or if it’s like Friday night, and you have something else that you want to attend to - human nature prompts you to take shortcuts. Or human nature prompts you to try to fix things as soon as possible, so you could get onto it. There’s nothing wrong with that, it’s human nature.

What he ended up saying, was, “What works is a system, an automation, that works. You want to rely on the computer to solve your own problem about human nature - being, we’re tired, we’re not consistent. When we try to do things for the thousandth time, we tend to make mistakes. What works is really - you offload those things to automation, to computer-assisted work. You free yourself up into the higher level of tasks that’s meant for human brains, such as designing your network the correct way.”

With that in mind, that just hit me really hard at that moment. It’s like - you could almost go back to several moments in your career where you’re like, “Oh my God, that was -“ Looking back, you know that was a moment that hit you.

For me, even if I talked about learning Python, writing books on Python or Kafka - all these other technology stuff, right? These are just tactical stuff. Fundamentally, that changed my mindset about which direction I should go on - what actually works?

Back to your question. He said, “Good intentions don’t work. What really works is automation. Realizing your own shortcomings as human beings. Find the right tools and the strategy to help you solve those problems.

Len: Which actually is another great segue that you’ve provided me with, or opportunity for a segue, thank you, which is to your first book, which you published with Packt, which was about automation.

For those who aren’t aware, it’s called Mastering Python Networking: Your one-stop solution to using Python for network automation, programmability, and DevOps. I was wondering if you could maybe talk at the same time about what you mean when you’re talking about automation there, with Python and things like that - how that really changes, how that’s a part of your journey as a network engineer. Learning how to scale through automation, and things like that. And maybe tell the story of how the book came about at the same time?

Eric: Yeah, for sure, thanks for asking that question.

I think in order to know the reason for writing that book, you go back to the days of Amazon. You’re growing super-fast. You’re barely keeping up. I said, it wasn’t always rainbows and unicorns, right? We were dying. We were just - we couldn’t meet the demand fast enough. With the amount of projection that we’re saying, “We need X amount of servers. We need X amount of data centers.” Everybody was racing toward that goal.

I think Andy Jassy, who’s now the CEO of Amazon - at the time he was the head of AWS. He said, “When elastic computing become inelastic, that is when the industry will die.” Your whole premise is built on that you’re - when somebody pays the credit card and turns on the X-Ray bucket, that X-Ray bucket’s got to be there, right? Or the EC2 instance has got to launch. Your elasticity depends on a solid back end, that meets your peak demand. That peak demand is a shifting target. That shift is dramatically different than what I was always used to.

Talk about Time Warner Cable - that projection was very linear. You know at a certain time you’d do what, and that’ll get you over the next hump. Your resources - you plan ahead - if you’re digging up some trenches, you’ve got to go apply for the government, and that’s three months of paperwork.

But it’s very different with Amazon. The scale was just a different scale, at all. Networking-wise, that becomes a very big challenge. In order to provision to - there’s some stuff that you cannot automate, right?

Rack and stack - you’ve got to physically put the box there, you’ve got to physically connect these cables. But if you think about provisioning, if you think about, “Okay, now the racks are there - now I need to connect these boxes together, provision the servers, the vlans, the port speeds - and all of that. Those could be automated. That is when you really - in order to survive that growth, you have to be able to use computers again to manage these boxes.

Unfortunately, at the time - the mindset of the networking industry was much more geared to a human managed network. The devices fundamentally are not automation-friendly.

Now you have these contradicting priorities, right? The vendors would give you one set of things, expecting a human on the other end to manage it. But at the same time, you need to be able to manage these boxes automatically. At least once the extendor’s been provisioned, and now you’re just rubber-stamping on the same notes.

That is where I started to think about, or - for survivability, really, “How do we provision so much of these devices in such a short amount of time, with the amount of people that we have?”

That is he path and the growing pain that me and a bunch of other engineers went through within Amazon. This extends also to other organizations as well, right? You talk about Facebook, Yahoo!, Microsoft. All these other companies who needs to provision these servers fast enough for the network.

I went through a period of time thinking about that. Eventually - I don’t want to say we solved that problem - but at least we have a fighting chance of solving that problem with Python, and with some of the newer packages and so on.

After a few years of doing that, I looked around - and of course, there were no resources back then to learn from. I made a lot of mistakes. We made a lot of mistakes. I poked myself in the eye constantly. And, so, a couple of years later, I started writing blog posts about these bits and pieces that I learned along the way. I started thinking about what I would - “If I know what I know now, what would I tell myself - younger self - when I was just starting out?”

Because I see a lot of other engineers trying to come down the same pathm, essentially making the same mistake that I made, or in the same boat of being confused and not knowing - oh, “What’s the scope? Why are we even doing it?”

