Preprocessing for Bare Metal ECSs

Scenarios

Before authorizing the replacement of a faulty local disk, you need to preprocess the local disk.

This section describes how to preprocess a faulty local disk of a bare metal ECS before authorizing the replacement of this disk.

Replacing a local disk will lose data on it. If you do not need these data, perform operations in this section to preprocess the disk to be replaced.

Replacing a local disk of a bare metal ECS may restart or power off the host. Before you preprocess a local disk, ensure that services on the bare metal ECS are stopped or that stopping the bare metal ECS during the replacement will not affect services running on it.

If you need to retain data on the local disk or if services on the bare metal ECS cannot be stopped, stop your operation here and contact technical support.

Procedure (Linux ECS)

Obtain the SN of the faulty local disk.

1. Log in to the management console.
2. Click in the upper left corner and select your region and project.
3. Click . Under Compute, click Elastic Cloud Server.
4. In the navigation pane, choose Events.
5. In the event list, check Description of the Local disk replacement event to obtain the SN of the faulty local disk.

   For example, the SN is 99K0A046FL3G.

Obtain the mount point of the faulty local disk.

1. Remotely log in to the bare metal ECS whose disk needs to be preprocessed.
2. Obtain the drive letter matching the SN of the faulty disk.

   ll /dev/disk/by-id/ | grep SN

   For example, if the SN is 99K0A046FL3G, run the following command:

   ll /dev/disk/by-id/ | grep 99K0A046FL3G

   The following information is displayed:

   /lrxxrxxrxx 1 root root 9 Sep 18 19: 20 ata-MG06ACA10TE_99K0A046FL3G ->./../sda
   lrwxrwxrwx 1 root root 10 Sep 18 19: 20 ata-MG06ACA10TE_99K0A046FL3G-part1->./../sda1

3. Obtain the mount point of the faulty local disk.

   df -Th | grep drive-letter

   For example, if the drive letter is /dev/sda1, run the following command:

   df -Th| grep /dev/sda1

   The following information is displayed:

   /dev/sdal ext4 9.1T 7.5T 1.6T 83% /srv/data

Preprocessing a local disk to be replaced will lose data on it. If you need to retain the data, stop your operation here and contact technical support.

Unmount the faulty local disk.

1. Unmount the faulty local disk.

   umount mount-point

   For example, if the mount point is /data, run the following command:

   umount /data

Check whether automatic mounting of the faulty disk is configured in /etc/fstab of the bare metal ECS.

If automatic mounting is configured for the faulty disk in /etc/fstab of the bare metal ECS, delete or comment out the configuration so that the bare metal ECS will not enter the emergency mode after its local disk is replaced.

1. Obtain the UUID of the disk partition.

   blkid disk-partition

   For example, if the disk partition is /dev/sda1, run the following command:

   blkid /dev/sda1

   The following information is displayed:

   /dev/sdal: UUID="e7100f3e-af65-49da-a013-c4ace8e5aba7" TYPE ="ext4"RTLABEL="logical"PARTUUID="cd358d32-e02b-4b23-bbd9–8a8bdae0e070"

2. Check whether /etc/fstab contains information about automatic mounting of the disk partition.

   cat /etc/fstab | grep disk-partition-UUID

   For example, if the UUID is e7100f3e-af65-49da-a013-c4ace8e5aba7, run the following command:

   cat /etc/fstab | grep e7100f3e-af65-49da-a013-c4ace8e5aba7

   If the following information is displayed, the /etc/fstab file contains information about automatic mounting of the disk partition:

   UUID=e7100f3e-af65-49da-a01 3-c4ace8e5aba7 /srv/data ext4 defaults,noatime, nodiratime 1 0

3. In the /etc/fstab file, delete or comment out the configuration of automatic mounting for the disk partition.
   1. Open the /etc/fstab file.

      vi /etc/fstab

   2. Press i to enter the editing mode.
   3. Delete or comment out the configuration of automatic mounting found in 2.
      To comment out the configuration, add # in front of it. For example:

      # UUID=e7100f3e-af65-49da-a01 3-c4ace8e5aba7/srv/data ext4 defaults,noatime, nodiratime 1 0

   4. Press Esc, enter :wq, and press Enter to save the change and exit the editing mode.

Stop the bare metal ECS.

1. In the event list, check the name or ID of the bare metal ECS (Instance Name/ID in the Local disk replacement event).

2. In the navigation pane, choose Elastic Cloud Server. Locate the bare metal ECS based on its name or ID. In the Operation column, choose More > Stop.
   

   After you complete preprocessing the local disk, restart the bare metal ECS to synchronize the disk information to the virtualization layer.

Procedure (Windows ECS)

Obtain the SN of the faulty local disk.

1. Log in to the management console.
2. Click in the upper left corner and select your region and project.
3. Click . Under Compute, click Elastic Cloud Server.
4. In the navigation pane, choose Events.
5. In the event list, check Description of the Local disk replacement event to obtain the SN of the faulty local disk.

   For example, the SN is WS22LX16.

Obtain the disk ID to confirm the disk to be replaced.

1. Remotely log in to the bare metal ECS whose disk needs to be preprocessed.
2. Open Windows PowerShell as an administrator and obtain the disk ID based on the SN.

   Get-Disk | select Number, SerialNumber

   Figure 1 Checking the mapping between the disk ID and SN
   
   

   Replacing a local disk will lose data on it. If you need to retain the data, stop your operation here and contact technical support.

Stop the bare metal ECS.

1. In the event list, check the name or ID of the bare metal ECS (Instance Name/ID in the Local disk replacement event).

2. In the navigation pane, choose Elastic Cloud Server. Locate the bare metal ECS based on its name or ID. In the Operation column, choose More > Stop.
   

   After you complete preprocessing the local disk, restart the bare metal ECS to synchronize the disk information to the virtualization layer.