In this Q&A session with Moshe Belostotsky, Director of DevOps at Tomorrow.io, we dive into the changing role of DevOps and how security considerations are changing the way software is being built and delivered.
Q: First of all, if you can tell me a little about yourself, what brought you into DevOps?
A: “I was in the world of DevOps even before it was called DevOps and before the Cloud became a thing. Ever since I can remember, I have been doing automation, CI/CD, treading this line between infrastructure automation and enablement, automatic tests, and later on, the Cloud.
I started working with automation at the age of 16 and have been doing it ever since with a 4-year break during my army service, after which I jumped right back in.
Q: What do you like the most about working with DevOps automation?
A: “A couple of things.
First, it’s the variety of work: the number of touch points with the platform, with the different teams, and sometimes with the customers. At its core, DevOps is about collaboration. It’s about breaking down silos between development and operations teams so that they can work together to deliver software faster and more efficiently.
Second, it’s never-ending problem-solving. You are always looking for ways to optimize processes, increase the velocity and optimize the way developers work. It’s also about the efficiency and stability of the production environment.
In a way, being in DevOps allows you to see a bird’s eye view of the entire system. What makes this role very interesting is not being limited to a single domain.
Q: What advice can you give to somebody who is just starting in DevOps?
A: “As an autodidact, I can say that the first thing to know is that you don’t know anything. That’s the baseline and the starting point. And the second thing to realize is that you can solve anything.
Once you know that everything is solvable and that you don’t need to panic when you don’t know how to solve something because you can always gather new knowledge, you can start enjoying the process of problem-solving and optimizing.”
And last but not least, understanding the developers and how they think, and how we can add value by translating the infrastructure to them.”
Q: What do you think makes a good DevOps engineer?
A: “A person with a can-do attitude, a people person, who is always learning, and a problem-solver.
Someone who has that basic understanding that everything is solvable and that we should not take for granted anything at all. Someone who strives to help the developers and understands that we’re here to communicate with people and solve their problems, not just communicate with the computers.”
Q: As a director of DevOps, what are your priorities?
A: “MTTR, MTBF, and LTV, so mean time-to-value, mean time between failures, and mean time to recovery. Those are the measurable KPIs to focus on. And, of course, cost efficiency.”
Q: As a DevOps leader in your organization, what role does security play in the decisions you make?
A: “Very important. I work closely with the security team.
Collaboration is key. As DevOps, we need to create a single language with the security office. Because eventually, we aim for a single goal – for the company to be successful, to grow, and to avoid security incidents, especially public incidents. These will undeniably be very bad for the company and very bad professionally for all involved. We never want to be in this situation.
But also an important part of DevOps is the developer experience. So when we apply security measures and security restrictions on production environments, we still need to maintain the mean time-to-value KPIs.
So when the developers can’t do their work or have to go a much longer road when trying to achieve their goals, we hurt the company, although we increase security.
If the developer cannot view his environment in production, you cannot access these environments in production, doesn’t have any breaking glass protocol, the recent environment in production, then we hurt mean time to value, and mean time to recovery, which eventually will hurt the company. Although our security is great, we will be out of business.
So balancing developer experience with security is something we constantly have to focus on as DevOps.“
Q: As DevOps, what’s the worst ask you get from other teams?
A: “Friday afternoon, a developer decided that he has some spare time to develop. He encounters an issue, and he’s starting to send messages in the DevOps channel.
The channel has two functions; we use it both for standard requests and for urgent requests. So we are always monitoring those channels. And those requests, especially when they’re ambiguous but turn out to be non-urgent, should really wait till Monday morning.”
Q: How can organizations assist their DevOps engineers to be more successful in their jobs?
A: “First is creating the space for and facilitating learning. Most of the DevOps teams are understaffed. And we don’t have time for learning, upskilling, going to meetups, and taking courses. To make the learning part of the job description.”
Q: What would you think is the next big change in DevOps?
A: “I think the two trends to be aware of are serverless and shift-left.
DevOps will require more and more coding skills. We will need to do more and more coding and less infrastructure maintenance. That is why we in DevOps always need to learn and adapt.”
Moshe Belostotsky is the Director of DevOps at Tomorrow.io. With nearly two decades of experience in the field, Moshe is one of the leading minds in the startup nation’s DevOps community. Having worked with companies such as Cisco, Hewlett-Packard, and Fiverr, Moshe has a wealth of experience in the field. In his current role at Tomorrow.io, he is responsible for managing the entire DevOps department and ensuring that the company’s products are released on time and meet customer expectations. In addition to his role at Tomorrow.io, Moshe is the in-demand thought leader in the DevOps community and frequently speaks at industry events.