That was my persona. I was writing that book for my younger self, and saying, “It’s okay, don’t panic. Here - the stuff that I wish I had known,” and put into a book format.

The second question was, “How did the book come about?” Like I said, I started writing blog posts. These are just notes for myself, really. I’ve taken a lot of notes, and I converted some of those notes that I think are useful for others. At the time - I think one of my blog posts about open flow, which is - at the time, it was a newer technology - and not a lot of resources out there, started ranking pretty high on Google search, if you do like, “open flow for POX controller,” I think it’ll rank pretty high. That got the notice of some editors at Packt. They reached out, and the rest is history.

Len: And was it very exciting to get approached by a publishing company? We normally talk about that stuff mostly at the end, but it must’ve been exciting to be approached by a publishing company.

Eric: It was exciting, but it was also - I didn’t know what the– I didn’t think that they knew what they were talking about. When you talk about Python networking, you’re typically talking about like sockets and servers and little protocols within - about Linux kernel, about networking, versus network management. I had to go a double take and say, “Are you sure this is what you were talking about, and this is what you want me to write a book about? Because that’s all I know.”

They said, “Yeah, that’s fine.” They confirmed it a couple of times. Then I started saying, “Okay, great.” I didn’t know about Leanpub back then. But I would’ve probably still written a self-published book, even if they hadn’t approached me. Because that’s just something I wanted to do.

But, great - now Packt was willing to give me a chance to do it. I wrote the book really, really fast - before they realized they’d made a mistake. So, yeah, that’s what it was.

Len: So you’d been writing blog posts, and then you published this book. Was that around the time that you started to make courses and things like that, and become a content creator on - I’m going to say, “on the side,” because you have always had a job the whole time, but you do make books and courses and blog posts - and podcasts now as well. Was there a point in time when you decided that this would be like a big part of your non-9-to-5 life?

Eric: I want to give credit to Michael Kennedy, as well has his podcast on Talk Python. Of course you know - during the time I was learning Python, I was trying to look for resources on Python itself. His podcast was a great influence on me. At the time, I believe he actually recommended two books. If I remember these titles correctly, one is called A Company Of One, and the other was The End Of Jobs.

Basically, the gist of these books was, “You are your own company. You want to treat your career as if you’re a start-up of one. You want to market your skills, and think of yourself as a marketer.”

What’s your demand? Who’s paying you to do what? Would your skillset fit somebody who would pay you to do it? I may love eating pizza, but nobody’s going to pay to watch me eat pizza. You want to find these circles. Then once you find that overlap, you want to spend all of your time working on that part. Because that’s the part where it’s most valuable to people.

After listening to that, I thought to myself about where I want to take my career. At the time, I had a young family. When my first kid was born, or I had my second kid - I can’t recall. But anyway, that got me to think about where I want to go. I still need that job, but also I want to take advantage of the opportunity to generate side income and treating my career as if it’s a startup. I wanted to know, where’s my most valuable skill set, and where’s the market demand for that - and so on, so forth.

That is when I started to - you’re right - that was around 2016, 2017. 2017 was when the book came out. But 2016 is when I started making courses, and trying to find that unfulfilled demand out there.

That goes back to what a finance and business degree taught you, right? You want to find a demand that is unfulfilled, and that is large enough to sustain couple of players in it - and so on. That was the trigger.

That continues on to today, where I enjoy making content and - But as you know, right? Making content is great and all, but it’s hard to scale. Because you’re really just depending on yourself. That’s what I see as the next challenge, is - how do we make user generated content beneficial, and have the influence that it deserves?

Anyway, so that’s some question I have. That’s why Leanpub is such a great platform. I was actually surprised to hear that it’s been around for 10 years. I mean, you must have a great story to tell - and I’m looking forward to like listening to your - I listened to several episodes, but not all of them. But I look forward to listening to your story, how you came up, and all the challenges that you solve.

Len: Thanks very much for sharing all of that, and for the kind words about Leanpub at the end. I mean - obviously enabling content creators like yourself, to do things efficiently in way that reaches an audience that scales - that’s what we’re all about, is trying to help people do that. It’s funny actually, the file number on the recording for this episode is 200. This is the 200th recorded episode.

Eric: Oh nice.

Len: I’ve been thinking about doing a like just solo, “What is Leanpub?” episode, where maybe I would just talk for 45 minutes or an hour about what is Leanpub - how long has it been around? All that stuff. What have been some of the challenges over the years?

Eric: Oh forget my episode, I’m looking forward to that episode, I want to listen to that one.

Len: I’m not going to go off about that too much, but yeah - the fun parts will be like - when you’ve been around that long, the things you stop noticing. You stop noticing that they’re not happening anymore. People going, “Why would I buy a book that’s not finished?” Like, of course - when we launched, we got that all the time. Because Leanpub books weren’t just unfinished in the sense of like serial novels that are published chapter-by-chapter. They were unfinished non-fiction books that are published after someone’s finished three chapters. And by the way, they’re probably going to be revising chapter one every time they go back in. So, even something you read is now going to be changed. That was a strange idea.

