Managed hosting — backup engineer for a hosting provider

For a managed hosting provider I work as a structural backup engineer. This party has no servers of their own but mostly supports IT clients in managing and monitoring their server infrastructure, working from the premise that the client should, in principle, never have to worry about their servers — downtime, recovery, disk space, certificates, monitoring, all of it is part of the service.

My role is to know and keep knowing all the varied aspects of that hosting landscape. Not to perform the same rituals daily, but so that the moment extra capacity is needed I can step in without a colleague first having to onboard me into the stack.

Concretely that means I'm comfortable across the chain: orchestration with Ansible, Terraform and Chef, container platforms on Docker and Kubernetes, OS work on CentOS, AlmaLinux, Debian and Ubuntu, observability through Sensu, and the server landscape around Cloudflare, Let's Encrypt and hosters such as Hetzner and Upcloud. On top of that, databases like PostgreSQL, MariaDB and MySQL plus messaging infra such as RabbitMQ.

A good backup engineer is more than someone with the right credentials. It demands people who know where to start looking in a large system, who stay calm during incidents and who document their work cleanly so the regular colleagues can pick it up tomorrow. I find that craft — working inside someone else's landscape without leaving a footprint — one of the more beautiful sides of this work.

At the same time I learn a great deal myself. A broad range of client setups, each with their own preferences and history, forces you to keep thinking instead of leaning on one standard pattern. That diversity feeds straight back into other projects where architecture and operations sit close together.

A role that looks unspectacular at first glance, but is of great importance to the client: the certainty that 'it runs' isn't a coincidence but the result of a team that stands ready and knows what it's doing.