VPS monitoring and logging #18

New issue

Open

opened 2026-04-19 07:23:31 +00:00 by forgejo-lila · 0 comments

forgejo-lila commented

2026-04-19 07:23:31 +00:00

Owner

Context

The VPS (Hetzner, Debian 13, ARM64) has no monitoring or centralized logging. If a service goes down or the server is compromised, there's no alert.

What to do

Set up basic monitoring and logging. Two concerns:

Security monitoring: detect unauthorized access or rootkits.
Service monitoring: know when containers are down or unhealthy.

Options to evaluate

Security: chkrootkit or rkhunter for rootkit detection.
Logs: logwatch for daily log summaries via email. Or monit for process monitoring with alerts.
Uptime: a simple external service (e.g. UptimeRobot free tier) pinging https://lilastudy.com and https://api.lilastudy.com/api/v1/health.
Container health: docker compose ps in a cron job, alert if any container is not healthy.

Acceptance criteria

Daily summary of VPS activity is available (email or log file)
Alert mechanism if the app goes down (at minimum: external uptime check)
Security scan runs periodically

Notes

Current VPS setup: SSH key auth, ufw (ports 22, 80, 443, 2222), fail2ban. Docker Compose stack with Caddy, API, web, Postgres, Forgejo, CI runner. See documentation/deployment.md for full details.

## Context The VPS (Hetzner, Debian 13, ARM64) has no monitoring or centralized logging. If a service goes down or the server is compromised, there's no alert. ## What to do Set up basic monitoring and logging. Two concerns: 1. **Security monitoring**: detect unauthorized access or rootkits. 2. **Service monitoring**: know when containers are down or unhealthy. ## Options to evaluate - Security: `chkrootkit` or `rkhunter` for rootkit detection. - Logs: `logwatch` for daily log summaries via email. Or `monit` for process monitoring with alerts. - Uptime: a simple external service (e.g. UptimeRobot free tier) pinging `https://lilastudy.com` and `https://api.lilastudy.com/api/v1/health`. - Container health: `docker compose ps` in a cron job, alert if any container is not healthy. ## Acceptance criteria - Daily summary of VPS activity is available (email or log file) - Alert mechanism if the app goes down (at minimum: external uptime check) - Security scan runs periodically ## Notes Current VPS setup: SSH key auth, ufw (ports 22, 80, 443, 2222), fail2ban. Docker Compose stack with Caddy, API, web, Postgres, Forgejo, CI runner. See `documentation/deployment.md` for full details.