Erez Berkner - Cloud computing
Erez Berkner is the co-founder and CEO of Lumigo, a monitoring and debugging platform for modern cloud applications.
In this episode, we take a deep dive into cloud computing. With over 16 years of experience working in the cloud, Erez tell us about the history and development of cloud computing, with modern cloud computing focused majorly around serverless and Kubernetes.
We also discuss the key considerations, challenges and most common misconceptions about the cloud, as well as how it impacts the developer experience.
Links & mentions:
“Moving to the cloud will allow you to scale with lower maintenance, will allow you to retain your employees with lower costs. But it is not the end of conceptual IT, ops, etc.”
Welcome to the Agile Digital Transformation podcast, where we explore different aspects of digital transformation and digital experience with your host, Tim Butara, content and community manager at Agiledrop.
Tim Butara: Hello, everyone. Thanks for tuning in. Our guest today is Erez Berkner, co-founder and CEO of Lumigo, a monitoring and debugging platform for modern cloud applications. Today we'll be discussing cloud computing. We'll go through a bit of history, the best practices and advantages of moving to the cloud, as well as the key challenges and the most common misconceptions here. So Erez, welcome to the show. Thanks so much for joining us today. Do you want to add anything or should we just jump straight to the questions?
Erez Berkner: No, sounds like a great topic. Thanks for having me. Excited to be here.
Tim Butara: Yeah, it definitely is a great topic. It's kind of weird that we haven't actually discussed cloud computing on the podcast yet. We're what, at about 70 episodes now and this is the first time that we're talking about it. So, yeah, it's great having you with us, discussing this with us, Erez. So can you first tell us, let's begin with this. In what ways is cloud computing a game changer? Why should companies consider the cloud for their digital strategies?
Erez Berkner: Yeah, so maybe it's a bit of a context – I’ve been, let’s call it, escorting the cloud from the last 16 years. So I've been seeing that through– my previous role was the director of cloud security at a company called Checkpoint. Now at Lumigo, where we deal with observability for the modern cloud. But I think that first and foremost, what I'm seeing in the last 17 years is that cloud computing is becoming definitely like the gold standard when it comes to modern application and new activities, but they also for legacy systems.
And the funny thing is that this is still– at the same time everybody is talking about the cloud, but it's still very early stage. It ranges from say, 10%, 15%, 20% of the world workloads today in the world are cloud, which always sounds very small to me compared to the hype around it. That's like kind of a background of the scope of cloud computing today.
And to your question about, is it a game changer? So I definitely think it is, just because it provides so many abilities to scale, abilities to execute in a whole different level compared to what was there before, like the on premise implementation of your application and servers. So we're going to get into that probably a bit later. But it is changing the way people scale, people operate their application, the cost structure. So in any of those, it's a real game changer.
Tim Butara: So you said that you've been actually working with the cloud for what, 16 years, 17 years? So you must have seen the entire evolution of the cloud. And can you give us an insight into a kind of comparison between what cloud computing used to look like when you first started out versus what the modern approach to cloud development is now in 2022?
Erez Berkner: Yes, absolutely. And it did evolve quite a bit. So early on, the concept of the cloud was, hey, you have those servers in your server room, and you're not a server expert, right? Like, you're a company, you're focusing on your business. You're expert in your business, but you're not an expert in running servers. So that's one.
And the second is – you don't really know. It takes a lot of time to set up a server. So if you need a new server, you need to plan for it. You need to order the hardware, you need to make room for it. Networking, application, operating, it takes several months sometimes to set up a new server. So if you need to scale, you're limited, that is, to scale and the responses you're giving to your business needs.
And then came, with the public cloud, AWS back, I think, 2005, 6, and said, okay, you're not expert in managing servers. We can do that for you. We have the experience from amazon.com. We're doing that. So let's do that for you. You will be able to rent a server, and that server will be a VM, and that virtual machine will have the server room, will take of cooling of networking, we'll do all of that for you. Just get access to a server and that's agile and that can scale, and that's from today to tomorrow. It can be instant.
So that's where the cloud started. Allow me to rent servers instead of buying them. Ever since, this concept really evolved to the next level, and we're seeing in the last couple of years the motion from rent me a server to rent me a service. And that's maybe a natural evolution of that tendency of you're not an expert in, so let us do it.
But today we see more and more Kubernetes, more and more serverless, more and more managed services to the point where when I need to set up a new web server instead of setting up, renting a server and building that I would rather use a service like API Gateway, for example, from Amazon. And everything's been taken care of for me and I don't need to do a lot on the technical part. I'm consuming it. Same goes with, not just AWS – like PayPal, for example, Stripe. I don't want to set up a payment service, I want to consume it.
So this entire concept of don't build it, use APIs to consume it as a service, this is really, really growing with the cloud adoption. And that's not just for services like I mentioned, it's for databases, for queuing, for web servers, for probably any service you can think of, you can get today as a service, which is what we refer to as a modern cloud.
Tim Butara: So let's say that the company does decide to move to the cloud because they see the obvious benefits that we talked about and they see how they will be able to scale and they're able to see maybe the long term ROI. So what should they prioritize or what should they keep top of mind when making this decision to move to the cloud, to make this cloud implementation as successful as possible?
