#88: 2020 was quite a year. Your year may have been great or it may have been horrible. Hopefully it trended towards great, even if did not live up to your expectations. In today’s episode, we look back over some of the items that we encountered and that you probably did too.
If you like our podcast, please consider rating and reviewing our show! Click here, scroll to the bottom, tap to rate with five stars, and select “Write a Review.” Then be sure to let us know what you liked most about the episode!
Also, if you haven’t done so already, subscribe to the podcast. We're adding a bunch of bonus episodes to the feed and, if you’re not subscribed, there’s a good chance you’ll miss out. Subscribe now!
Viktor Farcic is a Principal DevOps Architect at Codefresh, a member of the Google Developer Experts and Docker Captains groups, and published author.
His big passions are DevOps, Containers, Kubernetes, Microservices, Continuous Integration, Delivery and Deployment (CI/CD) and Test-Driven Development (TDD).
He often speaks at community gatherings and conferences (latest can be found here).
His random thoughts and tutorials can be found in his blog TechnologyConversations.com.
I'm going to be mean now, but I'm yet to see a virtual conference and I haven't seen all so don't doesn't mean that yours is like that, but I didn't like any of them.
This is DevOps Paradox episode number 88. DevOps in 2020 - A Year in Review
Welcome to DevOps Paradox. This is a podcast about random stuff in which we, Darin and Viktor, pretend we know what we're talking about. Most of the time, we mask our ignorance by putting the word DevOps everywhere we can, and mix it with random buzzwords like Kubernetes, serverless, CI/CD, team productivity, islands of happiness, and other fancy expressions that make it sound like we know what we're doing. Occasionally, we invite guests who do know something, but we do not do that often, since they might make us look incompetent. The truth is out there, and there is no way we are going to find it. PS: it's Darin reading this text and feeling embarrassed that Viktor made me do it. Here are your hosts, Darin Pope and Viktor Farcic.
So what else can we say about 2020? Dumpster fire?
One star. I'd give it one star only.
Give it one star. Did you give it, do you give it a full one star or does it only get a partial one star?
I'm I'm imagining the system where actually one star is a minimum. You cannot give no stars or, or, or half stars. Yes.
So maybe that is how you have perceived 2020. I have to admit at times I have, but I've also looked at all the things that 2020 made possible. One of the things is proving to people that work can actually happen outside the four walls of the office.
Exactly. That's the year we will be looking back, years from now as the year that made remote work and remote learning a common thing. Few years from now, we'll be talking about how schooling changed drastically and we will be pointing to this year. Same thing for work and many other things. This will be the year that Amazon will officially become a country by itself in terms of revenue and all that stuff.
That will be an interesting number to see when their results do come out. I wonder where in the national GDPs Amazon will be. That will be interesting. Let's think back a little bit more on 2020. Think about all the people that we talked with. Think about all the other things that happened just within our little microcosm of DevOps Paradox. Really, 2020 has been a paradox, right? It's been the best of times and it's been the worst of times to break out a literary reference.
Yeah, but it's been a very good year for DevOps. I think that the whole we're all working remote sped up quite a few things that normally would go slower than they went in this year. Switch to cloud became not a conversation anymore. It becoming mandatory. Using services for everything you can. Giving more responsibility to people, hence DevOps. I think that the positive side effect of this year is that it forced companies to do the things that we all knew they would be doing. But now, there are no more excuses, no more conversations. Hey, should we go to cloud or should we manage our own data centers and access to those data centers only from offices in which nobody is right now.
and probably will not be even into later 2021.
Yeah and those companies who try to force people to go back to offices after COVID passes are going to have real trouble. I don't know how it's in US. There has been actually exodus from huge cities. There are a lot of people who came to live in Barcelona or Madrid, and I'm talking for Spain now, because that's where job opportunities are and they accepted that they should be paying much higher rent, have much higher cost of living, compared to wherever they were born in because of the job opportunities. Now that everything is remote, there is a huge number of people, significant percentage of people going back to their hometowns or going somewhere else and they might not want to come back if that same company says, okay, now you do need to be in office because other companies are going to say, hey, we're okay with remote. Come to us.
We've talked around that on one of the episodes, in fact, fairly recently, when we were talking with Olaf about where you live shouldn't define your pay because as some people have moved from in the US from the coasts, whether it's from the West coast or from the East coast and move more into the Midwest, I mean, sure, some people may have their salaries adjusted for whatever reason, but quality of life for a lot of people, especially if they have families, can tend to be a little bit better than living in the big cities.
