Episode 19: Automate All the Things w/ John Capobianco
In this episode, we sit down with the master of network automation, John Capobianco! We talked about how John got his start in the IT industry and entry-level introduction to automation with stuff like Python and Ansible. We also talked about his new mind-mapping project where he turns public rest APIs into mind-maps. We also touched on his network automation book which is available everywhere!
Be sure to follow John on Twitter here
Vote for us in the Cisco Champion IT Blog Awards here
Check out the survey here
Follow the show on Twitter, Instagram, Facebook, Discord, and Youtube
Support the show here
Email us here
Like us? Give us a review on Podchaser or Apple Podcasts to let us know!
-
Pat: 0:19
Hi, everybody. Welcome back to this week's edition of, so you want to be in it if you've been here before, you know what we're about, but for all you new folks out there, we are a podcast aimed at those starting out their journey in the vast world of it. We talk about breaking into it from a beginner standpoint, right? So either they're right out of college or changing careers, or if you maybe just got into it and looking for the next rung of the ladder, we're here to help with all of that. So again, everything from climbing the ladder, management breaking in and everything in between. So I'm your host, pat, as usual hanging out here, you can find me on Twitter at pat Allen, 180 2, and then you can find the show on Twitter. @sywbiit the acronym for, so you want to be in it and we're pretty active on Twitter. So come say, hello, Dean Mack, hanging out over here. Dean's on Twitter at DeanMacUK. What's up being, how you doing?
Dean: 1:14
guys, hey Pat? How's it going man? Long time. No
Pat: 1:17
I know. No, it's been a a couple days. So we're trying to keep up with the weekly thing here. So, some schedules got knocked out of whack. So we're doubling up this week for as far as recording. So you guys will hear, stay on track with us and, trying to bring you some content and stay relevant. And consistency is the key. But, I was waiting for this podcast. This was the one I was super excited about because we have the man, the myth, the legend, Mr. John, Capobianco on with us. I hope I said your name, right, John?
John: 1:47
You nailed it. You nailed it. That was perfect.
Pat: 1:49
I.
John: 1:50
pat. That's
Pat: 1:51
Nice. Good deal. I studied in school. Yeah. we're all good. Oh yeah. John's here with us, John. man, it's been a long time coming. I know you reached out to us on Twitter a while ago, to come on the show and, you know, we've gotten more acquainted on Twitter and some of your work and things of that nature. And, we couldn't be more excited for you to be here and hang out and chat with us. So, thanks, man. How you doing?
John: 2:12
I'm doing? great. I'm equally excited. And I did a reach out a while ago now to offer my support. And, I, this is really close to my heart. We were talking before we got started. I've been in it a long time now, but I am so passionate about the impact it can have on someone's life, on society, on communities. it moving for me personally, getting into it changed absolutely everything about the quality of my life and my day to day experience where I got to go, do you know my career and. So thank you for having me. I'm really excited about this discussion.
Pat: 2:52
It's our pleasure.
Dean: 2:54
Absolutely. Yeah, she was at your're here and I really liked the way you framed that. I, um, I'm a true believer in that too. It is definitely enhanced then, uh, redefine d my lifestyle completely and accomplish a lot of my personal goal. So I think it's a huge that you brought that to the table and to the forefront, to our listeners.
John: 3:14
Well, I know for me personally, I've been fortunate and I don't want to, you know, I have to mention my privilege that I got a home computer when I was very young and when they were very new. So when they were still A hobby in your garage, you put together what they call them. IBM clone. And you built your own kind of Frankenstein computer. And that was in, for me, it was like 19 86, 88, somewhere around there. So a very long time ago. And, yeah, so I've been lucky that I've always had computing as part of my upbringing, but I found myself, you know, I went to university out of high school and didn't last, very long. I was in arts major and it just didn't click with me. And, I found myself on academic probation. So I took a job at the aluminum factory to help, you know, get a job. My dad worked there as a millwright and was able to get me into the production side of things, working on. heavy gauge, heavy metal, aluminum machines and forklift operations and pack lines, and they call them tension levelers. And I, after a few years of working 12 hour shifts doing that days and midnights rotating 12 hour shifts, I, I really wanted to change my life and really wanted to do something different. So I went back to school in the community college, in the computer programmer analyst course. So I know, I know I'm a network engineer now, but when I went back to school, I went to a three-year computer programing course that, you know, we wrote everything from C to JavaJava Cobal, kicks, uh, ual basic, we wrote some real applications and there was a component of networking to that community college program. So that I graduated in 2003 to give you a rough timeframe. Yeah.
Pat: 5:15
Nice.
Dean: 5:16
So you have been in industry for yeah considerable amount of time.
