Sun Enterprise 6x00, 5x00, 4x00, and 3x00 Systems Dynamic Reconfiguration User's Guide

Removing a Board

If you have not already done so, read "General Preparations".

Terminating I/O Devices

Terminate the use of all devices in the board. All I/O devices must be closed before they are unconfigured.

  1. Terminate all usage of devices on the board.

    1. To identify the components that are on the board to be unconfigured, use the ifconfig, mount, pf, or swap commands.

    2. To see which processes have these devices open, use the fuser(1M) command.

    3. Ensure that any networking interfaces on the board are not in use. All storage devices attached to the board should be unmounted and closed. See "I/O Board Unconfiguration".


      Note -

      DR does not automatically terminate network use or close devices. There currently is no way to ensure that the use of the network remains terminated or that all devices remain closed. Other clients may remount them between the time of the unmount and the unconfigure operations, so be careful.


    4. If AP (alternate pathing) is in use on the system, switch all board functions to the alternate board. Wait until all of the alternate paths are functioning before proceeding to Step 2.

    5. If AP is not available, warn all users to stop using the functions that the board provides.

  2. If the redundancy features of Alternate Pathing or Solstice DiskSuite mirroring are used to access a device connected to the board, reconfigure these subsystems so that the device or network is accessible by way of controllers on other system boards.

  3. Unmount file systems, including Solstice DiskSuite meta-devices that have a board resident partition. (Example: umount/partit)

  4. Remove Solstice DiskSuite or Alternate Pathing databases from board-resident partitions. The location of Solstice DiskSuite or Alternate Pathing databases is chosen by the user and can be changed.

  5. Remove any private regions used by Sun(TM) Enterprise Volume Manager(TM). Volume Manager by default uses a private region on each device that it controls, so such devices must be removed from Volume Manager control before they can be detached.

  6. Any Sun(TM) RSM Array(TM) 2000 controllers on the board that is being detached should be taken offline, using the rm6 or rdacutil commands.

  7. Remove disk partitions from the swap configuration.

  8. Either kill any process that directly opens a device or raw partition, or direct it to close the open device on the board.

  9. If a detach-unsafe device is present on the board, close all instances of the device and use modunload(1M) to unload the driver. If a detach-unsafe device is present on the board, close all instances of the device and use modunload(1M) to unload the driver.


    Caution - Caution -

    Unmounting file systems may affect NFS(TM) client systems.


Removal Procedure

  1. Terminate all usage of devices on the board.

    See "Terminating I/O Devices".

  2. Check the status of the board:

    • For a simple list containing board names, states, and conditions, enter:


      # cfgadm
      

    • For a more detailed list, enter:


      # cfgadm -v
      

    For a board removal or replacement, the states and conditions must be one of the following sets:

    • The board is ok:

      • Receptacle state--Connected

      • Occupant state--Configured

      • Condition--OK

    • The board is failing:

      • Receptacle state--Connected

      • Occupant state--Configured

      • Condition--Failing

  3. Unconfigure the board:


    # cfgadm -c unconfigure sysctrl#:slot#
    

    For sysctrl#:slot# (the attachment point ID) use the board name that was listed in the status report of the previous step.

  4. Use the cfgadm command to see if the board is unconfigured.

    If the unconfigure operation failed:

    1. See "Removing Boards that Use Detach-Unsafe Drivers".

    2. See "Quiescence".

    3. Resolve the problem.

    4. Unconfigure the board again (Step 3).


    Note -

    A failure of the unconfigure step results in a partially unconfigured condition. If this happens, attempt to unconfigure again. A configuration operation is not permitted at this point.


  5. When the board is successfully unconfigured, you can do one of the following:

    • Leave the board in the system unconfigured

    • Configure the board

    • Logically disconnect the board, in preparation for removal:


      # cfgadm -v -c disconnect sysctrl#:slot#
      

  6. The disconnection takes a few moments, so if you wish to remove the board from the card cage at this time, first verify the board status.

    1. Use cfgadm to verify that the board is logically disconnected.

    2. Check the LEDs on the board to verify that the board is electrically disconnected.

      The two outer LEDs must be off and the middle LED may be either lighted or off.

  7. After you have verified that the board is disconnected, you may physically remove or replace the board (see "Installing a Replacement Board").

If you wish, you can leave the board in place until a later time.


Caution - Caution -

If no replacement is available, leave the board in the slot, or fill the empty slot with a dummy board or a load board to maintain the proper flow of cooling air in the cardcage. For Enterprise 3000, 3500, 4000, 4500, 5000, and 5500 systems, use a dummy board. For Enterprise 6000 or 6500 systems, use a load board.