On Sun, 2009-02-15 at 15:28 +0000, Jan T. Kim wrote:
After running a whole night, fsck had gone to (about) 1200000 out of 200000000 blocks, and the hard disk made a rhythmic, clicking noise. It had obviously slowed down during the night (it was at about 1000000 blocks when I checked before going to sleep), so I stopped that.
If the rhythmic clicking noise is the one I think it is I believe the cause is as follows:
1. The controller attempts a seek, moves the head what it calculates is the correct distance but fails to read the correct track/sector IDs from the disk.
2. In case the error is caused by the controller's idea of the head position being out of step with reality the controller moves the head to extreme position where there is a physical stop.
3. From this known position it re-attempts the seek.
4. The seek fails again and the whole process repeats.
If the problem is gradually getting worse it may be caused by small particles inside the housing with the disk which, as time goes by, scratch the surface and make the corresponding data unreadable.
For writes the controller can relocate the failed tracks to start with onto spares but eventually it will run out.
I'd see if I could get a refund or get the supplier to replace the disk with a different model as the quality control for the model you have it obviously not good enough.
On a related issue I also have an external USB hard drive that is a little problematic. I bought it for use with a satellite box but that sometimes complains it is not ready and if I attach it to the PC (Linux) I get some strange SCSI messages and it is very slow to settle down:
[195236.064040] usb 2-4: new high speed USB device using ehci_hcd and address 5 [195236.197229] usb 2-4: configuration #1 chosen from 1 choice [195236.197824] usb 2-4: New USB device found, idVendor=1058, idProduct=1001 [195236.197829] usb 2-4: New USB device strings: Mfr=1, Product=2, SerialNumber=3 [195236.197834] usb 2-4: Product: External HDD [195236.197837] usb 2-4: Manufacturer: Western Digital [195236.197841] usb 2-4: SerialNumber: 574341535930333636373533 [195236.613571] Initializing USB Mass Storage driver... [195236.617336] scsi6 : SCSI emulation for USB Mass Storage devices [195236.617699] usbcore: registered new interface driver usb-storage [195236.617705] USB Mass Storage support registered. [195236.621508] usb-storage: device found at 5 [195236.621513] usb-storage: waiting for device to settle before scanning [195241.620133] usb-storage: device scan complete [195241.622372] scsi 6:0:0:0: Direct-Access WD 6400AAK External 1.05 PQ: 0 ANSI: 4 [195241.634480] sd 6:0:0:0: [sde] 1250263728 512-byte hardware sectors: (640 GB/596 GiB) [195241.636288] sd 6:0:0:0: [sde] Write Protect is off [195241.636293] sd 6:0:0:0: [sde] Mode Sense: 21 00 00 00 [195241.636297] sd 6:0:0:0: [sde] Assuming drive cache: write through [195241.636970] sd 6:0:0:0: [sde] 1250263728 512-byte hardware sectors: (640 GB/596 GiB) [195241.639217] sd 6:0:0:0: [sde] Write Protect is off [195241.639221] sd 6:0:0:0: [sde] Mode Sense: 21 00 00 00 [195241.639225] sd 6:0:0:0: [sde] Assuming drive cache: write through [195241.639230] sde: sde1 [195241.644432] sd 6:0:0:0: [sde] Attached SCSI disk [195248.284481] sd 6:0:0:0: [sde] Sense Key : No Sense [current] [195248.284490] Info fld=0x0 [195248.284493] sd 6:0:0:0: [sde] Add. Sense: No additional sense information [195251.393481] sd 6:0:0:0: [sde] Sense Key : No Sense [current] [195251.393489] Info fld=0x0
Any ideas?
Regards, Steve.
Now I've done things from scratch the traditional way with fdisk and mkfs (without -c). The disk made a lot of weird noises while mkfs was running, but eventually mkfs completed. I then mounted the file system and exercised it by repeatedly copying a Linux kernel source tree onto it. This worked ok for a few times, then I decided to have a look with du at the growing amount of stuff on the disk. Upon that, the copying process started to fail with messages like
cp: writing `/media/mnt/dummy0008/include/net/ieee80211.h': Read-only file system cp: cannot create regular file `/media/mnt/dummy0008/include/net/checksum.h': Read-only file system
and ``tail -f /var/log/messages'' at this time showed multiple repetitions of
Feb 15 14:01:31 paxarchia kernel: [ 8431.175590] Info fld=0x0 Feb 15 14:01:31 paxarchia kernel: [ 8431.175590] sd 5:0:0:0: [sdb] Add. Sense: Internal target failure Feb 15 14:01:32 paxarchia kernel: [ 8431.280559] sd 5:0:0:0: [sdb] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE,SUGGEST_OK Feb 15 14:01:32 paxarchia kernel: [ 8431.280559] sd 5:0:0:0: [sdb] Sense Key : Hardware Error [current]
The hard disk made a constant, whirring sound and its green LED was blinking at 1Hz (approx).
When I umounted the file system, the whirring continued for several minutes and the umount command hang. After several minutes, though, the hard disk made a few clicks and stopped whirring, and umount completed.
Can any of you make sense out of all this?
Best regards & thanks in advance, Jan