John: 5:19
So the one advantage of the community college program was that we got a job placement in our final year. So my placement was with the Ontario ministry of health. And yes, I would say since about 2002, I've been in, in it in one shape or form or another. what's interesting. And at the time it, it was difficult, but my placement was more in what I would Consider it. networking, client server, storage, connectivity. Building systems installing software, that type of stuff. But I went to school to learn how to write programs. So when I graduated, I sorta had a lack of computer programming experience in the field, but I had, a diploma in programming, but I had this sort of a plus N plus level of field experience with the ministry of health, without any certifications or diploma, like my diploma was in programming. So that was challenging because then I I wasn't confident enough in my own abilities as a programmer, even a junior programmer, I did do one QA contract. So that was my first contract. Once I graduated college, I got a three month contract in quality assurance in Toronto. And then at the end of that contract, I went back into the it side of things for a value added reseller in downtown Toronto for a six month contract. So it wasn't full time, but the way I looked at it. I doubled the length of my next contract. So things got little better right. With my second contract, but that was building windows servers primarily and doing break fix work in, in the bay street area of Toronto. So the financial sector, for example, and that may be a lot to unpack. Do you have any kind of questions about where I'm at in my journey so far?
Dean: 7:24
No, that's pretty clear. What made you gravitate to sort of the computer in industry? you said, you had a computer at a very young age, was that choice? A parent's choice? uh, someone reallreally who influence family family of bell war. How did that come into fluition?
John: 7:42
It's my, yeah, my parents at the time, you know, everyone was hearing about computers with were the future and, you know, kids should get involved in computers, I guess not much has changed. People still talk about that. But you know, in the mid eighties it was like, it was really revolutionary stuff. This was before windows or anything. Right. So multimedia wasn't here, the windows 95 error wasn't here, but my younger brother and I kind of together, you know, what really made it click Dean was, building a bulletin board system. And that's what got the hook into me into technology was letting other people dial into my computer through their modem and to figure out how to set up a BBS, you know, with that technology. And at that age, I was still, I was about 14, maybe 12. That really was neat that, okay. the computers standalone computers are pretty neat. We could play video games on them and make spreadsheets and do things. But once we connected them, once the three of us, like right now are able to, from different parts in the world, have a conversation with video and audio through the computer. that was different. That, that really something clicked. And I wanted to, I thought I would enjoy working with that, telecommunications side of things, I guess.
Dean: 9:11
okay. So it was like the broader aspect of multiple computers, like the internet, like the infrastructure of the internet that appealed to your sort of curious nature.
John: 9:21
Yeah. and just connecting with people and all around the world. I don't know. I think it all goes together, but one of the early computer programs was PC globe, which lets you pick on a country and you could hear their national Anthem and look at their flag and how many people lived in the country. now we have the ability to connect. Like I've had similar conversations like this with people in Nigeria, in Brazil. And you probably have a similar experience now where you're sharing and connecting with a global audience. It's just, I don't know. It's fascinating to me that we can come together to solve problems and to get to know each other. And I don't know, be just be more connected as human beings. Right. And that it's the technology that, that allows for that to happen. So, but as for why. I guess specifically why I gravitated. So I think that's why I moved back into the it side of things. building networks, for example, versus computer programming and writing software. But if you flash forward, you know, 15 or 20 years, that software lifestyle and writing programming, it's now available to us for things like networks or clouds or compute, you know, you used to have to build networks, network by network CLI by CLI. And now we can use things like Python or Ansible or whatever, and write programs to, automate that the deployment or the configuration or the monitoring or whatever. So it's almost come full source, full circle. that programming diploma that I felt like I never really put to use. Now, it's like a vital part of my day to day job because the network has evolved to become this programmatic thing.
Dean: 11:16
absolutely. And you're more efficient as an engineer when you're, when you start to automate repetitive.
Pat: 11:23
Yeah, I would agree with that. And I'll be the first one to admit that I am not the best at Python. I'm learning more of it and learning to build it from the ground up sort of thing. And I can take a Python script if you put one in front of me and tell you what's going on line by line, but if you put a blank page in front of me and say, okay, build this to change, switch ports on a VLAN and log in and do all this stuff. Eh, you know, it's one of those kind of things. So you have to with the programming language and John, correct me if I'm wrong with any sort of programming language. Well, especially I'm speaking of Python and Ansible, since those are the two that I have the most experience with in the network world, you have to start at the bottom and build block by block or brick by brick. Does that, is that a true statement in that aspect of, you know, then finally having a fully running automated script.
John: 12:13
Yeah, I'm a big fan and you're Right. It's kind of modular task by task. So Ansible, I think that's why Ansible is so successful because the playbook is made up of nice modular tasks it's written in the YAML. So it's human readable and you can kind of go block by block and say, okay, I see they're adding a VLAN here. They're adding a static route here they're writing memory here. They're making they're gathering information and making a CSV file here. It's nice and easy to read that YAML it might not be easy to write and maybe, you know, I'm being broad in general. When I say that YAML is easy to read, I think it is. I think that's the appeal to Ansible, but I find something like Python. You can almost line up that Ansible playbook task by task. And put it side by side and say, okay, how would I do that? Same command or same step in a Python way, right. Using a library or writing my own for-loop or my own API call. whether or not once easier or harder, I've heard mixed things. I, it comes up a lot because I wrote a book on Ansible, but now I seem to favor Python more. And I think that would be a true statement. I favor Python a little more now and why that is. and which one I find easier. I know, I think that either, either is better than nothing, right. Either is better than going device by device. And I'm not trying to insult anyone who's still there. And I know this is about beginners and you want to get into it. Just try to keep it in your mind. But automation really is achieving those same steps that a human would do. So you have to still understand those initial steps. What is a VLAN? What is the command to create one? How, what is a static route? What is a next hop? What is a sub-net mask? so there is a, still a barrier to entry that you have to have some basic, I would say CCNA and N plus level of network understanding if you're going to be successful in automating it. that's why I like the DevNet associate kind of track because there's still 20% foundational network in that course or that exam. But the rest of it explores more modern approaches to. working with that 20% of the network foundation. I think CCNA is obviously still one of the better entry. I, it's hard to even call it entry level. That's a very difficult exam. It's a lot
Pat: 15:08
agree.
