mirror of
https://github.com/RWejlgaard/pez-infra.git
synced 2026-05-06 04:14:43 +00:00
Cloudflared tunnels are no longer used. All traffic now routes through Cloudflare DNS to Caddy on helsinki-a over Tailscale. - Remove cloudflared systemd unit files (copenhagen-a, london-b) - Remove cloudflared from media_stack role and copenhagen-a host_vars - Remove cloudflared references from services README and host docs - Remove cloudflared deploy trigger from CI workflow Live service on london-b stopped and disabled. copenhagen-a was unreachable but the tunnel is unused regardless.
1.9 KiB
1.9 KiB
london-a
Dedicated monitoring server. Runs Prometheus and Grafana, nothing else.
Overview
| Location | London (NW9) |
| OS | FreeBSD 14.3 |
| Tailscale IP | 100.122.219.41 |
| Role | Monitoring (Prometheus + Grafana) |
Hardware
| Component | Spec |
|---|---|
| CPU | Intel i7-4790K (8 threads) |
| Memory | 32 GB |
| Boot disk | 1 TB |
Old gaming PC, now perfectly happy as a monitoring host. Very lightly loaded — disk at ~6%.
Services
| Service | Port | Status | Notes |
|---|---|---|---|
| Prometheus | 9090 | Active | prometheus.pez.sh |
| Grafana | 3000 | Active | grafana.pez.sh |
| node_exporter | 9100 | Active | Metrics exporter |
| Tailscale | — | Active | Mesh networking |
Both Prometheus and Grafana are behind Authelia (auth handled by Caddy on helsinki-a).
Unused services (audit 2026-03-30)
These services are enabled in rc.conf but appear unused. Pending cleanup.
| Service | Port | Finding |
|---|---|---|
| InfluxDB | 8086 (all interfaces!) | Only _internal database — never used. Listening on *:8086 is also a security concern. |
| Redis | 6379 (localhost) | Empty keyspace, no clients. |
| PostgreSQL | 5432 (localhost) | Has pez_vps database from a defunct VPS management project. Data may need backup before removal. |
| libvirtd | — | Zero VMs. Installed for the same pez_vps project. |
ZFS
- Pool:
zroot - Weekly scrub:
0 12 * * sun zpool scrub zroot(root crontab, not ansible-managed yet)
Why FreeBSD
This one runs FreeBSD instead of Ubuntu. For a single-purpose monitoring host it works well. No particular reason to change it — it's stable and does its job.
Networking
Connected via Cat 5 to the Ubiquiti switch alongside london-b.
Notes
Prometheus scrapes all hosts over Tailscale. See monitoring.md for scrape targets and dashboard details.