Linux Admin
systemd Timers for Maintenance
Replace mystery cron lines with timers you can explain, monitor, and silence when they misbehave.
Hybrid studio 2 weeks · 12 lab hours Associate
480,000 KRW informational until enrollment confirms
Program narrative
We compare cron vs timers with real maintenance windows, daylight savings edge cases, and jitter for thundering herds. Each learner ships two timers: one human-friendly, one machine-friendly.
What the labs include
- OnCalendar grammar with daylight edge cases
- Randomized delays for fleet jobs
- Unit coupling between service and timer
- Failure notifications without email storms
- Temporary suspend patterns during freeze windows
- Visualization with systemd-analyze plot
- Peer critique of naming conventions
Outcomes you can show a lead
- Convert a fragile cron line into a timer pair
- Explain jitter choices to application owners
- Produce a calendar others can trust
Straight FAQ
Is legacy cron migration covered?
Yes, with a staged checklist; we do not automate mass migration blindly.
Monitoring integrations?
We export timer metadata to Prometheus textfile example only.
Limitations?
Container-only workloads are out of scope for this cohort.
Recent participant notes
“Randomized delay lab stopped our patch reboot stampedes. Still wish we had 30 more minutes on calendar DST.”
Taeyang