John: 15:09
from zero to CCNA or maybe from zero to network plus to CCNA that I maintain it the most difficult Cisco exam I've ever written, possibly the most difficult overall exam I've ever written. And I've written a significant number of Microsoft exams co, Comptiaams, Cisco exams, obviously. but the doors that it opened for me, you know, talking about, so you want to be an it that CCNA made me very attractive as a candidate. it certainly elevated my status from at the time. A senior desktop engineer moved over and then a server, you know, moderate server level of responsibility. Well, they moved me Right. to the networking team. almost a senior level after I got my CCNA. So there was a lot of I've written them. You know, I started as a Microsoft engineer, and then moved to the network because of a reorg. And as part of that move, they sent me on a bootcamp. And that changed absolutely everything for me. and I still am very passionate about building networks. And, and then after my NA was, I got my design associate and it was just that, you know, people want to know about it. If you go to automate your network.ca/. That's my about page. I have my full career history there and every cert I've written, I didn't pass them all on my first attempt either. I'm not a superhuman, I'm not a living computer, anything I've been in the industry for 20 years. So if I maintain a one year, I tried to write an exam a year was my once I realized that, my A+ was valuable and I would overall, I could say that I enjoyed the experience of buying the big exam books, studying, writing the exam, passing the exam, getting a certification. That whole thing of really appealed to me early in my career. and I was getting more opportunities with these certifications. So my A+, and then my end plus, and then some Microsoft desktop exams. It was Vista at the time XP at the time. Right. Like, and I think the industry is very much the same that if you're, I think we used a trucking truckers, if you're a trucker who wants to move into it, I was a steelworker and I wanted to move into it. There are, you know, moderately difficult. It's not like going to four years, a university degree that you've got to, you know, finance your life away and go take a four year degree to become successful in it. That wasn't my experience, the combination of a community college diploma. And some certifications was enough for me. and I don't know that 20 years is a long time in technology. There was no like plural site or CBT nuggets or. educative io or I'm going to miss some platforms here, but you get the idea, public knowledge free online training learn at your own pace. If that was available to me, I don't know that I would have left, you know, worked 12 hour shifts and then gone to college in the daytime. I could done it at my own pace online
Dean: 18:38
well, what Web 1.0barely a thing
John: 18:40
yeah, I know. So what I, what I'm, I guess I'm, I'm trying to encourage people that if you're, if you're dissatisfied in your career or maybe what you're doing right now, or you want to go back to the workforce, if you've already retired, I used to be a professor and about a third of my students were, uh, Retirees like military retirees or other retirees who wanted to work? They, were getting a pension, they still wanted to go back to work and they thought it was a good, second career for them. You could just sit on YouTube and watch videos for a few hours a day and have a decent shot at probably your end plus a plus CCNA. Um, I don't know. What, What, do you find like you guys are really encouraging people to be in it. Is it still similar to that path of get an, A+ N+ given to the help desk and then keep climbing the ladder into the network or into the server? Yeah,
Dean: 19:46
Yeah, it's pretty much not changed at all. Just like, uh, the syntax of iOS has not changed very much. It's told them the way you built static route when you land, it's still the same way. I know you do fancy automation now, but it's still tall. It's still IP route some things don't change. It's just, it doesn't need to change it works, it works well, what I mean?
Pat: 20:10
Yeah. Now I'm curious of that John, cause you, you we of talked about this before we hit the record button, but you know, the ever-changing or the ever, fire burning question of, you know, do you need formal college or university versus getting your a plus net plus sec plus, right. And so for security folks out there, that's your meal ticket in, you know, you know, one versus the other, right. and I've known folks that, right, or that are in both camps, right. Going for formal school. And, and doing the, their information technology or information science, whatever they classify it as, and then go in that way. And then I know folks that just, you know, break in on an A+ or Net+ and gain experience that way, and their minds just click. In that space and they do very well with it. So there is no, I mean, to me, there is no cookie cutter way of saying, okay, yeah, you have to go to four year school and then do this and then do this everyone's path is different, I think. And I think it just depends on how your mind works and, you know, cause some people are great students. Some people are not, that's just the way the world works. So,
John: 21:20
I'm
Dean: 21:21
I'm going cut you off and say something even more controversial. That's just the kind of guy I
Pat: 21:26
write it down. Here we go.