But knock on wood, we basically don’t get that anymore. That’s one of the features of being around for a while like that. If you’re doing something that’s - insofar as Leanpub’s innovative, like in a fundamental way when it comes to publishing. I actually -

Eric: Yeah. I mean, that has also to do with - back to that business mindset, right? If you treat your skillset as a - quote unquote, “company of one” - you want to do the proof of concept. You want to have a lean process that’s advertised by - was it Eric Ries, with his book of - Eric Ries’ first -

Len: The first Leanpub book was Startup Lessons Learned, which was turning Eric Ries’ blog into a book.

Eric: Yeah, so for sure, right. Everything that you just said really resonates. I’m glad that you don’t - people start to have that buy-in, right? That you could drive the direction of the finished product, if you were able to just show it to people and put it in front of people - so that they know where the demand meets the skillset.

Len: Oh yeah, definitely. To talk about that for a moment as well - one of the things we always used to say is, “One of the biggest successes we can give an author is proof they should stop writing their book.”

Because in the the olden days, if you had that cartoonish image of how a book is written, right? It’s like - you get a contract with the publisher. Then you go live in a cabin in Norway for three years, working on your magnum opus.

Then you give it to the publisher. Two years later it makes it through their process, and it comes out. Then no one wants it.

Eric: That’s a tragedy.

Len: Well, if you were on a mission, or if it was valuable in itself to you - to go through that, and write that and have it out there - and maybe someone will discover it in 100 years, or maybe they never will? That’s time very well spent.

But if your goal was to get an audience for your work, for yourself, or for your cause that you’re advocating - to find out five years later what you could’ve found out if you just published two chapters in, that no one’s interested in a book on the subject, then you just saved that person 4.8 years.

Eric: Yeah. He could work on his next masterpiece, right?

Len: Yeah, yeah.

Eric: I think - absolutely, right? If your goal was just to write something that pleases yourself, then that’s - something down the line, of course. But I think the majority of our work - we want to see that reach people, and we want to be able to be the change that we want to be. And oftentimes for authors - that’s through their writing. If nobody reads it, then what are the chances of you accomplishing your goal?

Len: Yeah. It’s interesting - the old example I use, a funny example I used to give, to show how complicated it is - deciding what to write about - if you like writing and you like teaching, specifically with respect to prescriptive non-fiction, which is most Leanpub books. Which is, teaching you how to do stuff basically, or explaining things.

If you really like a subject, and you really want to write about it and share it with people, you still have to make a choice about what to do first, right?

Eric: Right.

Len: For example, you might write, The Expert’s Guide To Model Train Sets. You publish it, and no one’s interested. But if you’d started writing, The Beginner’s Guide To Model Train Sets, The Intermediate Guide To Model Train Sets, and The Expert’s Guide To Model Train Sets, and you started writing all three at the same time - and published each one when it’s three chapters in. You notice - no one was interested in The Expert’s Guide, but there was a ton of people interested in The Beginner’s Guide. Well, now you know which project to work on.

When it’s done, guess what next project you’re going to work on? All these people who’ve finished The Beginner’s Guide. They’re going to want The Intermediate Guide next, right? Publishing as you go with projects has these deep benefits - including abandoning projects.

That’s some of the stuff that we’ve learned over the years that was a bit unexpected. But–

Eric: Not ashamed to tell you that the first book I started with Leanpub was many years ago. I think I sold - like - what you said, I actually was speaking at DevNet Create for one of the Cisco events. Actually - before the event - it was on elastic stack. Before the event, I advertised the book, and so on. I sold a total of one copy. I know that’s not - probably I either shift my focus into a more useful way for people. Or I just abandon that.

So the Kafka book really wasn’t my first encounter with Leanpub. But that was still very useful information. Even just to pick the right project to work on. I do put my money where my mouth is, right? As soon as I made enough, I went back and got a lifelong membership with Leanpub. I am - for better or for worse, man - I’m counting on the platform to be around for 10 and 20 years.

Len: Well, thanks very much for that. Actually, just so people know, we do have a freemium model, so there’s a free plan.

You can do lots and lots of things - awesome things, as a self-published author. But there are some things that require maintenance, and things on our side. We do have paid Standard and Pro plans.

What Eric’s talking about, is - it’s a form of thing that we actually are very glad to be able to provide to people. Which is - one of the other values, or one of the other things of value that can come from publishing in progress, is that your book project can become self-funding.

If your book starts making some money, then you can use that to pay for like a week off to work on the book. Take a little leave from work. Or you can use it to pay for advertising for your book, if you want to do that. If you can set up a good self-funded advertising model for your book, well, now you’ve just got something that takes off on its own. Tied in a bow, as it were. Off you go.

