Skip to content

Latest commit

 

History

History
90 lines (63 loc) · 1.91 KB

scheduling.md

File metadata and controls

90 lines (63 loc) · 1.91 KB

Scheduling Backups

Using systemd timers (recommended)

systemd timers are the recommended way to schedule backups on modern Linux systems. They provide better logging, dependency management, and monitoring than cron.

Service unit

Create /etc/systemd/system/backups@.service:

[Unit]
Description=Run backup job %i
After=network-online.target
Wants=network-online.target

[Service]
Type=oneshot
ExecStart=/usr/local/bin/backups /etc/backups/%i.json
User=root
StandardOutput=journal
StandardError=journal
SyslogIdentifier=backups-%i

Timer unit

Create /etc/systemd/system/backups@.timer:

[Unit]
Description=Scheduled backup job %i

[Timer]
OnCalendar=*-*-* 03:00:00
RandomizedDelaySec=300
Persistent=true

[Install]
WantedBy=timers.target

Enable and start

# Enable and start a timer for a specific config file (e.g. /etc/backups/production.json)
systemctl enable --now backups@production.timer

# Check status
systemctl status backups@production.timer
systemctl status backups@production.service

# View logs
journalctl -u backups@production.service -f

The @ (template) unit lets you run multiple backup jobs from one pair of unit files, one per config file.

Per-job timers

If you need different schedules per job, create separate timer overrides:

systemctl edit backups@offsite.timer
[Timer]
OnCalendar=
OnCalendar=*-*-* 01:00:00

Using cron

Add an entry to /etc/cron.d/backups:

# Run backups daily at 3am
0 3 * * * root /usr/local/bin/backups /etc/backups/production.json >> /var/log/backups.log 2>&1

Or using crontab -e for the root user:

0 3 * * * /usr/local/bin/backups /etc/backups/production.json

Monitoring

Pair the backup job with the flag file notification for simple freshness monitoring, or the Prometheus notification for metrics-based alerting.