USB Fehler bei aufwachender Platte

[Wolti]

Hallo zusammen,

ich habe aktuell ein kleines Problem, was mich in den Wahnsinn treibt. Die Folgende Hardware spielt zusammen:

USB 3.0 über xhci_hcd an einem AMD Board mit GBit Lan (Samba)
Ubuntu 12.04 mit Linux 3.2.0-30-generic

Die Platten sind in einer 5 Bay von Sharkoon im Raid 5 über USB 3 am Board
Chipsatz: JMicron JMB393 und JMB 539
LVM mit Einteilung der gesamten Kapazität in Laufwerke und über LUKS mit Ext/4 per DeviceMapper im System.

Soweit funktioniert das System super. Die Platten erledigen ihren Dienst, die Gehäuse auch und Samba verteilt fleißig. Platten sind gechecked und OK.

Jetzt aber zum Fehler:
DIE PLATTEN SCHALTEN NACH 5 MIN IN DEN RUHEZUSTAND
Das ist soweit gewünscht. Bei reinen Directory Zugriffen wachen die Platten auch sauber auf. Werden jedoch mehr Daten z.B. über Samba angefordert und die Platten wachen dann auf (z.B. eingeschlafen, weil ein Film pausiert wurde und bei erneutem Play wird aufgewacht) bricht der Datenstrom ab und es kann nicht mehr auf die Platten zugegriffen werden, bis ein Neustart erfolgt ist.

System technisch sieht das dann so aus:

