Run checks on a timer. No persistent process, no daemon overhead.

Scheduled checks use cron or systemd timers to run netglance commands at fixed intervals. Each invocation is a standalone process — it runs, saves results, and exits. No daemon, no background service, no memory footprint between runs.

When to use this mode¶

• You're on Linux and prefer cron or systemd over macOS launchd
• You want lighter-weight monitoring than the full daemon
• You're running on a server or VM where the daemon feels like overkill
• You want fine-grained control over exactly which checks run when
• You're already comfortable with cron and just want to plug netglance in

What it looks like¶

# crontab -e

# Ping check every 5 minutes
*/5 * * * * /usr/local/bin/netglance ping check >> /var/log/netglance/ping.log 2>&1

# Full health report every hour
0 * * * * /usr/local/bin/netglance report --json >> /var/log/netglance/report.log 2>&1

# Device discovery every 6 hours (needs root — use root's crontab)
0 */6 * * * /usr/local/bin/netglance discover >> /var/log/netglance/discover.log 2>&1

# Daily baseline snapshot at 2 AM
0 2 * * * /usr/local/bin/netglance baseline snapshot >> /var/log/netglance/baseline.log 2>&1

# Weekly HTML report on Sunday
0 3 * * 0 /usr/local/bin/netglance report --format html -o /var/log/netglance/weekly.html 2>&1

Setup¶

1. Install netglance¶

uv tool install netglance

2. Create the log directory¶

sudo mkdir -p /var/log/netglance
sudo chown $USER:$USER /var/log/netglance

3. Add entries to crontab¶

crontab -e

Add the checks you want from the examples above. Save and exit.

4. Verify scheduling¶

crontab -l                              # confirm entries
tail -f /var/log/netglance/report.log   # watch for output after the next trigger

Storing results¶

By default, netglance writes results to its SQLite database at ~/.config/netglance/netglance.db. Scheduled checks contribute to the same data store as manual CLI runs, so netglance metrics and netglance baseline diff work on accumulated data from both.

For JSON log files, use --json output and pipe to a file. This is useful for external log aggregation (Loki, Elasticsearch, etc.).

Systemd timers¶

For more control than cron, use systemd timers. They support randomised delays, dependencies, and better logging.

Create a service unit (/etc/systemd/system/netglance-report.service):

[Unit]
Description=netglance health report

[Service]
Type=oneshot
ExecStart=/usr/local/bin/netglance report --json
User=netglance

Create a timer unit (/etc/systemd/system/netglance-report.timer):

[Unit]
Description=Run netglance report hourly

[Timer]
OnCalendar=hourly
RandomizedDelaySec=300
Persistent=true

[Install]
WantedBy=timers.target

Enable:

sudo systemctl enable --now netglance-report.timer
sudo systemctl list-timers | grep netglance

Daemon vs. scheduled checks¶

Both approaches write results to the same SQLite database, so netglance metrics, netglance baseline, and netglance alert work identically regardless of which mode collected the data.

Full reference¶

For crontab recipes, PATH troubleshooting, log rotation setup, and wrapper scripts, see the Scheduling guide.