Erez Berkner: Yeah, so there are several things that really change when you move to the cloud. First of all, the structure of the organization changes. All of a sudden in the cloud because it's APIs, because it's code, developers are dealing with infrastructure, with configuration. So it really collapses this separation between ops and dev. So you hear more and more about DevOps and developers that are actually doing operations. I think it's very healthy because it's well connected, but this is definitely something that is going to change the structure of the organization.
And I think the other thing that changes or to consider is that the cost structure of the cloud is very different. In the past, you can forecast very clearly how much you will spend on servers because you bought them. Now when you rent them, when you pay by the service, by the minute or even by the request, it's agile, it's flexible, you can end up like electricity. End of the month, you'll see that your bill is high, but it's hard for you to forecast that. So those are just two examples of the things that are to keep in mind we're moving to the cloud.
Tim Butara: What about– a similar question would be, some of the things that you have to consider and keep top of mind are also the challenges and the potential issues. So what are the biggest challenges and maybe potential problems of implementing cloud?
Erez Berkner: So I think one of the main problem in this migration is the team. Keep in mind that the team that knows how to maintain servers does not necessarily know how to work with the cloud. It's really, really different and you need different skill sets. So you need to educate the team, you need to grow the team. You need to bring experts that know how to work with API, with cloud providers because it's not the same type of writing code, let's call it. So that's one challenge. You need to make sure that you're ready for that.
Second is, I mentioned the forecasting. The cost structure is different, so you need to be prepared for that. Otherwise, like, you're going to have one very frustrated CFO because it's very hard to control costs and forecast. So it requires a new model of financing and planning and flexibility on the cost side.
And the last thing is the control. When you're running your own servers, you're in full control. You know what's going on. You can monitor everything very easily because it's your server in the other room. When you're going to the cloud, it's not your server, it's not your service, you're using other company services and all of a sudden you're blind to many of the things that are happening. And those cloud architectures are very, very complex, especially the modern one with serverless and Kubernetes.
So acknowledge that you will lose control by definition and this is something that is changing. And now, how do we deal with that fact that we're losing control? That's not a bad thing, by the way. You don't want to stay in control of everything because again, that's not your business. You just want to make sure that things are working. And that's the main question. How do you get things to work without having a very specific people that are focusing on controlling those?
Tim Butara: So it takes a bit of a mindset shift from this urge to have everything under your own control, to kind of relinquish this and kind of trust in the experts that you're relying on for this particular thing.
Erez Berkner: Yeah, it's not for people that are obsessed with control.
Tim Butara: I think it's just part of this general mindset shift that's kind of come with this digital information and this digitalization that we're experiencing. It's more, having more flexibility, being more open to change, being more open to handling even unpredictable things. And I think it goes in line with that. And I guess another aspect of this control, especially when it comes to the cloud, is also observability. Right? Obviously. So how can we monitor, how can businesses monitor operational workflows in a cloud-based IT infrastructure?
Erez Berkner: Absolutely. So that's the other side of that. So you lose control, that's fine, embrace it, but don't let go of your ability to monitor. I'm going to let this go only if I know that everything works well. And if something doesn't work well, I'm going to know about it and then I will be able to solve it. So this is where the ability kicks in and this is why it's so important for the cloud, especially for the modern cloud, to have observability baked in, in order to understand what's going on and have the visibility.
So that's all about understanding what's happening. You lost control. There are so many services out there, if you don't plan for it, you will be lost across your services. You're using so many Lego pieces and you need a one view that actually ties everything together and very clearly says this is what happened, this is where it started, this is the process along the way, this is the impact and allowing your developers to kick in if something requires the handling.
So that's a focus on observability of the modern cloud. That's why you're hearing a lot about distributed tracing, about the observability, tracing, monitoring, logging, all of this together come into the new concept of modern observability that is critical for the modern cloud.
And this is what we're trying to do in the last four years, like we've built Lumigo to answer that emerging challenge for the modern cloud from a very focused perspective on the modern cloud. Because what you used to do in the past on your servers isn't the same as what you need to do today in a growing serverless, API-based environment where everything has to be very, very code-based and API-based.
Tim Butara: One other thing that I've also been wondering, we talked earlier about DevOps, about how it's kind of bridging these two. We talked about the challenge of the team and how that changes. And I'm wondering, so how does, keeping all this in mind, how does the cloud impact the developer experience?
Erez Berkner: Wow, that's a big impact. I think, first and foremost, I would look at that from a manager perspective, and I've been managing R&D teams for many years. Even before the actual experience, the perception is also very important. So if I want to grow my team, if I want to retain them, if I want to make sure that they're happy and they're dealing with the modern technologies, I need to be at the cloud. It's not the only reason to go to the cloud. But no, many people don't want to go to a company that is not in the cloud today.
And we're seeing that actually with serverless, we see people that actually, I want to go to the edge of computing. I want to go to the edge of technologies. I want to deal with serverless, I want to deal with Kubernetes. So definitely it's true with the cloud. So that's one point about this very hard, let’s call it recruitment questions that we have with developers around the world.
