◆ kubedock / infrastructure consulting
Independent consulting

Containerizing the systems
nobody else wants to touch.

Legacy app dockerization, stateful workloads, and Kubernetes ingress — for EU teams modernizing infrastructure that has been running since well before microservices were fashionable.

Services
01 /
Legacy app modernization
Dockerizing monoliths with native dependencies, hardcoded paths, ancient JVMs, and license servers. Then shipping them on Kubernetes — on-prem or any major cloud (AWS EKS, Azure AKS, Google GKE) — with sane build, deploy, and rollback workflows.
02 /
Stateful workloads on Kubernetes
Clustered databases in containers — PostgreSQL with pgpool, MariaDB Galera, DB2 in HA configurations. Custom container images for systems the official charts do not handle. Persistent storage, backup, and restore that actually works under load.
03 /
Kubernetes ingress, mTLS & RBAC
Traefik IngressRoute, gRPC over h2c, ServersTransport, certificate-based user authentication, RBAC group bindings. The kind of Kubernetes networking and access-control issues that burn an afternoon when you do not work with them weekly.
Selected work
Stateful workloads
Custom DB2 image running as a cluster in Docker
Built and operated a DB2 container image with clustering support where the official image stopped short. Used as a foundation for a production data layer.
Read the full technical writeup
Stateful workloads
PostgreSQL + pgpool clustered Docker image
Custom Postgres + pgpool image surviving node failures, with read-replica routing and predictable failover — deployable on plain Docker or Kubernetes.
Read the full technical writeup
Platform tooling
Mitmproxy sidecar for live traffic inspection on Kubernetes
A reusable sidecar that attaches to any pod and exposes its HTTP traffic through a browser-based GUI — full request and response inspection without redeploying or instrumenting the target service. Required solving a Traefik HTTP/2 incompatibility along the way.
Writeup coming soon
About
Constantin Manea
Constantin Manea
Senior DevOps engineer · Bucharest, Romania

Ten-plus years building and operating Linux infrastructure, with the last several focused on Kubernetes, Docker, Traefik, and the awkward edges where legacy systems meet modern platforms.

I work with a small number of clients at a time on focused, scoped engagements — typically 2–6 weeks, with a clear deliverable and a written handover at the end.

Engagements
Have a system that needs containerizing, or an ingress issue eating your week?

Email a short description — the stack, the constraint, what “done” looks like. I reply within two working days with whether it is a fit and a rough scope.