You can also - of course - if you’re on one of our monthly plans, we have lifetime plans. You just don’t have to worry about the monthly payments anymore. A lot of authors who have our lifetime plans, it’s exactly that. They’re like, “Huh, this one’s working out. I guess I’ll use my proceeds to get a lifetime plan, and now I don’t have to worry about that anymore. I can just work on as many Leanpub projects as I want using all the features,” and away you go.

Eric: Yeah, something that - I think it’s both true in self-publishing, as well as open source project, is the issue of sustainability, right? What you were talking about. A lot of the technologies we work on are open sourced. Or started in open source, and the result of a lot of volunteers volunteering their time, effort and energy into this project.

But now the issue becomes the sustainability. Who’s paying for this, if there are no costs to do this? Same thing for content creators. Is - how do you sustain your operations?

Back to wherewe were talking about Leanpub - a traditional publisher, since I’ve worked with - I’ve also worked with O’Reilly and all these other publishers that produce on smaller scale. But their aim is a little different, right? Their overhead structure is different.

Their model is to have like 1% or 5% of the books that profit - that’s profitable enough, to make up for the rest of the books that do not make money. Which results in a smaller percentage of royalties for the authors that do end up in that 1%. Whatever person comes for that model, what ends up being is - it’s hard to sustain a writing career if you’re only getting 16%, 20% or up to 25% of the royalty that the book makes.

However, if you’re able to use a lower-overhead platform, and still get even just a fraction of the sales figures - first of all, you’re able to write on topics that may not be geared toward the general public. Very much service toward that long tail, but at the same time, you’re able to sustain your project.

I think those are two big issues that I see right now in open source community, as well as content creators. Is just the sustainability. You’ll see hyper success for companies like Anaconda, which is a Python package company for data science.

They’re uber successful, but they still have to go out and do series A, series B, series C funding, just to sustain their operation, and still contribute back to the open source community. If they have to do that, think about the 10,000 other companies that are doing similar stuff - but don’t have the recognition or visibility of Anaconda. It’s not a problem that we could solve right away. But what Leanpub is doing, is being part of the solution.

Len: That’s a very good description of how a lot of - particularly how the publishing industry works, right? Because if you’re a conventional publishing company, I mean - somewhat infamously, there’s been - typical publishers tend to be spending less and less on advances and things like that. Less and less on marketing, and stuff like that.

It’s a big problem for them, to attract authors, because if you’re not giving them a high royalty rate, if you’re giving them a low royalty rate, then one of the things that happens, specifically for what you’re saying - only a small proportion of the books that a publishing company publishes are profitable, abnd can make up for all the investments in all the other books that they’ve had to invest in. What happens is that, if your first book ends up in the remainder section of the bookstore -

Eric: Yeah, right - not doing too hot.

Len: No, you’re not going to have a very good chance of getting a second contract. I mean, I’m speaking at a very abstract level here. Keep going, don’t give up - you know what I mean?

Eric: Sure, sure.

Len: But still, and notoriously too, you can be really excited, “My first book is out.” But if the guy sitting next to you, his book is doing better - just in that first week - guess who’s going to get the next week’s marketing budget?

Now your book dies. But maybe if there had just been a different news article out that week on your subject, yours would’ve been the one that made it beat the other one in the first week - and you would’ve got the company’s marketing budget going forward.

I mean, all things in life involve risk, and there’s precariousness, and luck and bad luck - and things like that. But -

Eric: Yeah, you lose a lot of gems, right? I think if you think back on the uber successful comedy Seinfeld - the first three seasons were just - they’re going nowhere, right? I don’t know if you’re familiar with that show, but it’s like - I think they had a very successful run. It’s a US-based show, so you may not know. But the first three seasons, they got nowhere. All of a sudden, the fourth one - starting from the fourth season, if I remember correctly - it just exploded, and it become one of the most successful sitcoms ever.

But the thing is, if you give up on the first three seasons - well, you never get discovered. The fourth season was never made, you generally lose out on this excellent show - and you lose out on the stuff that would benefit a large chunk of people.

What you were saying is that the chance of being discovered is shorter and shorter. People’s attention spans are shorter and shorter. There’s a finite amount of money, in order to get your product be noticed. It’s hard to stay in the game if you’re not successful right away. I think that’s what you’re driving at too.

Len: Yeah, if your success is contingent on the efforts of your publishing company promoting your book, then you’re exposed in these ways.

You brought up the other thing too, which is that, with self-publishing, you typically make much higher royalty rates, which actually makes some projects sustainable - just in themselves, that wouldn’t be otherwise. For example, I bet Nick Russo would not have found a conventional book publisher to sell his one million word, $400 book, right?

