Cisco Networking

Locked Out of Cisco UCS? How to Recover the Master Admin Password | Lazy Admin Blog

Posted on Updated on

It’s the nightmare scenario: you need to make a critical service profile change, but the only admin password is lost or forgotten. Because Cisco UCS Manager doesn’t store passwords in a reversible format, you can’t “view” the old one. Instead, you must perform a password reset by power-cycling the Fabric Interconnects (FIs) and interrupting the boot sequence.

⚠️ Warning: This procedure requires a physical power cycle of the Fabric Interconnects. In a production environment, this will cause a temporary disruption in management connectivity and potentially data traffic if not handled correctly in a cluster.

Phase 1: The Pre-Flight Check

Before you pull the power cables, you need two pieces of information. If you still have read-only access or a lower-privilege account, gather these now:

  1. Identify the Roles: In a cluster, you must know which FI is Primary and which is Subordinate.
    • Path: Equipment > Fabric Interconnects > [FI Name] > General > High Availability Details.
  2. Verify Firmware Versions: You must know the exact Kernel and System firmware versions currently running.
    • Path: Equipment > Firmware Management > Installed Firmware.

Phase 2: Password Recovery (The Process)

Scenario A: Standalone Configuration

If you only have one Fabric Interconnect, the process is straightforward but requires downtime.

  1. Connect: Attach a console cable physically to the FI console port.
  2. Power Cycle: Turn the FI off and then back on.
  3. Interrupt Boot: As it boots, repeatedly press Ctrl+L or Ctrl+Shift+R until you see the loader > prompt.
  4. Boot Kernel: Load the kickstart/kernel image: loader > boot /installables/switch/ucs-6100-k9-kickstart.x.x.x.gbin
  5. Enter Config: Fabric(boot)# config terminal
  6. Reset Password: Fabric(boot)(config)# admin-password YourNewPassword123
  7. Load System: Exit config mode and boot the system image: Fabric(boot)# load /installables/switch/ucs-6100-k9-system.x.x.x.bin

Scenario B: Cluster Configuration (High Availability)

In a cluster, the order of operations is vital to ensure the database remains synchronized.

  1. Subordinate First: Power cycle the Subordinate FI and interrupt its boot to the loader > prompt. Leave it there.
  2. Primary Second: Power cycle the Primary FI and interrupt its boot to the loader > prompt.
  3. Reset on Primary: Follow the “Standalone” steps (4 through 7) on the Primary FI console.
  4. Bring up Subordinate: Once the Primary is back up and you can log into UCS Manager, go to the Subordinate console and boot its kernel and system images normally from the loader prompt.

Important Notes

  • Clear Text: When you type the admin-password command in the boot loader, the password displays in clear text on the screen. Ensure no one is shoulder-surfing!
  • Strong Passwords: UCS Manager requires at least one capital letter and one number.
  • Console Access: This cannot be done via SSH. You must have physical or terminal server access to the console port.

#CiscoUCS #DataCenter #CiscoProphet #SysAdmin #Networking #ITTech #Cisco #UCSManager #LazyAdmin #Infrastructure

This entry was posted in Cisco UCS and tagged , , , , , , , , .