SRE / DevOps

We're looking for an SRE / DevOps engineer to build the reliability layer for a platform that commands live power plants. No unified log pipeline, no dashboard anyone watches, no rehearsed restore - the app runs in production, the layer underneath it is a set of habits. You decide what the observability stack looks like, what an alert may wake a person up for, and how failover is designed and tested.
Second part is physical: with the Infra Lead, dispatch control room - workstations, UPS, local network, secured remote access to plants. Grafana in one tab, a rack in the other.

Must have
4+ years SRE / DevOps / systems engineering, Linux daily: services, systemd, storage, performance diagnostics
Real experience with physical or VPS servers, not cloud consoles only - UPS, RAID, remote management. Cloud-only profile is an automatic no
Cloud VMs: provisioning, patching, hardening, capacity planning, TLS, nginx
Docker / Docker Compose in production: dev/staging/prod layouts, registry, zero-downtime deploys
Observability you built yourself: Prometheus + Grafana or equivalent, log pipeline (fluent-bit / ELK class), and an alerting philosophy you can defend
Redundancy you tested: PostgreSQL backup/restore and replication, Redis failover, measured recovery. You can describe a backup you personally restored
Networking: WireGuard or similar, firewalls, DNS, TLS, diagnosing cloud ↔ office ↔ plant links
CI/CD as code (GitHub Actions or similar), promotion through environments
Bash + Python for automation
English B2