Kubernetes Storage Engineer (Rook/Ceph)

Senior Engineer to own and harden Rook/Ceph object storage for a Kubernetes-based data platform running Spark and Elasticsearch.

We’re looking for a senior engineer with deep hands-on expertise in Rook (Ceph on Kubernetes) to help stabilize — and potentially redesign — a production object storage stack. This is part-time and long-term opportunity.

Responsibilities

This project is focused on Rook-based architecture and production hardening, including:

• diagnosing root causes of instability and performance collapse\

• \designing a reliable Rook/Ceph architecture for Kubernetes

• improving upgrade safety, operational stability, and performance

• advising whether we should stay on Ceph (via Rook) or migrate awayWe are also open to alternatives (e.g., managed object storage like Wasabi), but the primary goal is to engage someone who can own Rook/Ceph decisions end-to-end.

Environment (current)

• ~10 Kubernetes nodes

• VMs running on Proxmox

Current challenge

We are currently running a self-hosted Ceph-based object storage setup that becomes unstable under:

• heavy read/write traffic (S3 + RADOS writes)

• rebalancing events

• ongoing Ceph upgrades

Additional context:

• No namespacing in the current design

• ~0.5TB stored data

• Current infrastructure costs: ~$15–20k/month (Leaseweb)

Requirements and Skills

• Strong production experience with Rook (Ceph on Kubernetes)

• Deep understanding of Ceph internals (rebalancing, OSD behavior, CRUSH, recovery tuning, etc.)

• Comfortable owning architecture decisions in real production systems

Fluent English (daily communication)


Please provide the position link when you apply in telegram or the position title in the email title. Please if you apply in two places mention it in both messages. Thank you!