Technical Knowledge Base

Failing A Predictive Failing Drive

Prerequisites

Access via SSH or Serial is required
Field Engineer will need to log in via the “root” account utilizing the same password as grpadmin (password will need to be obtained from customer)
Once the disk has been initiated, you must wait for it to finish before replacing the failed drive.

Failing the Drive

  1. SSH into the member of the group that has the predictive failing disk

    Note: Use the Member IP, not the Group IP

  2. Sign into the member as root
  3. Enable mirroring (should see the output: 'enable mirroring')
raidtool -m enable
  1. Enable mirror rebuilding (should see the output: 'enable mirror rebuild')
raidtool -m rebuild_enable
  1. Initiate the failover:
raidtool -m <disk number> initiate

The system will automatically copy to the lowest numbered hot spare.
Following messages are expected:
initiateMirror: disk\=9
copy to spare initiated

  1. Monitor rebuild status via the GUI or the CLI using the normal status commands
  2. Check rebuild status
raidtool

The system will give you a completion message once finished copying to the hotspare
ERROR, event from storage array member1; Disk x failed

Preparing for the Replacement

  1. While the rebuild is in progress, disable mirroring (should see the message 'disable mirroring')
raidtool –m disable
  1. After the rebuild has completed and you receive the error stating that there is a drive failure, disable mirror rebuilding
raidtool –m rebuild_disable

This allows the new drive to come into the configuration as a spare