Dean: 21:28
like, in my opinion, I find, and you don't have to agree with me on this, but most of the engineers I've come across are more sharper and more switched on and can gravitate the concepts quicker. I'm not the ones who are formally educated. The ones who don't have the four year degrees, the ones who have grinded it out and work their way up. Those, those guys are generally in my experience. I can't speak for everybody, but in my experience are the sharpest and I've, I've been around for a while. But if I'm comparing formal education to guyswho haven't had more formal education, the e guys who don't have formal education in tech, are way sharper, and you can even see that. But big people in tech, like Bill Gates, Elon Musk, uh, these guys are huge and Steve jobs and they even went back and got their degree. kind of, I'm not saying the degree slows you down or it's not required, or there's no value there, but just looking at it, comparing apples to apples, to people in industry who are at the top of their game. Usually the one who doesn't have the four year cookie-cutter degree is generally sharper.
Pat: 22:41
I would tend to agree with that. I tend to think that, or I tend to see, I should say. Folks that are very sharp minded like Dean said, and generally clicks for them. And they get it are the folks that though they, they tinkered with things right? with a computer on a kitchen table or in a bedroom. And you know, they tore it apart and put it back together and they they cut their teeth on there, you know, on that laptop or on that computer or whatever they're working on. And then they apply that knowledge is easily appliable to them in their head. It makes sense in their head when they hit the real world and they have, you know, real world it problems. I, I tend to agree with Dean as far as you know, those guys that are really high up there are guys that aren't formally educated. And again, not saying it's not worth it. It just depends on the person. It depends on who you are, how you learn things of that nature. And there's certainly value there. it just seems to me that some people just. Quicker than others, or it clicks better that in some personalities versus others
John: 23:45
Yeah, I don't want to be a gatekeeper at all. I I think that there's room for everyone. And I think that the numbers prove that there's more demand than supply for it as process, especially in the security field. I mean, Canada, I read it in the, I don't know if it was the globe and mail, but a big Canadian newspaper had some statistic of Canada has a 30,000 person. It deficit in it security. Now they will take you if you have a four year CS degree, or if you have your security plus. So I think there's a level of investment. You know, obviously a four-year computer science degree is, is an extremely meaningful and, and powerful achievement and accomplishment. I was teaching at a community college, I taught at St. Lawrence for about three years in Kingston, Ontario, there were, there was a few students that were CS graduates that went back to community college to get a little bit more of the hands on, you know, they weren't doing labbing and in their computer science degree, they were writing algorithms and, and it was closer to calculus than it was CLI commands.
Dean: 25:04
It's not really networking or it is, it is data science like, I've done. I've been a part of that. And yeah, you do learn Python, Python is not just for automating networks. It's for like data science, crunching numbers, doing calculations. It's it's a programming language. It's huge. And, um,
Pat: 25:23
broader
Dean: 25:24
Yeah, I mean, you just, there's so many multiple uses for it. part in the networking there could be there's other uses for it in multiple different industries. And yet trying to cross apply those skills
John: 25:36
It's so
Dean: 25:37
to manage.
John: 25:38
I know with blender. So a 3d animation software. It has BPY, blender Python. So if you're into like 3d worlds and doing animation software, things like that, and you want to get into it. I T is broad. I know we sort of had this narrow focus of Cisco network compute storage, cloud security, there's project management. There's there is um, I don't know all kinds of different niches and facets to it. And then if you look at the, you know, the development side of things, there's front end development, back end development, middleware, uh, you know, so it's, there's just the digital world. There's a lot of ways to participate in it and, and make careers in it. And, um, I tell you what, when I went to the factory there was not, there was a chance that I, I could lose a limb or lose my life. People in that factory that I worked in were killed on the job. I had never put my life on the lines is moving in it. Okay. CD ROMs, don't come flying at you taking your head off in the data center. It's a nice, safe. Comfortable living where you're working in, in a pretty good environment, you're a knowledge worker. So you're there ideas and your thoughts and your problem solving. I find it to be extremely rewarding just in general, as a career, if you're wondering, do you really want to be in it? It really is. You enjoy solving problems. And do you enjoy using digital tools to solve those problems, Right.
Dean: 27:27
Yeah. Even to be honest with you, I solve my most intrinsic problems with pen and paper still. Believe it or not. I don't use a computer to solve problems. I used to come,
John: 27:37
paper? Yeah.
Dean: 27:38
a computer to,
John: 27:39
code. Most of uh,
Dean: 27:41
them. Yeah. It's all pen and paper. You just can't beat that. Sadly.
John: 27:47
Cause you were talking about small tasks and automation and getting started most of my stuff does start as human ideas in pen and paper that write pseudo code. And it's not even really pseudocode, but right. Can I connect to the device, connect to multiple devices, authenticate issue, a command, right the memory, right? Like it really is just a sequence of ideas. And then it's a matter of finding the syntax. And that for me still means stack overflow and Google and API documentation. I don't just know this stuff. I have to look up a lot of it, but right. What is the syntax to make a for loop and add four VLANs instead of one VLAN right. Can I map one? Okay. Now can I add multiple VLANs and of scaffold on your success and over time you get better and better at it. Like any other skill, you know, like subnetting, if we just throw that out there, people say, I can't, I don't know anything about automation. Well, before you got your CCNA, you didn't know anything about subnetting either. And when you started subnetting, did it not seem like the most impossible thing? How, how did people just look at this and know that I need a router or I'm in the same broadcast domain? 15 years later, I can basically look at two IPS and say they need a router or not, but that's a experience.