Oct  6 14:56:12 Server kernel: [75749.940965] usb 9-1: reset SuperSpeed USB device number 2 using xhci_hcd
Oct  6 14:56:12 Server kernel: [75749.957910] usb 9-1: device firmware changed
Oct  6 14:56:12 Server kernel: [75749.957990] sd 6:0:0:0: Device offlined - not ready after error recovery
Oct  6 14:56:12 Server kernel: [75749.958017] sd 6:0:0:0: [sdb] Unhandled error code
Oct  6 14:56:12 Server kernel: [75749.958024] sd 6:0:0:0: [sdb]  Result: hostbyte=DID_ABORT driverbyte=DRIVER_OK
Oct  6 14:56:12 Server kernel: [75749.958035] sd 6:0:0:0: [sdb] CDB: Read(10): 28 00 ed 84 37 b0 00 00 f0 00
Oct  6 14:56:12 Server kernel: [75749.958062] end_request: I/O error, dev sdb, sector 3984865200
Oct  6 14:56:12 Server kernel: [75749.958131] sd 6:0:0:0: rejecting I/O to offline device
Oct  6 14:56:12 Server kernel: [75749.958148] usb 9-1: USB disconnect, device number 2
Oct  6 14:56:12 Server kernel: [75749.958166] sd 6:0:0:0: [sdb] killing request
Oct  6 14:56:12 Server kernel: [75749.958190] sd 6:0:0:0: rejecting I/O to offline device
Oct  6 14:56:12 Server kernel: [75749.958219] sd 6:0:0:0: rejecting I/O to offline device
Oct  6 14:56:12 Server kernel: [75749.958258] sd 6:0:0:0: [sdb] Unhandled error code
Oct  6 14:56:12 Server kernel: [75749.958265] sd 6:0:0:0: [sdb]  Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK
Oct  6 14:56:12 Server kernel: [75749.958276] sd 6:0:0:0: [sdb] CDB: Read(10): 28 00 ed 84 38 a0 00 00 10 00
Oct  6 14:56:12 Server kernel: [75749.958300] end_request: I/O error, dev sdb, sector 3984865440
Oct  6 14:56:12 Server kernel: [75749.966530] xhci_hcd 0000:06:00.0: xHCI xhci_drop_endpoint called with disabled ep ffff88022cc50380
Oct  6 14:56:12 Server kernel: [75749.966543] xhci_hcd 0000:06:00.0: xHCI xhci_drop_endpoint called with disabled ep ffff88022cc503c0
Oct  6 14:56:12 Server kernel: [75750.084167] usb 9-1: new SuperSpeed USB device number 4 using xhci_hcd
Oct  6 14:56:12 Server kernel: [75750.104937] scsi8 : usb-storage 9-1:1.0
Oct  6 14:56:13 Server kernel: [75751.107007] scsi 8:0:0:0: Direct-Access     SSI H/W  RAID5                 PQ: 0 ANSI: 5
Oct  6 14:56:13 Server kernel: [75751.110349] sd 8:0:0:0: Attached scsi generic sg1 type 0
Oct  6 14:56:13 Server kernel: [75751.117190] sd 8:0:0:0: [sdd] Very big device. Trying to use READ CAPACITY(16).
Oct  6 14:56:13 Server kernel: [75751.117551] sd 8:0:0:0: [sdd] 7813726208 512-byte logical blocks: (4.00 TB/3.63 TiB)
Oct  6 14:56:13 Server kernel: [75751.118268] sd 8:0:0:0: [sdd] Write Protect is off
Oct  6 14:56:13 Server kernel: [75751.118279] sd 8:0:0:0: [sdd] Mode Sense: 28 00 00 00
Oct  6 14:56:13 Server kernel: [75751.119676] sd 8:0:0:0: [sdd] No Caching mode page present
Oct  6 14:56:13 Server kernel: [75751.119692] sd 8:0:0:0: [sdd] Assuming drive cache: write through
Oct  6 14:56:13 Server kernel: [75751.122286] sd 8:0:0:0: [sdd] Very big device. Trying to use READ CAPACITY(16).
Oct  6 14:56:13 Server kernel: [75751.124190] sd 8:0:0:0: [sdd] No Caching mode page present
Oct  6 14:56:13 Server kernel: [75751.124199] sd 8:0:0:0: [sdd] Assuming drive cache: write through
Oct  6 14:56:13 Server kernel: [75751.144054]  sdd: unknown partition table
Oct  6 14:56:13 Server kernel: [75751.145085] sd 8:0:0:0: [sdd] Very big device. Trying to use READ CAPACITY(16).
Oct  6 14:56:13 Server kernel: [75751.147042] sd 8:0:0:0: [sdd] No Caching mode page present
Oct  6 14:56:13 Server kernel: [75751.147057] sd 8:0:0:0: [sdd] Assuming drive cache: write through
Oct  6 14:56:13 Server kernel: [75751.147067] sd 8:0:0:0: [sdd] Attached SCSI disk
Oct  6 14:57:03 Server kernel: [75801.061514] Buffer I/O error on device dm-12, logical block 0
Oct  6 14:57:03 Server kernel: [75801.061526] lost page write due to I/O error on dm-12
Oct  6 14:57:03 Server kernel: [75801.061548] EXT4-fs error (device dm-12): __ext4_get_inode_loc:3676: inode #144965633: block 579862560: comm smbd: unable to read itable block
Oct  6 14:57:03 Server kernel: [75801.061566] EXT4-fs error (device dm-12) in ext4_reserve_inode_write:4494: IO failure
Oct  6 14:57:03 Server kernel: [75801.061575] EXT4-fs (dm-12): previous I/O error to superblock detected
Oct  6 14:57:03 Server kernel: [75801.061752] Buffer I/O error on device dm-12, logical block 0
Oct  6 14:57:03 Server kernel: [75801.061759] lost page write due to I/O error on dm-12
Oct  6 14:57:03 Server kernel: [75801.061846] EXT4-fs (dm-12): previous I/O error to superblock detected
Oct  6 14:57:03 Server kernel: [75801.062026] Buffer I/O error on device dm-12, logical block 0
Oct  6 14:57:03 Server kernel: [75801.062032] lost page write due to I/O error on dm-12
Oct  6 14:57:03 Server kernel: [75801.062047] EXT4-fs error (device dm-12): __ext4_get_inode_loc:3676: inode #144965634: block 579862560: comm smbd: unable to read itable block
Oct  6 14:57:03 Server kernel: [75801.066975] EXT4-fs (dm-12): previous I/O error to superblock detected
Oct  6 14:57:03 Server kernel: [75801.067192] Buffer I/O error on device dm-12, logical block 0
Oct  6 14:57:03 Server kernel: [75801.067199] lost page write due to I/O error on dm-12
Oct  6 14:57:03 Server kernel: [75801.067219] EXT4-fs error (device dm-12): ext4_find_entry:935: inode #144973825: comm smbd: reading directory lblock 0
Oct  6 14:57:03 Server kernel: [75801.073130] EXT4-fs (dm-12): previous I/O error to superblock detected
Oct  6 14:57:03 Server kernel: [75801.073346] Buffer I/O error on device dm-12, logical block 0
Oct  6 14:57:03 Server kernel: [75801.073354] lost page write due to I/O error on dm-12
Oct  6 14:57:03 Server kernel: [75801.073372] EXT4-fs error (device dm-12): ext4_find_entry:935: inode #144973826: comm smbd: reading directory lblock 0
Oct  6 14:57:07 Server kernel: [75804.833780] Aborting journal on device dm-12-8.
Oct  6 14:57:07 Server kernel: [75804.834043] Buffer I/O error on device dm-12, logical block 291536896
Oct  6 14:57:07 Server kernel: [75804.834050] lost page write due to I/O error on dm-12
Oct  6 14:57:07 Server kernel: [75804.834193] JBD2: I/O error detected when updating journal superblock for dm-12-8.
Oct  6 14:57:26 Server kernel: [75824.553688] Buffer I/O error on device dm-11, logical block 0
Oct  6 14:57:26 Server kernel: [75824.553700] lost page write due to I/O error on dm-11
Oct  6 14:57:26 Server kernel: [75824.553719] EXT4-fs error (device dm-11): ext4_find_entry:935: inode #17828211: comm smbd: reading directory lblock 0
Oct  6 14:57:40 Server kernel: [75837.824995] Aborting journal on device dm-10-8.
Oct  6 14:57:40 Server kernel: [75837.825259] Buffer I/O error on device dm-10, logical block 12615680
Oct  6 14:57:40 Server kernel: [75837.825267] lost page write due to I/O error on dm-10
Oct  6 14:57:40 Server kernel: [75837.825410] JBD2: I/O error detected when updating journal superblock for dm-10-8.
Oct  6 14:57:41 Server kernel: [75838.886139] Buffer I/O error on device dm-10, logical block 0
Oct  6 14:57:41 Server kernel: [75838.886151] lost page write due to I/O error on dm-10
Oct  6 14:57:41 Server kernel: [75838.886306] EXT4-fs error (device dm-10): ext4_journal_start_sb:327: Detected aborted journal
Oct  6 14:57:41 Server kernel: [75838.886323] EXT4-fs (dm-10): Remounting filesystem read-only
Oct  6 14:57:41 Server kernel: [75838.886334] EXT4-fs (dm-10): previous I/O error to superblock detected
Oct  6 14:57:41 Server kernel: [75838.886594] Buffer I/O error on device dm-10, logical block 0
Oct  6 14:57:41 Server kernel: [75838.886601] lost page write due to I/O error on dm-10