The second is that this really also changes the developer experience, as you implied. And the point is that you need a new set of tools. You need to know how to work with APIs of Amazon, Microsoft, etc. You need to be able to architect. Every developer now becomes somewhat of an architect. Every developer becomes somewhat an ops because he needs to configure DynamoDB and the scale and the number of reasons, number of rights he or she needs to define the number of cards in Kinesis, in a messaging system.
All of these are things that are baked in via code by developers and are critical. I think that makes developer much more holistic, which I think is a great, great momentum, and again, also changes the roles. I think as we move to the modern cloud, we're going to see less and less of, I'm a developer. I just know Python, and that's all that I know, and lets me just write code all day. But I'm a full stack developer. I know Python, but I also know how to set up a DynamoDB, and I know how to connect those, and I know how to monitor those, and I know how to handle issues when they arrive and what to check and how to debug. So I think this is really opening the developer role and experience to much broader spectrum of the software lifecycle.
Tim Butara: It also positions them much more favorably for the more demanding and more competitive job market today. I mean, on the one hand, we have a higher demand for tech talent, but on the other hand, we also have a higher demand from the tech talent itself. Right. Many more companies are looking for full-stack developers as opposed to a few years ago. And I'm guessing as we just highlighted, that developers who have worked extensively in a cloud environment will be much better able and it will be much easier for them to land these roles or maybe find more prestigious roles in their industry.
Erez Berkner: Absolutely, because that's where the industry is going to. So if I'm– while maybe the cloud share is 20%. If you look at the statistics for openings, you'll see the number of cloud experience much higher than 20%. Because if I need to develop a new project and I need 50 developers and that's a cloud project, then I would try to hire those with cloud experience.
So I think definitely if you're a developer thinking about the future, you want to be versed within new technologies. Definitely the cloud, Kubernetes, serverless, these are things that you probably want to get into and that's very, very hard to find, like, expert in those. Even now, even five years, six years after the Kubernetes bomb, let's call it, it's still very hard to find experts in that. Definitely in serverless. So if I have one tip for developers, if you can get expertise in serverless or Kubernetes, that's gold. That's gold out there, today and in the coming four or five years for sure.
Tim Butara: That's definitely an invaluable tip, Erez. Any developer listening right now, I hope you follow it. But hey, this has been an excellent conversation. I really enjoyed it and I learned a lot of interesting stuff. And I just have one final question before we wrap up. What are some of the maybe most frequent myths and misconceptions about cloud computing and how would you debunk these?
Erez Berkner: If I need to choose one, I would say there is a concept, I think, out there mostly by managers, by executives, that get a feeling that let's go to the cloud and that's going to solve all of our problems. It's like a married couple that has some problems and then say, let's have a newborn and that will solve all of our problems. It provides a lot of value and a lot of excitement and a lot of pleasure and a lot of values. But it's not a silver bullet that's going to solve our problems.
So one of the things that IT organization are challenged with is maintenance. Like a lot of executive thinks, let's go to the cloud, there's no maintenance over there. And I think that's important to understand that even in the cloud, even Kubernetes, even in serverless. Maybe maintenance is lower, but it's still there. You need to have someone expert that deals with that. You need to have the right tooling in place, you need the right monitoring observability, security, CI/CD, entire ecosystem that you need to maintain in order to get your business delivered.
So moving to the cloud will allow you to scale with lower maintenance, will allow you to return your employees with lower costs. But it is not the end of conceptual, IT, ops, et cetera. It's just like a different form of it. So make sure that you understand that before and don't be disappointed after you move to the cloud. And hey, I still didn't need those ops guys, but I moved to the cloud.
Tim Butara: Well, Erez, this has been great. As I just said earlier, thanks so much for joining us today. Just before we finish, if listeners would like to reach out or maybe learn more about you, where can they do that?
Erez Berkner: Yeah, so first and foremost, I just want to add that we're very much focused on the modern cloud, serverless, Kubernetes, containers. So we're doing a very busy monitoring, but at the same time, we are very involved in the community. So if your team has any questions, want to consult on those topics, we have experts that are actually seeing hundreds of hundreds of environments. We know what is the best practices and feel free to reach out for anything.
Again, it doesn't have to be like related to Lumigo, but I think one of the nice things we see out in the community is that a lot of contribution, a lot of community work is being done and I want to offer this also as part of this, so feel free to contact with that as well. The best would be probably through Twitter. I’m Erez, E-R-E-Z, Berkner. Berkner. Or you can just look for Lumigo and feel free to send us a direct message and we'll be happy to help.
Tim Butara: Awesome. I'll make sure to add the relevant links in the show notes, Erez, and thanks again, it's been a pleasure.
Erez Berkner: Excellent. Thank you very much for having me. It's been fun.
Tim Butara: Well, to our listeners, that's all for this episode. Have a great day everyone, and stay safe.
Thanks for tuning in. If you'd like to check out our other episodes, you can find all of them at agiledrop.com/podcast, as well as on all the most popular podcasts podcasting platforms. Make sure to subscribe so you don't miss any new episodes and don't forget to share the podcast with your friends and colleagues.