While performing a firmware update on a rack-mounted application server located in a remote colocation facility, an administrator's SSH session hangs and the production NIC stops responding to pings. The server is still connected to AC power, and the administrator can reach the host's dedicated management interface on a separate management VLAN. Which action will allow the administrator to restore the host to service without waiting for on-site staff to intervene?
Re-establish an SSH session and run the shutdown -r now command with a five-minute timer.
Send a Wake-on-LAN magic packet to the server's primary Ethernet port.
Log in to the server's baseboard management controller and issue a chassis power cycle.
Use the hypervisor management console to reboot all guest virtual machines on the host.
The server's baseboard management controller (BMC)-accessed through technologies such as IPMI, iLO, iDRAC, or CIMC-operates on its own embedded processor and NIC. Because it is powered any time the chassis has line voltage, it remains reachable even when the operating system is hung or completely powered down. From the BMC console the administrator can issue a chassis power cycle (or separate power off and power on) command, providing hard-reset capability that is fully independent of the host OS, hypervisor, or production network path.
Wake-on-LAN can turn a system on, but it cannot force an immediate power-off when the machine is locked up. An SSH shutdown command requires a responsive operating system, which is not available in this scenario. Rebooting virtual machines from a hypervisor console affects only the guests, not the physical server that is frozen. Therefore, using the BMC's remote power control function is the only method that meets the requirements.