Run netglance checks automatically on a schedule without the overhead of the full daemon. Use cron, systemd timers, or a lightweight wrapper script to fit your infrastructure.

Why Schedule Instead of Using Daemon Mode¶

The daemon mode (netglance daemon install) is convenient on macOS, but scheduling checks directly offers several advantages:

• Lightweight: No persistent process consuming memory or system resources
• Flexible: Run different checks at different intervals — a ping every 5 minutes, a full report every hour, device discovery weekly
• Portable: Works on any Unix system with cron or systemd, not just macOS
• Observable: Each run is independent; logs are straightforward to inspect and debug
• Composable: Combine multiple checks in a shell script with custom error handling

Crontab Scheduling¶

Edit your crontab with crontab -e. The format is minute hour day month weekday command. All times are in your local timezone.

Quick Reference¶

# Every 5 minutes: quick ping check
*/5 * * * * /usr/local/bin/netglance ping check >> /var/log/netglance-ping.log 2>&1

# Every hour: full health report
0 * * * * /usr/local/bin/netglance report >> /var/log/netglance-report.log 2>&1

# Every 6 hours: device discovery + baseline diff
0 */6 * * * /usr/local/bin/netglance discover scan && /usr/local/bin/netglance baseline diff >> /var/log/netglance-discover.log 2>&1

# Daily at 2:00 AM: comprehensive scan + HTML report
0 2 * * * /usr/local/bin/netglance report --format html -o /tmp/netglance-daily-$(date +\%Y\%m\%d).html >> /var/log/netglance-daily.log 2>&1

# Weekly on Sunday at 3:00 AM: full baseline snapshot
0 3 * * 0 /usr/local/bin/netglance baseline snapshot >> /var/log/netglance-baseline.log 2>&1

PATH and Binary Location¶

Cron runs with a minimal PATH and no shell initialization. Either:

1. Use the full path to netglance (if installed system-wide):

   0 * * * * /usr/local/bin/netglance report

2. Or add PATH to the top of your crontab:

   PATH=/usr/local/bin:/usr/bin:/bin
   0 * * * * netglance report

3. If using uv run for development:

   0 * * * * cd /path/to/netglance && /usr/bin/uv run netglance report

Log Rotation¶

Redirect output to a log file with >> /var/log/netglance.log 2>&1. On macOS, use /var/log/ (requires sudo to edit crontab) or ~/Library/Logs/. On Linux, logrotate handles rotation automatically. For manual rotation:

# On macOS or Linux
0 0 * * 0 gzip /var/log/netglance.log && mv /var/log/netglance.log.gz /var/log/netglance.log.$(date +\%Y\%m\%d).gz
0 0 * * * /usr/local/bin/netglance report > /var/log/netglance.log 2>&1

Systemd Timer Alternative (Linux)¶

Systemd timers offer better integration with Linux services, automatic logging via journald, and dependency management.

Service File¶

Create /etc/systemd/system/netglance-healthcheck.service:

[Unit]
Description=netglance health check
After=network-online.target
Wants=network-online.target

[Service]
Type=oneshot
ExecStart=/usr/local/bin/netglance report
StandardOutput=journal
StandardError=journal
User=root

Timer File¶

Create /etc/systemd/system/netglance-healthcheck.timer:

[Unit]
Description=Run netglance health check hourly
Requires=netglance-healthcheck.service

[Timer]
OnBootSec=2min
OnUnitActiveSec=1h
Persistent=true

[Install]
WantedBy=timers.target

Enable and Verify¶

sudo systemctl daemon-reload
sudo systemctl enable --now netglance-healthcheck.timer
sudo systemctl list-timers netglance-healthcheck.timer
sudo journalctl -u netglance-healthcheck.service -f  # Watch logs

Preventing Overlapping Runs¶

If a check takes longer than your schedule interval, prevent concurrent runs with flock:

# In crontab or systemd service
0 * * * * /usr/bin/flock -n /tmp/netglance.lock /usr/local/bin/netglance report >> /var/log/netglance.log 2>&1

The -n flag makes flock non-blocking — if another instance is running, the new one exits silently instead of waiting.

Timeout Protection¶

Protect against hung or runaway checks:

# Kill the check if it runs longer than 5 minutes (300 seconds)
0 * * * * timeout 300 /usr/local/bin/netglance report >> /var/log/netglance.log 2>&1

Wrapper Script for Multiple Checks¶

For more complex scheduling, create a wrapper script at /usr/local/bin/netglance-scheduled.sh:

#!/bin/bash
set -e

LOG="/var/log/netglance-scheduled.log"
ERROR_LOG="/var/log/netglance-errors.log"

log() {
    echo "[$(date '+%Y-%m-%d %H:%M:%S')] $1" >> "$LOG"
}

on_error() {
    log "ERROR: Check failed with exit code $?"
    echo "netglance check failed at $(date)" >> "$ERROR_LOG"
    exit 1
}

trap on_error ERR

log "Starting scheduled checks"

# Quick connectivity check
timeout 60 /usr/local/bin/netglance ping check >> "$LOG" 2>&1
log "Ping check completed"

# DNS health check
timeout 60 /usr/local/bin/netglance dns check >> "$LOG" 2>&1
log "DNS check completed"

# Device discovery
timeout 120 /usr/local/bin/netglance discover scan >> "$LOG" 2>&1
log "Discovery completed"

# Generate HTML report
REPORT="/tmp/netglance-report-$(date +%Y%m%d-%H%M%S).html"
timeout 300 /usr/local/bin/netglance report --format html -o "$REPORT" >> "$LOG" 2>&1
log "Report saved to $REPORT"

log "All checks completed successfully"

Make it executable:

chmod +x /usr/local/bin/netglance-scheduled.sh

Then schedule it in crontab:

0 * * * * /usr/local/bin/netglance-scheduled.sh

Email Notifications on Failure¶

Install cronic (simple bash wrapper) to email only on failure:

# On macOS
brew install cronic

# On Linux (Debian/Ubuntu)
sudo apt install cronic

# In crontab
0 * * * * cronic /usr/local/bin/netglance report
# Emails output only if non-zero exit code

Or use a simple bash wrapper:

RESULT=$(/usr/local/bin/netglance report 2>&1)
STATUS=$?
if [ $STATUS -ne 0 ]; then
    echo "$RESULT" | mail -s "netglance check failed" admin@example.com
fi
exit $STATUS

Comparison: Cron vs Daemon vs Systemd Timer¶

Choose cron for portability and simplicity. Choose systemd timers on Linux for better integration and logging. Use daemon mode on macOS if you want a persistent process with native system integration.