Symptoms

The platform-deployment script used bare echo statements for all output, meaning log messages were not persisted to a dedicated log file and were not captured in the systemd journal. This made it difficult to troubleshoot platform service deployment issues (e.g., container startup failures, registry setup, OS rebase errors) after the fact, as the output was only visible in the service status and could be lost on rotation or reboot.

Impact

Reduced observability during platform service deployment. Operators had limited ability to diagnose historical deployment failures, Docker Compose errors, or registry setup issues because log messages were transient and not written to a persistent, searchable location.

Conditions

Any F5OS-A or F5OS-C system running the platform-services-deployment service during upgrade/downgrade, reboot, or service start/stop operations.

Workaround

Prior to this fix, operators could manually inspect systemctl status platform-services-deployment.service -l for recent output, but historical logs were not reliably available.

Fix Information

Added structured logging to the platform-deployment script. All output is now timestamped and written to a dedicated log file (/var/log/platform-deployment.log) with logrotate support, and simultaneously forwarded to the systemd journal via StandardOutput=journal on the service unit. This enables operators to use both journalctl -u platform-services-deployment and the persistent log file for troubleshooting platform service deployment issues

Guides & references

K10134038: F5 Bug Tracker Filter Names and Tips