The BackupWall console will not open, or a service is down
Check that all six backupwall services are active, that the health check answers, that port 8087 is open in the server and cloud firewalls, and read the API and worker logs. A watchdog also checks the services every 5 minutes.
1. Check the services
systemctl list-units 'backupwall*' --no-legend --no-pager
You should see six units, all active running: backupwall-api, backupwall-worker, backupwall-download-worker, backupwall-beat, backupwall-watcher and backupwall-redis. On DirectAdmin, after you install the plugin, you also see backupwall-download-proxy.timer; that is expected. systemctl status shows the worker and download worker as active even when they hold no task process, which is normal.
Start a stopped one with systemctl start <unit>.
2. Check the API
curl -s http://127.0.0.1:8087/api/v1/health/live
curl -sk https://127.0.0.2:8087/api/v1/health/live
curl -s http://127.0.0.1:8087/api/v1/health/ready
The first two should print {"status":"ok", ...}. The first is the plain-HTTP listener (reachable only from the server); the second reaches the HTTPS listener. The third also checks the database and shows the version, build and dr_configured.
3. Check the address and firewall
- The console is
https://<server-ip>:8087. On a cloud server use the public IP and open 8087 in the cloud firewall (VPC rules) as well as the server's own. The installer prints the private address if it could not read the public one. - The browser warns about the certificate: that is expected (it is self-signed). Accept it once. Opening
http://<server>:8087/redirects tohttps://. - If you use CSF, add
8087toTCP_INyourself; the installer only prints a reminder. The installer opens the port only in firewalld or an active ufw. - On a server with no panel the installer does not open 8087; it serves the console on port 80 (plain HTTP) unless you installed with a
DOMAIN. See Servers with no control panel.
4. Read the logs
journalctl -u backupwall-api -n 50 --no-pager
tail -n 100 /var/log/backupwall/worker.log
systemctl status backupwall-redis
df -h /var/lib/backupwall /opt/backupwall
The API and the watcher log to the system journal (journalctl -u backupwall-watcher). The worker, the download worker and the scheduler write to files in /var/log/backupwall/ (worker.log, download-worker.log, beat.log, redis.log, watchdog.log). Logs rotate daily and 14 days are kept.
If the private queue is down
While backupwall-redis is down, BackupWall stays up but cannot queue work: Run now shows a failed row. When it is back, new tasks are picked up again. On builds from v267, stopping, restarting or crashing it does not restart the other services or hold back the worker (installs made before that release keep older service units, which stop the API, workers and scheduler and kill the running job).
A full disk
A full /var/lib/backupwall (working space) or /opt/backupwall stops backups, restores and downloads. Free space, then restart the failed services.
Still stuck
Run backupwall-support-bundle as root and follow Report a problem and make a support bundle.