Restarting a service before fixing the underlying config error puts it in a crash loop. Read journalctl output first. After editing unit files, always daemon-reload before restart.
What you will achieve
Recover a failed systemd service on Ubuntu or Debian — diagnose, fix configuration, reload units, restart, enable at boot, and verify it stays running.
1) Check current state
systemctl status nginx.service
systemctl is-active nginx
systemctl is-enabled nginxfailed or inactive (dead) with non-zero exit code needs investigation before blind restart.
2) Read recent logs
sudo journalctl -u nginx.service -b --no-pager -n 50
sudo journalctl -u nginx.service -p err --since "1 hour ago"Config syntax errors, permission denied, and port-in-use messages appear here — fix the cause, not the symptom.
3) Validate config before restart
sudo nginx -t
sudo sshd -t
sudo apachectl configtestService-specific test commands vary. Running restart on broken config wastes time and fills logs.
4) Reload systemd after unit file edits
sudo systemctl daemon-reloadRequired when you change /etc/systemd/system/myservice.service or drop-ins under .d/ directories. Skip this and systemd runs the old unit definition.
5) Restart and enable
sudo systemctl restart nginx.service
sudo systemctl enable nginx.servicerestart stops then starts. Use reload when the service supports graceful config reload without dropping connections:
sudo systemctl reload nginx.service6) Reset failed state
sudo systemctl reset-failed nginx.serviceClears the red failed badge after you fixed the issue — cosmetic but helps monitoring tools.
7) If restart keeps failing
systemctl cat nginx.service
sudo systemd-analyze verify nginx.serviceCheck ExecStart path, User= permissions, and After= ordering. Dependency not ready at boot may need Wants= or correct target.
8) Start vs restart
sudo systemctl start nginx.service # first start if never run
sudo systemctl try-restart nginx # restart only if already activetry-restart avoids error noise in automation when the unit was intentionally stopped. Pair with OnFailure= units for alert hooks on production services.
Prerequisites
Root or sudo access, correct unit name (use tab completion with systemctl restart nginx<Tab>), and config validated with the service's own test command before restart.
Verify
systemctl status nginx.service
systemctl is-active nginx
curl -I http://localhost/Active (running) with clean recent log lines. Reboot test on critical services: sudo reboot, then confirm auto-start.