Yeah. I wouldn't be able to classify big cities as worse quality of life or better. I don't think it's much about that. It's more about being able to choose where you want to live independently of location of your company. If you like in big cities, that's great. If you like countryside, that's great. What is not great necessarily is that you have to move somewhere where you might like to be less because that's the only place where companies are. Silicon Valley.
So if I would have done my homework, I would have listened to our four predictions podcast for 2020. Of course I didn't do the homework and that was pre transcriptions. So I can't just quickly go and look and see what the four points were. But thinking about it a little bit, at least one of those had to be serverless. Do you think serverless really made any kind of strides or did it go in a direction that we didn't expect in 2020.
In 2020, I don't think that serverless wasn't the hot topic. It's not something that everybody spoke about. That crown is still on Kubernetes head. On the other hand, there is undoubtful, drastic increase in usage. Maybe not drastic compared to percentage, compared to previous year, but there is a very clear increase in usage and that tendency will likely go up. The reason why there is increasing usage, because it's useful for certain scenarios, but the reason why I believe we are not talking much about serverless is I believe that 2020 is the year where companies are starting to rethink how that concept should work and we are not necessarily ready to have those solutions. Still a planning phase, I would say. Just to clarify, when I say serverless, I don't mean only functions. I mean I don't care about underlying infrastructure. Here's my code. Here's my image. Run it.
We're seeing that more and more recently. Amazon LightSail added that feature in to where here's my container. Much like ACI. Much like Heroku. Here's a container. Go run it.
Exactly. We're in a way going, I think, back to where cloud was maybe five years ago. It was relatively simple back then. You create a virtual machine or a few virtual machines, you copy your binary and that's it. A child could do it. Then we drastically increased the complexity in the meantime with containers partly, with Kubernetes, definitely. Now I think are going back from simplicity perspective. Now, of course, where we are right now, we are accomplishing things that five years ago we couldn't dream of. So it's not that now it's more complex for no good reason. But we are moving towards simplification of the new increased complexity and that simplification will most likely be under the umbrella of serverless like LightSail that you mentioned, right?
Yeah. Right. And it's not necessarily functions. It's just serverless. Run it, keep it simple and then go do it again and make more money.
Yes. There is still somehow silent, I think, not a big bang, but if you look at the products that are coming up, yes, it's shaping up into an explosion. We'll probably speak about explosion in the next, that will be 2021. Next episode, right?
Yes. If you're right here in the middle of this one, our next episode is going to be our predictions for 2021. So when you're listening to this or the release date for this as the last Wednesday of 2020. Goodbye, 2020. Hello, 2021 type of deal. That line of demarcation moving from December 31st to January 1st, we're still going to be working from home. Those things are still happening. But in 2020, did Kubernetes truly become the real hotness or is it now a toddler? When the baby is birthed, it's like, Oh, it's nice and I can swaddle with it and it's beautiful. But now that it's a toddler, it's having its tantrums and it doesn't know what it wants. It just knows that it wants it right now. Is that where we're at with Kubernetes? Let's put it in one phrase. Let's look at the CNCF landscape poster.
Insane. That's the work of a lunatic. It's absolutely insane. I'm sorry if I'm offending anybody. It's supposed to help us. That's not helping anybody. Is it already three digits? It must be three digits.
It's gotta be. Yeah.
Just insanely complicated. There is an explosion of opportunities. Kubernetes is the thing. Let's start with there. Every single company that has any self-respect is using Kubernetes today in some capacity or another. It could be 0.5% of the workload, but I don't think that there is a company right now that doesn't use Kubernetes in some capacity or another. Let me actually rephrase. I don't think that there is a company that doesn't use Kubernetes in some capacity or another, and will exist a few years from now. Was that harsh?
Probably, I don't know that it's realistic because there will be plenty of companies that aren't using Kubernetes that will still be around because their primary software is still running on mainframe.
Not even Java? There must be something right. 100% mainframe?
Not a hundred percent, no. But their moneymaker is mainframe.
Of course. But I was referring to there must be some, three people in that company that are playing with Kubernetes right now while we're talking with some pet project, something, right? Somehow and those same three people might be same type of people that were playing with Java five years ago and that company now has some applications in Java. Still maybe 3% of applications and 97 in mainframe, but there must be something.
Maybe. Let's also think back to 2020. We had our 50th episode, which I need to plan for our hundredth episode because right now this is episode 88. We had the godfather of DevOps, our phrase, some other people refer to him as that, Patrick Debois. I'm hoping Patrick, that you're listening to this right now and you're willing to come on for 100. I think Patrick should be our divide by 50 guest. That's what I want because Patrick's had a lot of changes this year because he went to Snyk in 2020 if I remember right.
I don't know when he joined. Was it this year?
