[CentOS] Odd disk automount issue on C7.3.1611

Tue Apr 25 15:11:00 UTC 2017
Lamar Owen <lowen at pari.edu>

I'm posting here before going through all the fun to do up a real bug 
report to see if anyone else has seen this behavior.

I have two identical Dynex external USB3 drive enclosures with identical 
3TB drives in each enclosure.  The dmesg output shows:
$ dmesg |grep TOSHIBA
[   59.942546] scsi 7:0:0:0: Direct-Access     TOSHIBA 
DT01ACA300            PQ: 0 ANSI: 6
[   86.301123] scsi 8:0:0:0: Direct-Access     TOSHIBA 
DT01ACA300            PQ: 0 ANSI: 6
$

HOWEVER, the 'Disks' utility program only sees the first one plugged-in, 
and not the second one.  It doesn't matter which one I plug in first; 
the second one won't get seen by Disks nor will it get automounted.  It 
doesn't even do the 'USB device inserted' sound, for that matter.  As 
another oddity, one of the drives has a LUKS-encrypted ext4 filesystem; 
the other has an XFS filesystem; if I plug in the XFS-formatted drive 
first, the LUKS-encrypted on will no longer automount even if the 
XFS-formatted drive is unmounted and unplugged first, but the the 
converse does work (mount LUKS+ext4; unmount and unplug; plug in XFS, it 
gets automounted).

If I plug in the LUKS+ext4 disk first, enter the passphrase, and let it 
mount, then plug in the XFS drive, I can then manually (from the command 
line) mount it just fine and both drives work just fine. Both drives 
work fine when used singly, but not when both are plugged-in.

So, has anyone else seen this?  I don't mind filing a bug report, but I 
am at a bit of a loss as to what component to file it under. I haven't 
dug into too many logs or other files as yet, but I can if I need to do 
so.  I ran into this making an encrypted copy of the contents of the XFS 
disk, which will be wiped and reformated LUKS+ext4, and I'd like to make 
these two drives back each other up.

If I plug in a different USB3 drive, such as my 4TB Seagate, everything 
works as expected; it seems to only be the identical drives that confuse 
whatever process is detecting this.

-- 
Lamar Owen
Chief Information Officer
Pisgah Astronomical Research Institute
1 PARI Drive
Rosman, NC  28772
828-862-5554
www.pari.edu