Dean: 29:15
for sure. Yeah. I agree. Even with a little automation, I've uh, I done, um, like you said, starting real simple connecting to a device. I just changed the name of it, right. To memory reload it. And then you just build on top of that, you take that function, you add a little bit more, it breaks you fix the function and then you create a new functional ideas and you reuse a bit of code there. And then you're like, oh yeah, this is like a full fledged thing. And I don't actually have to do anything anymore. I can just press run and here we go. So it is, I agree with that. It is just baby steps and just playing with it really just like, have an idea in your head and yeah all I wanted to do my one accomplishment for the day was all I want to do is just change the name of this switch to my name. That's what I want to do. And I don't want to press a key. I just want it to ask me what what name I want to name it. That's all I wanted to do. And it was just look that up, people have done that. And, um, yeah, just implement it and try it. Try and trial and error. A lot of programming, I kind of feel was trial and error. So, um, yeah, that was that's my 5 cents worth. But, um, I had a really good time, programming in, uh, Rex Python and a tiny bit of Java too. So yeah,
Pat: 30:35
Yeah. I do feel like the network automation space is really going to take off in the next couple of years. and it's here now, right? it's not, that's not to say it's coming. Right. it's here. We all know it's here. I just think that I think that network automation is going to be another tool in your tool. If you're going to be on the network side of things too, to know, right. As far as you know, how do you make things more efficient? how do you free up your time? how do you, reduce human error? Right. Cause a lot of, you know, I saw a a statistic somewhere. I can't remember the actual thing, but the gist of the statistic was a lot of outages. Big, broad outages are caused by human error. So if you can automate that, if you can automate that test it automate it, make sure it works. Then you're eliminating the human error aspect of, I put a static route in wrong. Oh, now I blackholed YouTube, you know, that kind of thing. So
Dean: 31:27
Yeah, but you have to be careful with automation because you can, you can faster too. It's like having a machine gun going wrong. Like you can Like if your programming is crappy and you automate stuff, you break it and trust me, you break it fast and fix it, fix it. Make sure your code is tip top. But no, I agree with you in that saying, and, um, what also led me to automation was, uh, we had a problem at a clinical facility where we had 52 switches where where I had to basically get the show text and I don't want to log into 52 switches. That's just
Pat: 32:03
Yeah.
Dean: 32:04
So you, it is stupid and run the same command over and over again. You just like, it gets to a point where you just. It's human nature. That takes you become lazy like that. There's got to be a better way for that. I can't, I can't do that. Like logging into 52 switches the same username and password and doing show tech, capture that output and put it into a text file is relatively simple. It's brain dead for a human. So it's a very, it's a relatively simple task in automation to do providing you sort of have, your kind of like cooking or bakery steps kind of deal. That's how I kind of view it. So just, it just, like you said, frees up time gives you more
John: 32:47
You, nailed it, Dean. So that what I like about what you just, said there, I like how everything, everything that both of you guys just said, I'm like nodding in agreement for anyone who can't see me with a big smile uh, one, like I am lazy That's the secret driving me. If I had 50 switches and they want the show tech, you go really like I'm gonna spend my whole day logging into these gathering. Now just think of the friction there. Did I hit every one of the devices? Did I save the file with the right name, for the right switch that I was on for each one of the, that's still not a trivial thing for a human to do successfully manually. Right? There's a lot of opportunity to get that wrong. And then automation comes along and it's like, you press enter. And because of the speed of Python that would be done in a few seconds, I really would cause pike. Yes, like forks and connect stuff all 50 at once. And the other thing is Dean it's safe. It's safe. You didn't start with like, can I change the default route everywhere through automation? Can I add you? Right. Like some people get really hung up on automation equals configuration I think configuration management is a part of network automation. It show is perfect at scale. What's the one thing. Every network is lacking documentation. What is the perfect use case for network automation, documentation? They go hand in hand. We can solve long standing problems. We can, and I don't want to say lazy. I can focus on higher value propositions for my organization. They're not paying me the money they're paying me because I know the show tech command and know how to log in to 50 devices to run it. Like that's really not adding the value to your enterprise. If you were to step back and say, give me an hour and I, and I'll script this in something we can reuse every time we need a show tech from every device. Like it's, there's so much value there. Dean
Dean: 34:59
issues. Yes. Huge value. Like you said, you can reuse that you can give that to lower tax and part that knowledge show them automation stuff, it just builds from there really. And then everyone in the organization and not just the text but leadership management, they see the value in what automation really does provide and how much it's, we're able to leverage by just piecing together. Some code is, is huge.
Pat: 35:26
Yeah.
