Trucule

I’d use the same baseline approach for logs and compare counts instead of reading lines. Count yesterday’s warnings per unit with journalctl --since yesterday -p warning -o json | jq -r ._SYSTEMD_UNIT | sort | uniq -c, then alert on any unit missing from the baseline or whose count more than doubles. For network traffic, record process and remote address pairs from ss -tnpH state established every few minutes and alert on a pair the baseline doesn’t have.

Drafted with AI.

For monitoring, I’d start with alerts on changes you didn’t make. Save a baseline of listening ports with ss -tlnp and of running containers with docker ps --format ‘{{.Names}} {{.Image}}’, then have a daily cron job compare both against the baseline and push any difference to your phone with ntfy.

Keep the baseline files outside your Docker volumes.

Also send a notification on every successful SSH login with a short script in /etc/ssh/sshrc that posts to the same ntfy topic.

Uptime Kuma covers whether Nextcloud, Immich, and DocuSeal are still reachable through the tunnel, and it can send its alerts to ntfy as well.

Drafted with AI.

2a06:98c0:3600::103 is the address Cloudflare uses as the client IP for requests sent from a Worker, so these probes most likely come from someone else’s Worker and have nothing to do with your DNS-only records. Cloudflare also adds a CF-Worker header to every Worker subrequest, set to the zone name of the account that owns the Worker. I’d log that header in your reverse proxy, then send the zone name and a few sample requests to Cloudflare’s abuse report form. You can also drop any request that carries a CF-Worker header, since your Lemmy instance shouldn’t need traffic from Workers you don’t run.

Drafted with AI.