I think it was. But regardless, it's always interesting to see what's happening with Patrick. We had Carlos on. Carlos Sanchez, one of our old time friends talking about his day job. If we, and we didn't say the name of who he worked for, but go out to KubeCon US and once they actually release the videos, which they may be out by now, I'm not sure what their release schedule was, to where he talks about running multi cluster at a global scale. Would you want that day job?
That company is too big for me.
it is, but ignore that for just a second. Would you want to be the one right now where Kubernetes is at right now? He's probably running 1.15 or 1.16. Probably. Would you want to be the one in charge of running a globally distributed Kubernetes cluster or clusters because of how they're doing it?
assuming that by charge, you mean I can make decisions as well, not only apply them? Yes. Yes.
Okay. In your case and thinking back now, currently 19 is the current version. So that means 16 came out at the beginning of the year. No, I'm doing the math wrong. No, that is right. There were four releases this year, right? Help me with the math.
Uh, yeah, 16, 17, 18, 19.
19. Yeah, so either it was either 16 came out about this time last year or whatever. All right. 16 or 17. If you have had to have made the decision in December of 2019, and you looked ahead to December 2020, would you have chosen Kubernetes as your platform? In doing that decision-making, would you have chosen if you had to make a decision December 2019, am I going Kubernetes right now or not? Would you have made that bet?
Without doubt. I think that this time last year, Kubernetes has been around for awhile. We have success stories from many companies. It would be a no brainer for me at that time. Now maybe two years ago, I would have doubts, but year ago, end of 2019, yes, without doubt. Because it's important to put it in a context. You say large-scale Kubernetes, would that at that time look like a less challenge than a large scale alternative, whatever the alternative is because it's not about large-scale Kubernetes against five servers without Kubernetes, right? Managing hundreds of servers is a complex thing without Kubernetes as well. Actually, it's more complex if you ask me. Just imagine all the networking and routing troubles that we had before Kubernetes.
Oh, yeah. I consider 16 probably the first really good release. 15 was okay, but 16 was like production ready I think. 14 was okay. 15 was okay. But there was, I can't remember. There were like two big features in 16 that were really, that was like, okay, big checkbox.
In retrospective we can say that actually right now I think that in 18, all the core resources are finally v1 without betas, alphas and what so not as a suffix, right? I think that ingress was the last one to actually become v1 and that's 19 or 18, not sure. 18. I think.
18. I think it was 18. Okay. So we've gone down the rabbit hole of Kubernetes. We've talked about serverless. I would be remiss in not bringing up GitOps. Now, I don't think we would have made a bet on GitOps this time last year.
No, it wasn't big. It was very, very experimental. I think that if you came to me this time last year, and I haven't heard about GitOps but you explained to me in easy to understand way the principles of GitOps, I would say last year or year before that, or any year, I would say yes, that makes perfect sense. I think that principles are a no brainer in a way. Now from tooling perspective, we were not there for serious usage, I think this time last year.
Do you think we're even there right now? We'll go into more depth on that on our 2021 episode, but do you think we're even close right now or are we closer?
Yes. I mean, we have tools that been around for a couple of years now. Used by both small, but also big and serious companies. I think that we are there right now, just as we were there this time last year with Kubernetes. Of course, three years from now, we're going to look at now, now and say, Hey, that was silly. Was it really? Yes, those tools work decently well, right now.
We had some courses come out. We finished up a book. Launched a new book that's still at this point still being added to. Did you ever see yourself producing courses, at least in the way that you've produced them?
No, I didn't. I don't even remember any more how that all started. I saw myself much more comfortable writing than in a video format, but then just happened.
well, it just happened, but why do you think it happened?
I don't remember anymore.
I don't remember. I think probably part of it was since we didn't have to travel anymore, it changed the way that we work.
No, but it started two years ago, right?
No. or maybe it did. I don't remember anymore.
This is the third course we have, right.
Yes, it's all. Okay. So I can't blame it all on 2020. We talked about at the beginning of the episode. Online learning. Did we even consider online learning being a thing? I mean it's always been a thing. You have the masterclasses. Our courses are on Udemy. Did we expect it to be like the way it's turned out?