John: 35:26
Yeah. And pat mentioned, pat said you brought up tools, pat, and that. That to me is like, like, you know, you had mentioned a Dean, like 20 years ago. It wasn't even web 1.0 or whatever. And it's kind of funny how long I've been around, but like almost overnight, I've gone from Putty and notepad, maybe notepad plus plus to vs code and WSL and Ubuntu and Docker and Kubernetes and get an Azure dev ops and GitHub and right. It's like, holy cow, we finally have more than just Putty right. And a console cable to do our jobs. So pat, I think you nailed it, that the tools are here now that empower us to work a little bit differently moving forward.
Dean: 36:20
yeah, we have virtualization now via the cloud. We can leverage even though, I hate the cloud, but yeah, we have enormous amounts of tools now and your come up was a lot harder than mine and Pat's, I got to admit I'll come up was difficult. The fact that like, yeah, you guys barely had like a shoe string and a can so it was like, it must've been hard to just the point I'm making is it was harder to go off and do your own research. It was a time where like, if you wanted to write a paper, you have to go to a library and go find the book and really research that night. The accessibility of the internet, even computers is just wasn't available as it is today.
John: 37:04
today. Yeah. It's funny. You mentioned that I I'm in a shopping mall in my hometown where I live near my apartment and I was broke. I was just at a college, but I was trying to find resources to study for these certifications. There was like an old used bookstore that happened to have. A small computer book section. And they had like the exam cram books that I needed for my A+ at the time. It's so funny. You mentioned libraries pop because I remember finding these discount bookstores that had like, you know, big sections of the, you know, the big, giant reference books for your A+ at the yeah, I think
Dean: 37:49
I remember the computer magazines too.
John: 37:52
Yeah. Computer magazines and all that stuff. Yeah. It's a whole different whole different
Dean: 37:55
well, but that literally you have to get your information then like latest tech from like three and a half inch floppies and like the freebies they used to give you and all that freeware like it was, it used to be a grind and you thought you were special when you had that stuff. And that was like a 10 bucks subscription. Like, you know, I mean, like you had to, you're going to be poor like getting computer magazines. Could you want.
Pat: 38:20
yeah, I hear ya. Yeah, you're right. I think John was hanging around with a chisel and a stone, trying to write code back in the day on a stone tablet or whatever. But no, I agree. I think Dean and I's come up where
John: 38:31
I was just going to say I had to type park before I could shut off. my first computer to park the hard drive heads from spinning. So that's the date myself, my first computer. You had to type. Before you brought the power off on the computer, or you would fry the hard drive because the heads would still spin. So you had to say park to stop the heads and then you could successfully power down and I, I love, it's funny. I have all this gray in my beard. Now when I was a young man, we used to derogatorily say the gray beards, the old guys that worked at the company 20 years from the mainframe days, we all called them the gray beards. And here I am 20 years later And I've got a literal gray beard. I can't believe it. so ironic.
Dean: 39:22
so the way it goes, it's definitely the way that's funny though. The fact that you have to part, I don't even know what park in the heart, this is where that is beautiful. It sounds like rewinding VHS to me.
Pat: 39:33
Uh,
John: 39:35
it's yeah, same era. Same era Dean, same era. And if anybody listening wants to know what a VHS is, it was a video cassette tape that, yeah. I used magnets, same era Dean, same era.
Pat: 39:50
Oh
Dean: 39:51
bad was it when you wanted to watch your movie and no, would rewind it. I used to be the advice. Yeah. I wouldn't let someone wants to watch a movie. They pull it back and they never hit the rewind button. Yeah. But no, that's a fantastic story. Fascinating and there's a lot of insight there for sure. Absolutely. So what are you doing with yourself now? Job
John: 40:14
I'm currently at the Canadian parliament, the house of commons of Canada. I'm a by day. I'm engineering and architecting, integrating their network all the time with new things and automation. And then in my free time, I I'm publicly learning Python and working on some open source projects. I've done a lot of mind-mapping. If you will look at markmap.Js.org is the website. This technology will take any markdown file and turn it into a mindmap. So I've been using things like the rest APIs on the internet or something like DNA center or ice and getting the JavaScript back from their API and turning it into mindmaps. It's actually been a pretty successful, it's a nice format actually for displaying network state information. We can also use PYATS data. So Python, like you mentioned, getting the show tech command. Well, certain Python libraries, let us get the structured data back that we can then mind-map as well. So like a routing table or something like that. So I've had a lot of fun with that
Dean: 41:30
Is that depending on the API you're interfacing with?
John: 41:32
Yeah, it depends on and you can do anything major league baseball's API to Pokemon's API star wars has an API. There's a lot of fun. APIs out there. If you're, if you, like you said, you want to be an in IT, play with some of your favorite. If you have an interest or a hobby, there's likely an API on the internet. There's one for space. There's one for Marvel comics. There's one for the X-Men you name it? There's probably an API? to play with. Yeah.
Dean: 41:59
That's a good learning
John: 42:00
then download a program called Postman a nd then start making API calls to some of your favorite APIs just to, just to start playing with the cloud and APIs and Python and postman. That would be my takeaway if you're really want to get started. And you don't know where to start, just Google API, and maybe how many people are in space. That's a good one. If you Google, how many people are in space, they'll show you the API call you can make, and it will come back and tell you from Nassau NASA's database,
Dean: 42:36
that's solid.
