systemd timers instead of cron

Cron has one enormous advantage: everyone knows it. And one disadvantage: when it fails, you find out by email to root@localhost, which nobody reads.

systemd timers fix exactly that.

The basics

A timer is two files — .service (what runs) and .timer (when).

# /etc/systemd/system/backup.service
[Unit]
Description=Nightly backup
After=network-online.target

[Service]
Type=oneshot
ExecStart=/usr/local/bin/backup.sh

# /etc/systemd/system/backup.timer
[Unit]
Description=Run backup nightly

[Timer]
OnCalendar=*-*-* 02:30:00
RandomizedDelaySec=15m
Persistent=true

[Install]
WantedBy=timers.target

What cron can’t do

Persistent=true — if the machine was off at the scheduled time, the job runs at next boot. Cron just skips it.

RandomizedDelaySec — spreads load when fifty machines would otherwise all fire at 02:30 exactly.

Logs in one place. journalctl -u backup.service shows the last run and the history. No redirecting to files.

Dependencies. After=network-online.target means the job won’t start before the network is up. In cron you solve that with sleep 30 and hope.

Resource limits. MemoryMax=, CPUQuota=, IOWeight= — cron has nothing like it.

Monitoring

This is the main reason we switched. Prometheus node_exporter collects systemd unit state, so a failed job raises an alert like anything else:

node_systemd_unit_state{name="backup.service",state="failed"} == 1

No MAILTO=, no log parsing. A failed job is just a metric.