Linux Systems Administrator

services such as GitLab and CI, and our resources in AWS.

You will keep these systems reliable, secure and documented, plan and

carry out changes to them, and lead the response when something breaks.

This is a hands-on senior individual-contributor role with direct

responsibility for technical execution. We need someone whose knowledge

runs from disks and kernels up to routing tables, firewalls and DNS.

RESPONSIBILITIES

Linux systems administration

• Build, configure, administer and harden Linux systems: installation,

system services, packages, users, permissions, authentication and

access control.

• Operate a fleet of remotely managed physical Linux servers and the

virtual machines and containers on them: storage, capacity planning

and hardware replacement.

• Own the operating-system lifecycle: patch testing, staged deployment,

operating-system and kernel upgrades, tracking exceptions, and moving

systems off releases that reach end of life.

• Keep configurations consistent, and find and correct configuration

drift.

Networks, cloud and development services

• Design, run and troubleshoot the network: routing, traffic shaping,

firewalls, DHCP and DNS, and the overlay network that connects the

data centers, the cloud and remote staff.

• Manage our resources in AWS, mainly EC2 and S3.

• Run the self-hosted development services, including GitLab, CI

runners, build machines and package mirrors.

Troubleshooting and incident ownership

• Independently diagnose production problems involving CPU, memory,

disk I/O, filesystems, systemd, networking and service dependencies,

such as failed services, boot problems, resource exhaustion, package

conflicts and intermittent performance issues.

• Use logs, metrics and command-line tools to find root causes and to

confirm that a fix worked.

• Monitor the systems, and improve monitoring and alerting so that

problems are caught early without needless noise.

• Lead incident response, contribute to incident reviews, and complete

the corrective actions that keep a problem from recurring.

• Work with developers, IT, data-center staff and hardware vendors when

a problem crosses system boundaries.

• Escalate when needed, with clear findings, the supporting evidence,

and a record of the troubleshooting already done.

Automation and configuration management

• Write, debug and maintain Bash scripts for provisioning, diagnostics,

remediation, reporting and routine operations, and Python automation

where it makes operational work more reliable or simpler.

• Write and maintain Ansible roles and playbooks to the team’s

standards.

• Replace repetitive manual tasks with reliable, maintainable

automation.

• Keep configuration and automation in git, test changes, and make

changes through merge requests, review and automated deployment

rather than by editing live systems.

Security and operational reliability

• Own access control and security: SSH keys and hardware security keys,

network access policies, least privilege, security baselines, timely

patching, vulnerability scanning, audits of exposed services, and

software supply-chain risk.

• Run the backups and prove that they restore.

• Plan changes and maintenance with a written procedure that gives the

steps, how to check the result, and a way back.

• Write and maintain documentation, including runbooks, configuration

records and troubleshooting notes, and review other engineers’

changes.

• Take part in a shared on-call rotation and in scheduled maintenance, mdadm RAID, LVM, ext4 and XFS, filesystem repair, SSD caching such as

bcache or dm-cache, and handling failing disks (SMART, NVMe health).

Virtualization and containers

• KVM/QEMU or Xen, LXC and Docker, and enough Kubernetes to debug a

cluster.

Monitoring and backups

• Monitoring and alerting, and backup and recovery, including testing

that restores work.

AWS

• EC2: instances, images, security groups, EBS, instance profiles and

the instance metadata service (IMDSv2).

• S3: bucket policies, versioning, lifecycle rules and encryption.

• IAM roles and policies written to the principle of least privilege.

Git, GitLab, GitOps and configuration management

Experience: 8+ years of relevant hands-on experience, Typically eight or more years administering Linux servers in production

with root responsibility, including physical hardware and not only cloud

instances, with recent hands-on responsibility for operations and

troubleshooting. Solid working knowledge of most of the following.

Linux systems

• Linux processes, services, users and permissions, package

management, filesystems and logging, in depth.

• Strong experience with Debian and other Linux distributions.

• Boot and kernel: UEFI, GRUB, initramfs, kernel configuration and

building, kernel modules (including DKMS) and sysctl.

• Resource control and diagnosis: cgroups v2, memory management and the

out-of-memory killer, pressure-stall information (PSI), and tools such

as strace, perf, iostat, ss and /proc.

• Both systemd and sysvinit.

Networking

• TCP/IP fundamentals: IPv4 and IPv6, subnetting, bridges, VLANs,

bonding and NAT, and troubleshooting connectivity.

• Linux routing: multiple routing tables, policy routing with ip rule,

source-based routing, and routing between sites.

• Traffic shaping with tc: queueing disciplines (HTB, fq_codel, cake),

classification with tc filters or firewall marks, and ingress shaping

through ifb.

• nftables, including sets, maps and connection tracking, and enough

iptables to read and migrate older rule sets.

• DNS and DHCP: unbound as a validating recursive resolver with local

zones and forwarding, dnsmasq for DHCP and local DNS, split-horizon

DNS, and debugging name resolution with dig.

• WireGuard, and NetBird or a similar WireGuard-based overlay network

such as Tailscale or Headscale: peers, groups, access policies,

routing peers for subnets, and DNS. Experience self-hosting the

control plane is a plus.

• Reading packet captures with tcpdump or Wireshark.

Storage

• ZFS in depth: pool layout, datasets and zvols, snapshots, send and

receive, scrubs and resilvering, ARC tuning, and recovering a degraded, Expert git: branching, merging and conflict resolution in a

merge-request workflow, without rewriting or destroying history.

• GitLab administration: a self-hosted installation and its upgrades,

groups, projects and permissions, protected branches, approval rules,

and CI/CD pipelines and runners.

• GitOps and infrastructure as code: configuration that lives in git and

is applied automatically after review, drift detection, and OpenTofu

or Terraform.

• Writing and maintaining Ansible roles and playbooks, or substantial

experience with a comparable configuration-management tool.

Security

• Host hardening, least privilege, firewall design, and finding and

closing services that are exposed by mistake.

• SSH in depth: sshd configuration, SSH certificates, and FIDO2 hardware

keys such as YubiKeys. GPG signing and verification., and commands run over SSH, which you can write, troubleshoot and

maintain in production on your own.

• Good working Python for systems automation and operational tooling.

Judgment and communication

• A record of personally resolving complex technical problems and

following through on the corrective actions.

• Sound judgment about change risk, testing, rollback and escalation.

• Clear, precise written English for change plans, incident reports,

reviews, tickets and documentation.

• The ability to work independently within a remote team.

PREFERRED

• Server hardware: IPMI, iDRAC, iLO and other BMC consoles, RAID and

HBA controllers, and firmware updates.

• Diagnosing PCIe devices and their drivers and firmware from kernel

logs.

• Physical data-center work, and working with data-center remote hands.

• GPU servers, NVIDIA drivers and CUDA environments, and infrastructure

for AI workloads.

• Other virtualization platforms, such as Proxmox or VMware.

• SAN and NAS storage, and multipathing.

• Monitoring and host intrusion detection, such as Prometheus, Grafana,

netdata or Wazuh.

• Jenkins or other build farms, and caching mirrors or proxies for apt,

PyPI, npm or Maven.