But from the author’s perspective, to make it worth his time and a sustainable thing - he didn’t need to reach a million people.

Eric: Precisely.

Len: I wish he had.

Eric: Right.

Len: But he didn’t need to. There are lots of– I mean I say it in a joking way - there are books that exist because of self-publishing, that wouldn’t have existed otherwise. All 300 experts in that area in the world, now have an excellent book. The market is 100% saturated. 300 copies were sold. But it’s an excellent book that wouldn’t have existed otherwise, and authors now can pay for their kids’ college or something like that.

These are the great successes that come from having these different kinds of models.

Just speaking of things that you experience over the years - particularly at Leanpub, is - there’s still a sort of - we’re still using the word “still”, one thing that hasn’t changed about a certain negativity that some people will have when it comes to self-publishing books.

But when you’ve discovered the fact that you can actually pay for your own time to write that book you always wanted to write, and have everyone in the world who cares about that issue learn about it and thank you for the book, your attitude towards it starts to change, and you stop worrying so much about like, “Oh, what does this say about my social status, that I self-published a book?”

You know what I mean? You start asking more serious - it’s funny, when it becomes truly possible to do it well, self-publishing I mean, you start thinking about what you can accomplish, rather than what people are going to think - if you know what I mean?

Eric: I agree 100%. It’s really about freedom. Working with a traditional publisher - like you said, they really worry about - the first goal is to sell books. They want to capture as big of a base as possible. You’re talking about like SEO ranking. You’re talking about - most likely, beginner books, right? There’s only so much experts or intermediate who are interested in a topic. But the base is like a pyramid. The base - the introductory books, usually capture the most potential readers.

If you realize that part - it’s sometimes not in sync with the author’s objective. From the author’s objective, or perspective - they really want to write the book they want to read. They want to write the book that has the influence that they’re looking for. When there’s delta in the traditional sense, the publisher wins, right? That you’re bending your original goal into including this chapter, that might not have otherwise made it into it. Or you made every chapter 6,000 words, instead of 3,000 - because, whatever.

I think it’s really about the freedom that it provides people. For authors, like you said, you don’t need to worry about all this other stuff. You just focus on what you want to write. If the reader’s very happy, and you’re good with it - then so be it.

Prior to Leanpub, I don’t think there’s anything that was available. Of course - CreateSpace, and all that - but that’s another story.

Len: Actually that’s actually a really interesting thing to bring up in the context as well. Because particularly in the print book world, actually, the trend is - in a very good way - not exactly the opposite from what I was describing. But basically because print-on-demand is how books are produced now, what they used to call the mid-list or back-list books, the books that the company’s not promoting anymore, because they’re 10 years old - a book could be 20 years old, it may not have sold a copy in 20 years. But then, for whatever reason, the subject that the book is about just starts trending. Everybody just goes onto Amazon, clicks a button. Then a bunch of automated messages get sent back and forth. There’s a machine somewhere printing that book off now.

Eric: Exactly.

Len: The reason that that’s revolutionary, is because, in the past, you could only make money by printing a big -

Eric: Inventory of books, yeah.

Len: A big inventory of books, right? Then if someone ordered them, you’d be like, “Thank God I can get one of those out the door.”

But now that they’re being printed when they’re being ordered, you can do that profitably. It means any book that has been written and is existing in a digital file somewhere, can have it’s printing trigged by you on your phone.

That’s actually made a lot of books profitable in a way now that they’ve like - or at least, they’re generating some royalties for their authors in a way that they actually wouldn’t have in the past.

Eric: I agree. I was actually with the fulfillment center branch within Amazon, when they acquired CreateSpace, as well as the other company, I forgot the name, but for CreateSpace, for books, right? The other company for digital media. Movies and pressing dvds. They were very much on-demand as well.

You think of movies and DVDs as if - when they release it, they make a bunch of discs - maybe that’s true for the first run. But afterwards - Braveheart, for example, right? 1995.

If somebody wants to buy a DVD of Braveheart, it’s not like they dust off from the shelf in some old warehouse and ship it to you. It’s a digital file that’s stored somewhere, and now they’ll say, “Oh well, great - Braveheart,” we’ll just make a disc right there and then, and ship it to Eric.

That allows - like you said - a lot of the long tail people who might not be served otherwise, would be able to just make it a possibility. You’re right - print-on-demand’s really revolutionary, and it’s great.

Len: I was slyly hoping there to find a way into talking about your book, Kafka Up and Running for Network DevOps: Set Your Network Data in Motion

Eric: Sure, no worries.

Len: I was trying to do it by setting the idea that like messages are being sent back and forth, between what you might call “producers” and “consumers.” But these are all machines.

Len: so, I was wondering if you could just take a few moments to talk about the book, and what Kafka is?

