**Motivations:** - Alert when the LPLDF storefront HTTPS endpoint is down from the proxy **Root causes:** - N/A (monitoring gap) **Correctifs:** - N/A **Evolutions:** - watch-https-lpldf.sh: curl check, state file, syslog tag lpldf-https-watch, optional webhook/email via env file - systemd oneshot + 5-minute timer; install script via SSH/scp **Pages affectées:** - tools/proxy-https-watch-lpldf.sh - tools/proxy-https-watch-lpldf.env.example - deploy/proxy-units/lpldf-https-watch.service - deploy/proxy-units/lpldf-https-watch.timer - deploy/scripts/install-lpldf-https-watch-on-proxy.sh - deploy/README-lpldf-https-watch.md
1.4 KiB
1.4 KiB
LPLDF HTTPS watchdog on the 4NK proxy
Purpose
Detect downtime of https://xn--lespetitesleonsdefrdric-89b1db.fr/ from the proxy and emit alerts (syslog tag lpldf-https-watch, optional webhooks / mail). Acts as an availability watchdog; a SIEM (e.g. Wazuh) can ingest these syslog lines.
Repository paths
- Watch script (installed to
/opt/proxy-config/scripts/watch-https-lpldf.sh):tools/proxy-https-watch-lpldf.sh - Optional env example:
tools/proxy-https-watch-lpldf.env.example - Systemd units:
deploy/proxy-units/lpldf-https-watch.service,deploy/proxy-units/lpldf-https-watch.timer - Installer (from ia_dev root):
./deploy/scripts/install-lpldf-https-watch-on-proxy.sh
Behaviour
- Accepts HTTP status 200, 301, 302, 307, 308.
- State under
/var/lib/lpldf-https-watch/. - First DOWN:
daemon.warning+ optionalALERT_WEBHOOK_URL/ALERT_EMAIL_TO. - Repeats while down at most every
ALERT_REPEAT_SECONDS(default 3600). - Recovery:
daemon.info+ optionalALERT_WEBHOOK_URL_RECOVER.
Optional proxy config
Create /opt/proxy-config/scripts/env/watch-https-lpldf.env (e.g. chmod 600), see tools/proxy-https-watch-lpldf.env.example.
Operations
- Manual run on proxy:
sudo /opt/proxy-config/scripts/watch-https-lpldf.sh - Logs:
sudo journalctl -t lpldf-https-watch - Timer:
systemctl status lpldf-https-watch.timer
Nginx is not modified for this check.