I don't think that it really changed yet. It increased in adoption for obvious reasons. COVID what so not. But I don't think that it really changed. We are still missing rethinking of how online learning works. For software engineers, but also in a bigger scheme of things. Because if you look at it, it's not really much different than it was three years ago, five years ago. More people watch it. Yes. More people learn online. Yes. But nobody really sat down I feel and thought, Hey, how would that really work? I think that online learning is still in a way a reflection of school rooms in a way
which is not efficient
No. What I believe more and more is that I don't know whether that's me, but I feel that our attention span is becoming shorter and shorter. We were switching from reading books to reading blog posts, and to now reading tweets, which we can have a debate how good or bad it is. But I believe that things needs to be split into smaller and smaller, easier to digest pieces. I don't think it's going to be, Hey, I'm going to sit down for hours and to learn. Rather, hey, I'm going to figure out this tidbit while cooking and then I'm going to dedicate 15 minutes of real work to it and then listen to it while driving the car and it's going to be spread around very small frequent pieces I think
The one thing that we didn't see coming that I will call out is virtual conferences. Everybody had to change quickly if they were going to have anything and some people have done it well. Most have not.
I'm going to be mean now, but I'm yet to see a virtual conference and I haven't seen all so don't doesn't mean that yours is like that, but I didn't like any of them. The only one actually I liked in a way, which was the only conference done without any real budget and within few weeks of preparation was All The Talks. I think that was the name.
Which was the one that Patrick did.
Yeah, exactly. Man, those conference apps that are being used now. To begin with, they're horribly ugly and they're complicated. I always have trouble finding things in conferences online and how it all works. I'm not very fond of them.
It goes back to the online education model. We're trying to replicate an in-person conference experience online and that's logically doesn't work.
I spoke in quite a few conferences in 2020. To begin with, I don't understand why are they live? They asked me to record it, not to do it live so they play my recording live. I don't understand why. Why would that exist? Why would we have live talks, unidirectional talks, you know, I talk to everybody or somebody else. If it's live, it needs to be about the moment about some kind of conversation with the audience or about commenting on a live soccer match, right, because it's happening right now. The typical talks, why aren't they live? I don't understand.
Well, the one reason that I could say, and it's reasonable, is not everybody has great upload on their internet. So that would be the only reason why I would, if I was producing a conference is the reason why I would want prerecorded talks.
I understand that. I probably didn't explain myself well. What I meant, why would there be a conference in the first place? Why wouldn't those videos just play? Why don't I just watch that, those videos when I want to?
Yes. Maybe, maybe the videos get premiered at a certain time
like, like a movie premiere
Yeah. I would rather have a completely, I'm not sure how to say this. A completely offline conference that isn't being driven by booth duty. I can't think of a better way to say that. But you got to have sponsors typically, but in the case of Patrick, I think he did have, for All The Talks, they did have some sponsors, but it was a very bootstrapped way of doing things.
I mean inject commercials inside those videos. Just launch them with YouTube and say on December 1st, we're going to launch a hundred videos that are well curated from people we selected on subjects that we want to hear. We're going to launch them on December 1st. I'm inventing the day or maybe one every day or whatever, and sponsors embed, inject those commercials inside of those videos. That's okay. I understand. You need to get money from somewhere. Hey, you know what, actually, I guarantee you, if you well execute it and if you measure it throughout the year, the benefits will be so much bigger because you can reach hundreds of thousands of views for each of those videos if you do it right, which is much bigger than any conference can never do.
And by the way, the do it right isn't showing slide decks as your primary presentation model if you're online.
Here's the exercise. Think about three creators on YouTube that you follow. First three that come to your mind, and how many videos of any kind doesn't matter whether it's software or it's about drones or photography or whatever it is, how many of them use slides?
If you can answer that question and your number is greater than zero, you're watching the wrong people on YouTube, including us. Okay. So 2020 without reviewing our old episode, which I should have done more than likely we probably hit on some and we missed on some. So in next week's episode, which will be the first week of 2021, we're going to look ahead and make our bets for 2021. Do you think it will be 50/50? Do you think it will be 10/90?
I don't remember the bets from last year, but I normally don't tend to make those bets like Elon Musk, we will be on Mars by 2030. My predictions are usually very short term. Which is probably not so pompous as people expect. People want artificial intelligence will control your lives type of predictions. Right?
It already is because if you're listening to this episode, probably AI recommended it to you. So thank you for listening. And if you have made it this far as again, thanks for listening this year. If you haven't subscribed, please go ahead and subscribe. You'll hear me say that again here in just a minute, but we do thank you for listening and putting your confidence in us as a couple of people to sort of rambling most of the day and you actually listened to us. You may throw out everything we say to you, but you still listen to us.
We hope this episode was helpful to you. If you want to discuss it or ask a question, please reach out to us. Our contact information and the link to the Slack workspace are at https://www.devopsparadox.com/contact. If you subscribe through Apple Podcasts, be sure to leave us a review there. That helps other people discover this podcast. Go sign up right now at https://www.devopsparadox.com/ to receive an email whenever we drop the latest episode. Thank you for listening to DevOps Paradox.