Eric: Sure. Kafka is really just a messaging system. If you think about the use cases of Kafka, we have a lot of databases and storage options, like your large scale - whether it’s SQL no SQL database. We have a lot of options, where you need to write a record persisted somewhere, and then retrieve it later - right? You write it, like you’re signing up a library card or you do your driver’s license. You fill out your information, it’s stored in a database somewhere - and when you need it, you retrieve it.

But what we are lacking is this system where we’re treating data, as in, something about data in motion. That data comes in, and it’s only stored temporarily in the storage capacity, until we could enrich that data, and we influence that data and transport it.

For example, in a network security sense, that’s when I first started to think about treating data as in, data is this moving, always morphing subject, right?

We would have, say, a log somewhere, and then we will have an IP address.

In the log itself, an IP address is not very meaningful. What we need to do, is - we put the IP address into a place where we could create a IP database table, or database, that we know this IP address is from Russia, and it’s from the company of - I don’t know? Russia Telecom, for example. Then we feed it into the Kafka topic. The Kafka subscriber receives that information, and it will say, “Okay, now we’re correlating that IP address with other information that we may have.”

Maybe we have a list of botnets. We have a list of IPs. We can match it up. Whether that’s the hit or miss. Now that’s a second topic for publishers and subscribers. Then we can publish it to a third place, where we could store it into a place where we could visualize. We could have it aggregated into a map, and so on and so forth. If you do that enough, then you have a more aggregated view.

But the point is - if you need to write that data - you write that log entry into a database, retrieve it, enrich it, and then write it to a database again - and then retrieve it. That pattern is very different than my use case of always treating the data as in the stream of flowing data. That’s essentially the problem Kafka’s trying to solve - is trying to treat data, as in, data in motion. You’re always trying to get data, retrieve data - and then it becomes a permanent storage at the very end.

Len: Just to give people - I’m going to try and come up with a concrete example of this.

Eric: Yeah, of course - sure, please do.

Len: Based on some of the research I did from your youtube interviews and stuff that I’ve seen. But let’s see if I can collapse it down to a very specific example, which is, when you pause a video on Netflix. Netflix uses Kafka, I think? That pausing is a “topic” in Kafka parlance, right?

Eric: Right.

Len: It’s not like storing the whole movie paused in that state somewhere, it’s writing a quote unquote, “topic,” that’s like basically a pointer to where in that movie you want to go, if the person clicks “play” again.

Eric: Right. If you’re like me, I’m always sampling different movies. Within the span of 30 minutes, I might be looking at movie A - pausing it - and then looking at movie B, pausing it. Then, “Oh no, let me go back to movie A now.” I would watch it for 5 more minutes. Then I’ll stop. Then I’ll go back to my - maybe I’m doing it on my TV, and then I’ll go back to my treadmill - I want to continue on on my iPad. Then I have to do that again.

Netflix is using Kafka to store this temporary information - my first pause from movie A, that is only meaningful until I resume it on movie B. Then that information’s no longer useful for anybody, right? I don’t care that on Sunday night, 7 pm I paused it at 5 minutes and 42 seconds. All I care is, my latest progress for movie A is - an hour and a half into that movie. I want that to be transparent to all of the devices, right? Whether it’s on my phone, my iPad, and so on.

Those are the problems that Kafka will solve. Just that technology in general. It’s not for everything, and it’s not for -

Back to what we were talking about. It’s so interesting, Len - right? Everything goes back into what we’re talking about. I knew this book would not be as popular as Mastering Python Networking. Or I knew this topic wouldn’t be useful for everybody. Especially in a network engineering community.

Most people are still learning about automation, they don’t really worry about data. They don’t really worry about - just like I didn’t worry about it in the beginning. But I wanted to write this book, right? I don’t care if two people read this book. I wanted to have this book out there, so that it could serve a purpose. It could serve people who does need it.

I forget who said this, but it’s like, “Do you want to write a book that’s like a vitamin, or do you want to write a book that’s like a painkiller?” Right? When you’re in pain, you’re like, “Oh my God, I want that painkiller, this is the exact book that I wanted, it will solve my problems tomorrow, I’ll be a superstar at work.” Versus, it’s like, “Oh it’s so nice to have.” It’s like, “Oh, I want to learn about automation at some point, and I’ll just buy this book. I eat this vitamin, it’s good for me.”

I look at Kafka as being - okay, I was the one looking for this book in a network engineering context. I want this book to be that one book, that when somebody else needs it, they could be able to download it. And, yeah - it’s only a fraction of the sales figures, but I’m happy with it, right? I’m glad I did it.

Len: Thanks very much for that image. I’ve never heard that before, but that perfectly captures so many like - My brother and I have this joke that there’s an infinite number of ways you can split the world up into two kinds of things. You’ve given me a new one - which is vitamins and painkillers.

