This is the third post in our blog series about the Weaveworks CRE team. As experts in both cloud native technologies and the GitOps methodology, they help our customers deliver solid Kubernetes solutions and platforms.
Sebastian Bernheim, Customer Reliability Engineer
Sebastian is a code slinger and system wrangler with an abiding passion for learning and sharing technology. After one score and three years spent building, breaking and fixing software, his current mission is to help customers succeed by fostering developer innovation. Sebastian is also a three-peat progenitor, maker of omelets, and an accomplished dog wrestler. He lives in Jersey City, New Jersey.
What does a typical day look like for you and what are you currently working on?
My dog usually wakes me up when he’s ready for his morning walk. Walking Winston helps start the blood flow to my brain. He spends the rest of the morning napping on the couch or at my feet under my desk.
I eat breakfast at my desk and read the news or listen to technology podcasts.
I try to get meetings and administrative tasks done in the mornings, so I can spend my afternoons coding or digging into more complex issues. I’ll catch up on Slack and email or write client reports before lunch, take a break around 1 or 2 PM, and then go for a deep-dive on a project until I need to stop for the day.
I’m currently working with a large financial institution to create tools that help internal application developers deploy their projects to Kubernetes clusters. The client’s platform team is using a mix of open-source and proprietary projects to automate cluster setup, application deployment, and many of their internal processes. Weaveworks helps them make the best possible use of the projects and products already out there in the Cloud Native ecosystem as well as doing some bespoke coding.
Most recently, I’ve been writing a Controller to automate granting temporary exec access to application Pods in Kubernetes. Security is always a high priority for large Enterprise firms, especially in the Financial sector, and they tend to want to lock things down as much as possible. In this case, that means their developers and operators can’t connect to the containers running their applications in production and other higher environments. The new automation would kick off at the end of their existing exception approval process to give them the access permissions they need with an enforced time-limit.
Any words of advice for others trying to dive into cloud native and learn, build, implement Kubernetes platform?
The most important advice I can give is, START NOW! Whatever your role is and whatever project you’re working on today, you’ll benefit from containers and Cloud Native technology almost immediately. There is a lot of low-hanging fruit.
Tools like Docker for Desktop, Kind, and Firekube make getting started with a local Kubernetes cluster much easier than it used to be, and the Fork, Clone, Run model can help you create a running GitOps setup from scratch with very little effort. From there you can try packaging your favorite application in a container, or building a container image for your project using Docker, buildkit or Kaniko. Then deploy it to your cluster and start poking at it!
AWS, Azure and GCP all offer free trials you can use to learn about their services and try hosted Kubernetes. Cloud providers want to make it as easy as possible to try their managed Kubernetes services - like EKS, AKS and GKE. They all publish web-based tutorials to help you get the ball rolling.
The Cloud Native ecosystem is so big and so varied that it can be overwhelming, but there’s no reason to be intimidated. We’re all learning this together. No matter how long you’ve been working in Cloud Native there are so many new projects and the tech moves so fast in so many directions that nobody can keep up with it all. Everyone is always a newbie at something, and people in the community tend to be friendly and enthusiastically helpful. Ask questions, and someone inevitably steps up to help.
What’s your top 3 Talks/Books?
The Hitchhiker’s Guide to the Galaxy by Douglas Adams is hands-down my favorite book of all time. Other than that, I could never narrow it down to just three. I’m a fan of Kurt Vonnegut, William Kotzwinkle, T. C. Boyle, Terry Pratchett, Neil Gaiman, Neal Stephenson, and a long list of others. All my favorites books are fiction.
I think technology books are historical artifacts now. Technology moves too fast for the paper publishing cycle, and all those guides and manuals and O’Reilly grimoires we used to carry around with us years ago are obsolete today. Reference materials are maintained on the web, and the best tutorials tend to be written by our peers and fellow enthusiasts. The old barriers to entry in technical publishing have all been dismantled.
There are so many great talks and fantastic speakers out there right now. Check out Kris Nova if you haven’t seen her present live. Her FOSDEM talk from February this year is up on YouTube. Don’t forget the Weaveworks YouTube channel! Maybe this is a shameless plug, but I like watching my Weaveworks colleagues share their expertise. When they’re enthusiastic about the subject it really shows, and it’s a joy to behold.
What do you wish other people knew about Weaveworks?
The people here are amazing. My colleagues really do inspire me, and I’m constantly learning from them. It’s a pleasure to work alongside so many people with deep knowledge and true passion for what they do. I feel lucky everyday that I get to share this experience with them.
What’s your favourite cool new technology?
Firecracker, Ignite and Firekube!
Does Kubernetes still count? As an orchestrator it’s mostly commoditized now, so maybe it’s not what people notice anymore but I still think it’s pretty cool.
I also think Serverless engines are interesting, but I haven’t had time to play with them lately so everything I once knew is probably already obsolete. That’s part of the fun!