[SOLVED] Question about how /etc/hdparm.conf is populated by OMV

If in OMV Storage | Disks I select all drives one by one and press the Edit button. Then I set Advanced Power Management, Automatic Acoustic Management, and Spindown Time all to Disabled and save the changes.

When I look in /etc/hdparm.conf I see the following configuration for most of the disks:

}
/dev/disk/by-id/ata-WDC_WD30EFRX-68EUZN0_WD-WMC4N0FAHMNJ {
    write_cache = off

However, for three of the disks (all are identical WD Red 3TB) I see this instead:

/dev/disk/by-id/scsi-SATA_WDC_WD30EFRX-68_WD-WCC4N3YP9KRP {
    acoustic_management = 254
    apm = 1
    spindown_time = 120
    write_cache = on
}


}
/dev/disk/by-id/scsi-SATA_WDC_WD30EFRX-68_WD-WMC4N0FAHMNJ {
    acoustic_management = 254
    apm = 1
    spindown_time = 120
    write_cache = on
}
/dev/disk/by-id/scsi-SATA_WDC_WD30EFRX-68_WD-WMC4N0F63YTM {
    acoustic_management = 254
    apm = 1
    spindown_time = 120
    write_cache = off

Note that two of the disks have write_cache set to on, and one is set to off.

Is there something special about these disks that makes OMV override the settings I have made in the Storage | Disks panel? What makes OMV set different values for write_cache on otherwise identical disks?

It's been triple checked.

I even went so far as to delete the /etc/hdparm.conf file, make a small change to one drive in OMV Storage | Disks, save the changes, apply the configuration, and revert the changes. The /etc/hdparm.conf file is recreated and its contents are as above.

Solved.

Looking in config.xml <hdparm> section, the problematic disks were listed twice with differing UUIDs but the same serial numbers. I deleted the duplicate entries and /etc/hdparm is now normal.