Senior Linux System Engineer – Proxmox, Ceph & PostgreSQL (m/f/d)

95 % Remote · Full time · unlimited

IT has been part of the business model at HAPEKO since its foundation – and we are building it new: our own rack in the data center, Proxmox-VE-Cluster with Ceph on HPE hardware, PostgreSQL with Patroni, two GPU nodes for own LLM-Inference. For this we build our own platform team. The hardware stands, the clusters run. What is missing is the operation that carries it permanently – automation, standards, monitoring. Whoever builds that decides how this landscape looks in two years. That's you. If you understand operation as an engineering task – automated, measured, documented – and PostgreSQL is an operating theme for you and not a query tool, then this is your place.

This makes the position interesting You are the first person in the team with this focus and shape how the stack is operated – which standards apply, how automated what must be documented. You report directly to the IT management and work closely with it: architecture and standards decide together, implementation and operation are your responsibility. The team is six heads tall, decisions fall in days, not in quarters. Hands-on mentality is explicitly required: You are an architect and operator at the same time. With the growing landscape, your role is growing – towards Kubernetes, Net, Security or GPU, depending on where you want.

You operate and develop our platform: Proxmox VE, Ceph, Proxmox Backup Server – on your own HPE hardware in your own rack, not in a strange cloud. You are responsible for PostgreSQL as a platform service: high availability with Patroni, PgBouncer, backup and point-in-time recovery with pgBackRest, extensions such as pgvector, performance at the instance level and the access model. You bring our data platform (PostgreSQL, Airbyte, dbt) to operating standards: inventory, runbooks, restore tests, monitoring. You automate with Ansible and Git. What you did twice by hand is the third time code. You are 3rd level: What our support does not solve is analysing over application, database, network and infrastructure – and make sure it doesn't happen again. You work hand in hand with our support: You announce changes before You arrive with the users and explain what You have to adjust. What triggers your change in the background on the surface is your ticket – not his. You document that the next can take over – Runbooks instead of head knowledge. What awaits you in the first year: You take a freshly built Proxmox cluster with six nodes, two new backup servers, our new VPN (NetBird) and two GPU reference nodes. Everything is running – but not yet as it has to run in two years. You bring it there: Ansible instead of manual work, monitoring with alarms someone reads, restore tests with protocol, runbooks that work without you. At the same time, you build what's new in 2027: the data platform in regular operation and the platform for new applications – PostgreSQL, container operation and an API gateway that grows with them.

Multiannual practice in operating Linux platforms – you have built clusters, not only created VMs. Proxmox VE or KVM/libvirt in production; Ceph or a comparable distributed storage, ideally with own hand on the OSD. PostgreSQL in production: replication, backup and recovery, tuning. SQL is self-evident, but not your focus. Ansible, Git, Shell or Python. Network bases: VLAN, Bonding, Routing. You write what you do – so that someone else understands without you. Whether training, study or cross-entry, we don't care. We'll look at what you did.

The pump – no need, but a plus Kubernetes (with us RKE2): If you can, there are clusters for you. Network beyond the basics: routing, firewalling, high availability. GPU operation: vLLM, VFIO passthrough, MIG. Experience from hosting, data center, university or research – environments where hardware is still touched.

95% remote work: You work almost completely mobile. Every two weeks a day of presence in the branch in your area – we have 25 of them. 28 + 2 days holiday Trust work time: You shape your working time flexibly and independently. Ribbon: € 75,000–85,000 depending on the experience. Continuing education: We support training and certifications, e.g. for Proxmox, Ceph, PostgreSQL, CKA or others. Own hardware, own rack: You work on systems that belong to HAPEKO – with root on everything. And if you prefer to touch server than just SSH: If you live around Hamburg, you will get your own data center card – all the others will get iLO and one day before the rack. Short paths: You get the freedom to introduce your own ideas and actively shape technical solutions. .