Pat: 42:37
I love space. I'm such a space nerd and it's not even funny. I love it.
Dean: 42:43
pat is on the case.
Pat: 42:45
I am, I am on the case.
John: 42:46
All that you should see the, you should see the NASA library of APIs they have the weather on Mars. So does the probes and the Mars rovers feed API data back to NASA and they make it publicly available. You could find out what the temperature on Mars is right now from the API. It's pretty cool.
Dean: 43:06
it's
John: 43:07
I say pretty cool. Maybe cool's not the right word for a space API,
Dean: 43:13
I'm just being cynical. I'm just messin' but I'm not actually really insightful. I actually had, no, I wasn't aware of any of that good stuff NASA provides, but that's um, that is actually a great way because it's practical when I liked your Marvel one, because, um, I kind of feel like you can make your own, like super hero or you could call out your own iron man or your own Spider-Man or something like that. But that sounds really, uh, up my street.
Pat: 43:43
what's wrong with the little Superman right now. And then, you know,
Dean: 43:46
DC
Pat: 43:47
That's true. That's DC. All right, iron man. So, John, I wanted to turn a little bit,
John: 43:53
Although I heard the new one with John Cena is pretty
Dean: 43:56
Well, peacemaker, I haven't seen it yet. I've heard. It's good though, too.
John: 44:00
Yeah. I
Pat: 44:02
One. I'll have to check that out, but to. I wanted to turn a little bit to you wrote, you wrote a network automation book back in, what, 19, something like that, to that 19. And I was reading I'm slowly making my way through it, and I'm about a quarter of the way through the book. And it's really well done. So kudos on the book and, you know, being a successful author and all that comes with it, but can you take us a little bit through the process of, you know, why the book, was it just a project for you? Was there something more than that, or your thought process around the book and then.
John: 44:33
Sure. I, I have, I read a few other books, so I would recommend Jason Edleman and Matt Oswald's and I'm missing another author there, but network programmability and automation. Once I read that book, and once I had done a few, once I had about a year of successful network automation with Ansible under my belt, I felt I had found kind of the formula and I wanted to share it. I really just wanted to share. I think you can use IDE vs. Code with. And it get repo and his model template approach with Ansible to be successful with network automation. I don't think it needed to be any more complicated than that. I thought I had enough there and enough of a journey from an introduction to a CICD pipeline. Then it would be worth people reading at the time and even still today. So my wife was my editor. She was like excited and we had a lot of fun. I would write a chapter, she would read it. And she's not into it at all, she doesn't know anything about any of this stuff, but she, you know, she would read a chapter and I would write another chapter and back and forth. And um, I sort of reach out to a few different publishers and I kind of shopped around. And then I decided just to self-publish on Amazon, because it was a pretty easy process. You download the Kindle create software. So if anyone listening is an author would be author, you want to write a book, you've got something you want to say, it's pretty painless. You get this Kindle, create software, you bring in your word file of your book. It, it puts It into chapters. It gives you pagination and page headers, and it gives you the option to let people print it, buy a physical copy. It was all pretty painless. and then I kind of just released it and went from there. And it's, I really, you know, I, it had kind of the response I was hoping for. I would do things a little differently now, you know, obviously any, nothing is perfect and you know, it was my first attempt and I did learn a lot along the way, but in terms of, if you're listening to this. And you want to be in it. I know the market for this podcast is people who want to get in to it, honestly, reach out to me on Twitter. So is John_Capobianco and I can get you a PDF of the book. No problem. Just reach out to me. No questions asked. I can send you a PDF and if you want to buy a copy down the road, that's fine. if it, if it would help you get into it or get a sense of what network automation is all about, you know, just say, Hey, I heard you on the so you want to be it podcast and I'd like a copy of the PDF. No problem.
Dean: 47:25
that's huge. Thank you, John. Yeah. Hear that listeners? Free copy. That's amazing. Huge.
Pat: 47:32
all the stuff you can get.
Dean: 47:34
I know. So quickly, John, the way you wrote this book, do your scripts or programs run on your laptop or machine itself or do they run on like. And node in the network or do they run natively on the networking gear? How, how, or if you've got a mixture of both, because I know so many people would do script all my automation and it's different flavors for different types of hardware and some have sort of like a API built in some pieces of hardware. Don't is there an, like an, approach that fits all I've done a little bit, but I don't have like So is that something you could share with, uh, myself and our listeners?