Wie man hier schön erkennen kann sterben alle Operationen hier:

Oct  6 14:56:12 Server kernel: [75749.958148] usb 9-1: USB disconnect, device number 2
Oct  6 14:56:12 Server kernel: [75749.958166] sd 6:0:0:0: [sdb] killing request
Oct  6 14:56:12 Server kernel: [75749.958190] sd 6:0:0:0: rejecting I/O to offline device
Oct  6 14:56:12 Server kernel: [75749.958219] sd 6:0:0:0: rejecting I/O to offline device
Oct  6 14:56:12 Server kernel: [75749.958258] sd 6:0:0:0: [sdb] Unhandled error code
Oct  6 14:56:12 Server kernel: [75749.958265] sd 6:0:0:0: [sdb]  Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK
Oct  6 14:56:12 Server kernel: [75749.958276] sd 6:0:0:0: [sdb] CDB: Read(10): 28 00 ed 84 38 a0 00 00 10 00
Oct  6 14:56:12 Server kernel: [75749.958300] end_request: I/O error, dev sdb, sector 3984865440
Oct  6 14:56:12 Server kernel: [75749.966530] xhci_hcd 0000:06:00.0: xHCI xhci_drop_endpoint called with disabled ep ffff88022cc50380
Oct  6 14:56:12 Server kernel: [75749.966543] xhci_hcd 0000:06:00.0: xHCI xhci_drop_endpoint called with disabled ep ffff88022cc503c0

Dann wird das Gerät neu erkannt, (sd 8:0:0:0), kann aber nicht zugeordnet werden. Die dm's schmieren ab und alles andere als ein Reboot endet unsauber.
Danach ist alles wieder komplett OK.

Falls jm. einen Rat geben kann, würde er mir viele graue Haare ersparen. Vielen Dank im Voraus,

Daniel