I recently migrated a production server from CentOS 7 to AlmaLinux 9, and I’m running into an issue where the network doesn’t fully come up during boot. The system is using a static IP configuration managed via NetworkManager, and while everything works normally after the system is up, the interface stays unconfigured until I manually run nmcli connection up <profile>.
For context, I’ve been reviewing and documenting the network configuration and boot sequence notes using a small internal setup where I keep configs and troubleshooting logs in a structured editor (I’ve been using a CodeMirror-synced with windows video edit for this, mainly to keep YAML and network profiles consistent and easier to track changes across revisions). In the logs, systemctl status NetworkManager shows it running, but ip a doesn’t reflect the expected interface state until manual intervention. journalctl -b shows a few early boot timing warnings, but nothing clearly points to a misconfiguration.
Has anyone seen similar behavior on AlmaLinux 9 after migrating from CentOS 7? I’m trying to figure out whether this is a systemd ordering issue, a NetworkManager dependency problem, or something subtle missed during migration. Any suggestions on where to dig deeper in the boot process would be helpful.