John: 48:22
sure. So if it's hard to do, uh, uh, fits all solution in, unless you have a single vendor. Right. So even if you have Cisco, if you're just a Cisco shop, you still might have NX OS and iOS. And IOS-XE and iOS XR all with their own little subtle differences in syntax, all with their different APIs or no APIs, NX API in the data center or rest confident the campus, my most successful approach is to use the PI ATS library. So if anybody listening, if you want to Google PI, P Y ATS, all one word that's the Cisco open source. And it originally started as a testing solution, but you can do everything from configuration management to show commands. if it has a parser, it will give you the JavaScript object and patient back. So meaning it's almost like having an API. For show commands. So you can do showip ospf neighbors, and it will give you a JavaScript back of the neighbors. Yeah. And then, because it's JSON you can work very easily with it with templating languages like ginger two or with Python or right. It's a pretty easy format to work with, more so than standard CLI output. Right? So sort of the formula that I found is PYATS and you can do things like PI Ts dot device dot configure, and then pass it the command you want, you know, vlan 20. So it's not limited to just testing or show
Dean: 50:17
So that was my question. Does it, does it do an API call and then the API called, does the configuration show in the backend, or is it automation where you've kind of running the program and then you're kind of sending commands to the device? Is it, if that makes sense, because I've seen both I'm not saying one's worse than the I've just like, two different approaches Okay.
John: 50:45
Okay. Um,
Dean: 50:47
Gotcha.
John: 50:48
approaches for sure. And, if all things being equal, I prefer working with an API, but really there's no difference because it's the Java script here. Dean, the, the key to all of it is structured data, right? So whether or not it's a parse CLI command or an API call, it's the fact that it's structured data that makes the difference over. You know, standard running commands at the CLI?
Dean: 51:23
Kind of filled that the APIs return better data because sometimes, but then I suppose if you're doing real command calls, you get raw data, which you don't really get from the parsar, it's it's I suppose it's what you want to, what you're achieve, I suppose, is it's all horses for courses and that was, yeah. Interesting. Very interesting.
John: 51:47
Well, I hope I didn't do too much talking tonight, I get really excited and get a little carried away. So if you have any other questions or anything else that you wanted to cover before we wrap it up?
Dean: 51:57
Nothing else from me. no, anything for new pat
Pat: 52:00
no, I'm good. Yeah, we're just about at the hour. So I was gonna do the, the wrap-up spiel and, get out of here. So I know there's a ton of podcasts out there these days and you have your choice, so we appreciate you hanging out and, doing one, listening to us and John, I as always, man, this is, this has been awesome. Thank you very much for coming and hanging out. it's been a pleasure talking some automation stuff and, you know, talking to, just a wide variety of topics and things of that nature. So, an absolute pleasure, man. we'll certainly have you back, you know, in the near future and talk some more.
Dean: 52:35
for sure. Yeah, it's been super insightful. Thank you, John. And thank you for the book and all the insight and great info you shared with, uh, ourselves and our guests to spin. It's been wonderful. It's really has been, it's been really good.
John: 52:49
I, can probably talk with you guys for another hour or two easily. I, really think what you're doing for the community is great. And, um, it changed my life. And if I can give back, uh, any, you know, in any way at all to the community, I'm happy to help. So I'd love to be back. And I I'm really excited to follow your success in 2022 guys.
Pat: 53:12
Yeah, thanks a bunch, John. We really appreciate your coming and hanging out. So that's gonna be it for this, for this week, for the show. just a quick note. we were nominated as a finalist for the Cisco champion, it blog awards. So there's some really great content creators out there that Cisco has picked, to, to win the prestigious Cisco champion, it blog award. So, we are a part of that. So, the link to vote for your favorites, and I think they give you a couple votes. They break it out into actual bloggers and then podcasts. So we're obviously in the podcast space, that's where we play in. So I think they give you like a vote of five each. I think. So, if you like the show, which we hope you do, there is a link to the vote or to the website in the show notes. So go and vote for your favorite, content creators. And, we hope that's us and some other folks out there. So, that, and then we also have a survey for you folks that just helps us with some of the content, just some general questions out there. Like what day of the week do you listen? Or, you know, what do you like best about the show technical content or interviews? Yeah, gas, things of that nature. So just helps us tweak the show a little bit, get some feedback from what you folks like or don't like, and that helps with the possible advertisers down the road and, things of that nature. So if you want to check that out, that is, S Y w B I t.com/survey. and then our Facebook and Twitter handles and all of our socials, that'll be in the show notes as well. we always have our website, so you wouldn't be in it that calm. You can subscribe to the show, right? From there, some of the bigger popular platforms that we have links to, iTunes, Spotify, Stitcher, Google podcasts, things of that nature, or just playing an RSS feed that's up there as well. So, you could do it that way. If you prefer, throw us a rating on iTunes that always helps with our algorithms. It gets us to the closer, I should say to the top of the list, top of the charts, things of that nature, and, Yeah know, just tell a friend, right? Word of mouth works just as well. These days in the big, bad technical world that we live in. So word of mouth is just as good. So again, Instagram, Facebook, Twitter, all of our socials will be in the show notes, grab them there. The discord invite. we have a discord server for you guys to come hang out in. So trying to build that out a little bit more and more attraction, over the last few weeks on that. So that's always been a good discussion in there, during the day. So, the invite is in the show notes as well. And, that's it again. Thanks John. Thanks Dean. And we'll see you folks next week.
Dean: 55:43
Thank you. Take care of John. Thanks again for coming on. and uh, yeah, we'll be in touch shortly. I've got lots to learn.
John: 55:52
I'm always open. Thanks for having me.