That captures it very well too for me. From the author’s perspective as well, right? Because sometimes people write books for the same reason, right? Like, “It’ll be good for me.” Or, “It’ll get rid of this terrible pain that I have.”

Or it could be, I’ll just put these thoughts down and force myself to do it in an orderly, clear way. Because then I’ll actually be solving my own problem, by forcing myself to do that. It’s one version of the old line, “You don’t really understand something until you have to teach it.”

Eric: Precisely.

Len: By setting yourself that challenge, like then you’ll really understand it.

Len: The last question I always like to ask guests on the podcast, if they’re Leanpub authors, is - if there was one terribly awful, broken thing about Leanpub that you could ask us to fix, or if there was one magical feature you could ask us to build for you, can you think of anything you would ask us to do?

Eric: I can’t think of any like glaring thing. I mean, the fact that I bought a lifetime membership, right? It means I believe in the system, I believe in it enough that I want to work with this platform long term.

But what I would think that - it would really help, is to - right now, in order for me to do a more thorough distribution channel, I have to go to somebody else’s. For my use case, I have to go to PublishDrive - which is another startup that pushes it to other marketplace, such as - what is it? IBooks for the Apple store. They also do Amazon, and they also work with Ingram. You could sell it to foreign companies and so on. Especially if you want like a print version of your book as well. If I could ask for a component - I know it’s a heavy lift, and it’s probably against the idea of “lean.” So, but that’s what I would say.

Len: You mean like a make a print book and distribute it to channels?

Eric: Yeah, make it easier to distribute it, and maybe market it - buy ads on some other platforms. That’s another thing that these other vendor provides, which is - what I do is, I use Leanpub to write the book, and make it in a print-ready state. Then I could upload it to somebody else, in order for them to make a print copy - as well as buy ads if I wanted to. I didn’t - but if I wanted to promote my book more, I’m able to buy ads on, such as, the Amazon platform, maybe Google search - and so on and so forth.

Len: That’s really fascinating. This is something we’ve been asked for , generally speaking, in the past. I mean, very specifically, as Eric’s mentioning, we have a print-ready PDF export feature.

If you’ve written your book in one of our workflowsm, you can just click a button, and we give you the PDF. I mean, you set some parameters and stuff, but then you get the PDF file that you need to upload to like Amazon KDP or Lulu or all these other services, that can then be print-on-demand services, like we were just talking about.

But more generally speaking, what people are often asking - at a more basic level, what they’re asking for is like, “Look, I don’t want to have to go to five - the book publishing part of my life is one room in the mansion.” Or, “I wish it could just be one room in the mansion, and I’ve got the room in the mansion, which is Leanpub. But now I’ve got to go to three garages, to do the other things I need to do,” right?

So there is a desire out there for that. I’ve got a very specific question for you on that actually, if you don’t mind?

Eric: Sure.

Len: If you could click a button on Leanpub and we then published it on Amazon ourselves - which would then have to be under like a Leanpub account or imprint or something like that - and obviously we’d have to get some cut of whatever the royalties that came in from that - would that be something you’d want to do, even though you could just go and set yourself up on Amazon independently and get 100% of the Amazon returnsfrom the Amazon sales? I’m sure you know what I’m asking - would it be worth saving your time and the efficiency to just have all of that handled by Leanpub? Even though you might’ve made some more money if you’d done Amazon independently.?

Eric: Yeah, I don’t think - I know other authors that i’ve talked to, that would be useful for them. They would totally go for that. But for me, I wouldn’t do it, and I’ll tell you why.

It’s not so much about the money. It’s more about the freedom and the control that I would get, right? This book, since it was the first end-to-end finished book for Leanpub - I actually did a lot of experiment, right? For people who don’t know, Leanpub also - once you publish the book, you can also pause selling of the book, for any reason that requires that.

In this case, because I wanted to experiment with Kindle Unlimited, and all of that - and they require exclusivity, I could pause Leanpub, and I could pause the distribution on PublishDrive. Then Kindle becomes the exclusive provider. I just wanted to see how they come up with like the royalty percentage, and all of that.

The added benefit of Kafka - knowing that it might not be as popular of a topic, is - I get a lot of freedom to try out these different things. I did try all this other stuff, right? Give Amazon exclusivity. Edit the page within Amazon. Get different ISBN numbers from different sources, and including from Amazon itself. Read through all the agreements and all of that.

I think for me - I wouldn’t do that, because of the control. What I think would - this is just from experience, right? Most of the books, I think, selling in the US - probably 50% comes from Amazon. What I would want - in this very specific case, is - I want to have that A+ page. If you spend a little bit more time, and if you have a consulting gig, right? If a customer generates 50% of your revenue, you want to be able to treat that customer as a VIP, and give it as much love, tender care as possible. As opposed to all the other ones that you maybe just want to hit the mark as being pretty good.

