Disk health check

[ldrrp]

Problem

Currently there are no checks for disk health. This is a problem if we want to ensure data is not lost.

Solution

On a cron check for the health of the disks and if any are unhealthy we should notify the user and log the issue.

• We must scan for all physical drives ignoring raid groups and partitions.
• We should add those drives to a list of disks to configure in config.yml.
• Include config.yml settings disk checks with a custom golang struct.
• We will need a new model for system commands.
• If smart is not enabled we should let the user know to enable it and how.
• Cron should be configured to run at a regular interval of at least once every week.

Example:

 - disks:
    - uuid: AISDYIAYSIDYUA
      options:
        - notification:
            enabled: true
        - log: true
    - uuid: BISDASDSAJFYUA
      options:
        - notification:
            enabled: true
        - log: true

Notify may be "true", "false" or "email" etc. Parsed as string. True will use server default.

Additionally we should allow the user to call the disk check directly.

Steps to Reproduce the Problem

1. ./server.bin -system disk-check

[ldrrp]

root@cloud:/docker-nas/server# udevadm info -q property -n /dev/sdq | grep -i type
DEVTYPE=disk
SCSI_TYPE=disk
ID_TYPE=disk
ID_PART_TABLE_TYPE=gpt
root@cloud:/docker-nas/server# udevadm info -q property -n /dev/md127 | grep -i type
DEVTYPE=disk
ID_PART_TABLE_TYPE=gpt
root@cloud:/docker-nas/server# udevadm info -q property -n /dev/md0 | grep -i type
DEVTYPE=disk
ID_FS_TYPE=ext4
root@cloud:/docker-nas/server# udevadm info -q property -n /dev/sda | grep -i type
DEVTYPE=disk
SCSI_TYPE=disk
ID_TYPE=disk
ID_FS_TYPE=linux_raid_member

[ldrrp]

https://github.com/dell/csi-baremetal/tree/master/pkg/base/linuxutils