USB mass storage device split to 2 devices

debugcn Published at Dev

pts

When I connect an Hitachi Deskstar 7K4000 4TB hard drive in a Fantec DB-ALUE2e external SATA enclosure to a Lenovo T400 laptop using the USB 2.0 port, the hard drive gets split to 2 devices:

[ 1537.560160] usb 1-1: new high speed USB device using ehci_hcd and address 4
[ 1537.714687] usb 1-1: configuration #1 chosen from 1 choice
[ 1537.814573] Initializing USB Mass Storage driver...
[ 1537.814847] scsi4 : SCSI emulation for USB Mass Storage devices
[ 1537.815088] usbcore: registered new interface driver usb-storage
[ 1537.815094] USB Mass Storage support registered.
[ 1537.815818] usb-storage: device found at 4
[ 1537.815823] usb-storage: waiting for device to settle before scanning
[ 1542.810488] usb-storage: device scan complete
[ 1542.812251] scsi 4:0:0:0: Direct-Access     Generic  External         0200 PQ: 0 ANSI: 4
[ 1542.813976] scsi 4:0:0:1: Direct-Access     Generic  External         0200 PQ: 0 ANSI: 4
[ 1542.818156] sd 4:0:0:0: Attached scsi generic sg2 type 0
[ 1542.820658] sd 4:0:0:1: Attached scsi generic sg3 type 0
[ 1542.822331] sd 4:0:0:0: [sdb] 4294967295 512-byte logical blocks: (2.19 TB/1.99 TiB)
[ 1542.829215] sd 4:0:0:0: [sdb] Write Protect is off
[ 1542.829224] sd 4:0:0:0: [sdb] Mode Sense: 38 00 00 00   
[ 1542.829229] sd 4:0:0:0: [sdb] Assuming drive cache: write through
[ 1542.830485] sd 4:0:0:1: [sdc] 3519069872 512-byte logical blocks: (1.80 TB/1.63 TiB)
[ 1542.838112] sd 4:0:0:1: [sdc] Write Protect is off
[ 1542.838120] sd 4:0:0:1: [sdc] Mode Sense: 38 00 00 00   
[ 1542.838125] sd 4:0:0:1: [sdc] Assuming drive cache: write through
[ 1542.844988] sd 4:0:0:0: [sdb] Assuming drive cache: write through
[ 1542.844997]  sdb:
[ 1542.876247] sd 4:0:0:1: [sdc] Assuming drive cache: write through
[ 1542.876255]  sdc: unknown partition table
[ 1542.901098]  unknown partition table
[ 1542.907979] sd 4:0:0:0: [sdb] Assuming drive cache: write through
[ 1542.907988] sd 4:0:0:0: [sdb] Attached SCSI disk
[ 1542.916738] sd 4:0:0:1: [sdc] Assuming drive cache: write through
[ 1542.916746] sd 4:0:0:1: [sdc] Attached SCSI disk

I've tried this with Linux kernels 2.6.32, 2.6.35 and 3.6.6.

If I connect it over eSATA, the complete device gets recognized:

[3867153.075703] sd 9:0:0:0: [sdd] 7814037168 512-byte logical blocks: (4.00 TB/3.63 TiB)
[3867153.075708] sd 9:0:0:0: [sdd] 4096-byte physical blocks
[3867153.075862] sd 9:0:0:0: [sdd] Write Protect is off
[3867153.075868] sd 9:0:0:0: [sdd] Mode Sense: 00 3a 00 00
[3867153.075906] sd 9:0:0:0: [sdd] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
[3867153.082355]  sdd: unknown partition table  
[3867153.098386] sd 9:0:0:0: [sdd] Attached SCSI disk

Oddly enough, 1 sector is missing: 7814037168 - 4294967295 - 3519069872 == 1.

My questions:

Is it the inherent limitation of the USB mass storage (usb-storage) protocol that a single device can be at most 2 TB - 512 bytes large -- or is it the limitation of the Fantec external enclosure -- or is it a Linux kernel issue?
Is it the inherent limitation of the USB mass storage (usb-storage) protocol that sectors are always 512 bytes large, and the minimum I/O size of the hard drive (4096 bytes in this case) is ignored -- or is it the limitation of the Fantec external enclosure -- or is it a Linux kernel issue?
Would it work any better with USB 3.0?

Mark Johnson

I don't think there are any limits like that in the USB mass storage protocol. There are >2TB exeternal drives on the market with USB interfaces that work just fine with a single volume.
I don't think the USB mass storage protocol has a sector size limit, since there is a Seagate external 3TB drive with a USB interface that is known to report 4K logical sectors.
I doubt USB 3.0 would be any better

I think your problems are with the USB<--->SATA bridge chip in the enclosure. It has to bridge the ATA-derived commands the SATA drive is using to the SCSI-derived commands USB is using. Why it's apparently splitting the physical drive into two logical drives, I have no idea. Try another enclosure.

Collected from the Internet

Please contact [email protected] to delete if infringement.