But you want this VIP customer to feel excellent, and be at home - in this case, because Amazon generates about 50% of the book sales or more, than I want to be able to take that print-ready copy - upload it to Amazon, and generate what they call the, “A+ page.” Which means there’s more summaries, and you can upload images and give more details on wthe– What the book was about. That’s very purely personal taste, I would say.

But I would also think that - just from chatting with other authors in the networking industry, who also do publishing on Leanpub - is that they would totally go for that. They just want a few print copies that they can show their parents or that they could give out at conferences, on speaking engagements. So, I could see both ways. It’s just that, for me, personally, because I take it more seriously, I wouldn’t go down that route.

Len: Thanks very much for going into so much detail, I really appreciate that. It’s one of the reasons we leave these kinds of questions till the end. By the way - there actually are people who skip to the end. They’re like, “I don’t care about all that other stuff.”

Eric: Oh is that right?

Len: Because what you just described there - for someone who’s starting out in self-publishing or something like that, that’s just a golden three minutes of insider knowledge that would take so much time to learn about on your own - to hear about that experience directly, and then the various needs and the choices - like when you’re self-publishing, you’re making choices all the time.

Unless you’re not, right? Because because you don’t want to be making choices all the time.

It really does come down to you and your personality, and maybe the individual project, and stuff like that. For a lot of people, if they’re going to be selling on Amazon, they’re going to want control over how their books are portrayed, and how they’re portrayed, in fine detail. There might be some people who want to tweak it every day, to see, “Does this make a difference, does that make a difference?”

Eric: Yeah. A/B testing.

Len: Yeah.

Eric: That would be a killer feature. You mentioned people might find it useful, right? Feel free to reach out to me on Twitter. I’ve learned so much, just from this one book of self-publishing experience, right? Who would’ve known that there’s only one company who sell ISBN numbers in the whole of United States, right?

Or, for example, how thick of a page - it’s usually in the print format - how many pages does it take, in order for you to have like words on the spine, right? And less than that page, because of the depth - it doesn’t even make sense to have words on the spine, and so on.

If anybody’s interested, I’ve learned so much, and I’m happy to share that knowledge. Just reach out to me on LinkedIn, on Twitter - whatever. I’m happy to just bounce off ideas. Because we’re all connected, and we should be helping each other out.

Len: And if you do ever write a blog post about it or something, please don’t hesitate to get in touch with me, and I’ll make sure to - even the Leanpub accounts, we’ll tweet about it and stuff like that. Eric: Oh awesome.

Len: Not only would it be obviously inherently valuable in itself, but it really helps people who are starting out, or who are frustrated or who don’t know what - they see these two paths before them, and don’t know which one to take. Hearing about those kinds of personal experiences can really help.

Eric: Yeah, I mean - not to derail, but I did benefit from a lot of these tribal knowledge as well. I was approached by Packt, but I didn’t know anything about it, right? Is this percentage they give good or bad? How many copies do I anticipate to get?

There’s surprisingly not as many people to write about the publishing experience, as I would’ve imagined. But I did go out there. It’ll be like - I don’t know? The second or third page of Google search results. That’s how deep I go, and look for those blog posts.

For those few that actually published the numbers, it helped me tremendously to set the expectations right about - you’re not going to get rich, but if - what’s considered good, right? Is 90,000 good? On Amazon ranking that is, right? Is a million good, or -? It really benefited me, and I want to contribute back and be part of the solution, in providing people who are confused maybe.

Len: Everybody, Eric means it, he’s a very friendly guy online. I mean, we’ve never met, but doing my research, I got a sense of your personality - how open you are, and how much you like helping people - and stuff like that.

Just to close that off on that note - when it comes to like you have to go three pages down - when it comes to publishing contracts, one of the really - that whole world, there’s this woman named Kristine Kathryn Rusch.

Len: She’s been a guest on the podcast in the past. Her blog is just excellent about that. You’ll find her pretty quickly.

Eric: Oh nice.

Len: If you ever get approached by a publishing company, the specific amounts, things like that are changing all the time.

But the principles of like, “How should I approach reading a publishing company contract?” - stuff like that, her blog is really excellent about that.

Eric: Awesome, i’ll check it out.

Len: Well, Eric - thank you very much for taking the time out of a beautiful Seattle area evening, to talk to all of us here. Thanks very much for using Leanpub as the platform for your latest book.

Eric: Oh it’s a pleasure, Len. Yhank you for creating the platform, I’m a believer. Hopefully I get to work with Leanpub a lot more, and I’m sure I will - and I’ve enjoyed the whole experience. Thanks for having me, it’s been a pleasure.

Len: Thanks very much.

As always, thanks to all of you for listening to this episode of the Frontmatter podcast. If you like what you heard, please rate and review it wherever you found it, and if you’d like to be a Leanpub author, please visit our website at leanpub.com.