Hot-swapping a module

You can hot-swap a faulty disk shelf module, removing the faulty module and replacing it without disrupting data availability.

When you hot-swap a disk shelf module, you must ensure that you never disable the only path to a disk; disabling that single path results in a system outage.

Attention: If there is newer firmware in the /etc/shelf_fw directory than that on the replacement module, the system automatically runs a firmware update. This firmware update causes a service interruption on non-multipath HA AT-FCX installations, multipath HA configurations running versions of Data ONTAP prior to 7.3.1, and systems with non-RoHS AT-FCX modules.
  1. Verify that your storage system meets the minimum software requirements to support the disk shelf modules that you are hot-swapping. See the DiskShelf14mk2 and DiskShelf14mk4 FC, or DiskShelf14mk2 AT Hardware Service Guide for more information.
  2. Determine which loop contains the module you are removing, and determine whether any disks are single-pathed through that loop.
  3. Complete the following steps if any disks use this loop as their only path to a controller:
    1. Follow the cables from the module you want to replace back to one of the nodes, called NodeA.
    2. Enter the following command at the NodeB console: cf takeover
    3. Wait for takeover to be complete and make sure that the partner node, or NodeA, reboots and is waiting for giveback.
    Any module in the loop that is attached to NodeA can now be replaced.
  4. Put on the antistatic wrist strap and grounding leash.
  5. Disconnect the module that you are removing from the Fibre Channel cabling.
  6. Using the thumb and index fingers of both hands, press the levers on the CAM mechanism on the module to release it and pull it out of the disk shelf.
  7. Slide the replacement module into the slot at the rear of the disk shelf and push the levers of the cam mechanism into place.
    Attention: Do not use excessive force when sliding the module into the disk shelf; you might damage the connector.

    Wait 30 seconds after inserting the module before proceeding to the next step.

  8. Recable the disk shelf to its original location.
  9. Check the operation of the new module by entering the following command from the console of the node that is still running: environ shelf The node reports the status of the modified disk shelves.
  10. Complete the following steps if you performed a takeover previously:
    1. Return control of NodeA’s disk shelves by entering the following command at the console of the takeover node: cf giveback
    2. Wait for the giveback to be completed before proceeding to the next step.
  11. Test the replacement module.
  12. Test the configuration.