MetalBear

With mirrord for CI, you can run tests against real services, real traffic, and real data without spinning up extra infrastructure. Learn more here 👇 https://lnkd.in/eD7GCTEP

We just launched mirrord for CI which lets you run code inside your CI runner while connecting it to an existing Kubernetes environment. This way you get fast feedback, realistic tests, and no extra infrastructure to manage. Learn how you can test against real services, real data, and real traffic, while easily saving 20–30 minutes per CI run 👇 https://lnkd.in/eWjMSqKA

Traditional CI pipelines slow developers down because every run involves building container images, deploying them, and often spinning up ephemeral environments that are costly and slow to become ready. mirrord for CI changes this by running your changed microservice directly inside the CI runner while connecting it to an existing, real Kubernetes environment, like staging or pre-prod. Incoming and outgoing traffic, environment variables, and files are all proxied, so your tests behave as if they’re running in the cloud, without building images or deploying anything. Read the full announcement here: https://lnkd.in/gP9W7GF3

Friends in London if you're attending #ContainerDays come stop by the MetalBear booth!

AI tools are everywhere right now, and our engineering team uses them daily. In this post, we're sharing how we actually use AI coding tools and agents while building mirrord, what's been useful and what hasn't. Read here: https://lnkd.in/gss4A-Xa #ai #engineering

#Kubernetes 1.35 came with a lot of new features and improvements. Here are some of our favorites! If you’d like to learn more about them, check out our detailed blog: https://lnkd.in/gS-pVDC7

If you’re using AI coding tools today, this probably sounds familiar: ❌ Code looks correct ❌ Unit tests pass ❌ CI is green …and then it breaks in staging. Why? AI-generated code often lacks full awareness of: Microservice dependencies Runtime behavior Real cluster configuration In Kubernetes environments, these issues surface late and painfully. In our upcoming hands-on workshop, we’ll show how to test AI-generated code against real services in your staging cluster without deploying and without risk. 👉 Read more and join us live: https://lnkd.in/dGNw_JCV Anton Weiss Arsh Sharma MetalBear PerfectScale by DoiT DoiT #k8s #devops #sre #kubernetes #ai #genai

Happy New Year! We wrote down our top predictions for the cloud ecosystem in 2026 based on conversations with project maintainers, CNCF ambassadors, and experiences at KubeCon. Read here: https://lnkd.in/gpsGHvt4

"Developer velocity is critical, and debugging in real-world environments is still messy and slow. MetalBear cuts that pain point sharply, improving feedback loops and productivity in a way engineers immediately feel." Thanks for the mention Ophir Shay!

MetalBear co-founder and CEO Aviram Hassan recently sat down with Authority Magazine for an interview, sharing his experience of starting a company, lessons learned along the way, and debunking myths about what it means to be a CEO. Read the full piece here for insights: https://lnkd.in/g-9P37-u