Symptom

The vdb disk of a node is damaged and the node cannot be recovered after reset.

Error Scenarios

On a normal node, delete the LV and VG. The node is unavailable.
Reset an abnormal node, and a syntax error is reported. The node is unavailable.
The following figure shows the details.

Fault Locating

If the volume group (VG) on the node is deleted or damaged and cannot be identified, you need to manually restore the VG first to prevent your data disks from being formatted by mistake during the reset.

Solution

Create a PV and a VG again. In this example, the following error message is displayed:

root@host1:~# pvcreate /dev/vdb
Device /dev/vdb excluded by a filter

This is because the added disk is created on another VM and has a partition table. The current VM cannot identify the partition table of the disk. You need to run the parted commands for three times to re-create the partition table.

root@host1:~# parted /dev/vdb
GNU Parted 3.2
Using /dev/vdb
Welcome to GNU Parted! Type 'help' to view a list of commands.
(parted) mklabel msdos
Warning: The existing disk label on /dev/vdb will be destroyed and all data on this disk will be lost. Do you want to continue?
Yes/No? yes
(parted) quit
Information: You may need to update /etc/fstab.

Run pvcreate again. When the system asks you whether to erase the DOS signature, enter y. The disk is created as a PV.

root@host1:~# pvcreate /dev/vdb
WARNING: dos signature detected on /dev/vdb at offset 510. Wipe it? [y/n]: y
Wiping dos signature on /dev/vdb.
Physical volume "/dev/vdb" successfully created

Create a VG.
Check the Docker disks of the node. If the disks are /dev/vdb and /dev/vdc, run the following command:
```
root@host1:~# vgcreate vgpaas /dev/vdb /dev/vdc
```
If there is only the /dev/vdb disk, run the following command:
```
root@host1:~# vgcreate vgpaas /dev/vdb
```
After the creation is complete, reset the node.

What Should I Do If the vdb Disk of a Node Is Damaged and the Node Cannot Be Recovered After Reset?

Symptom

Fault Locating

Solution