In my miserable post, I forgot to mention that the 5 month old 160GB hard disk that houses my now unused Windows XP installation is starting to flake. I keep getting a shitload of these in my logs:
Apr 29 11:14:23 localhost kernel: hda: dma_intr: status=0x51 { DriveReady SeekComplete Error }
Apr 29 11:14:23 localhost kernel: hda: dma_intr: error=0x40 { UncorrectableError }, LBAsect=6323152, high=0, low=6323152, sector=6323152
Apr 29 11:14:23 localhost kernel: end_request: I/O error, dev hda, sector 6323152
Apr 29 11:14:23 localhost kernel: NTFS-fs error (device hda1): ntfs_end_buffer_async_read(): Buffer I/O error, logical block 6323089.
Apr 29 11:14:23 localhost kernel: NTFS-fs error (device hda1): map_mft_record(): Failed with error code 5.
Apr 29 11:14:23 localhost kernel: NTFS-fs error (device hda1): ntfs_read_locked_inode(): Failed with error code -5. Marking corrupt inode 0x3dca as bad. Run chkdsk.
There seems to be about 4 bad sectors (48078695, 48078696, 6323151 and 6323152) from what I can gather, however I can read the disk fine and have copied all my important stuff off. Windows doesn’t seem to be complaining yet, well I only boot into it about once a week to check hardware for comparison against Linux (like my faulty TFT monitor) and I don’t actually do anything with it as such. Strangely, these errors only appear in the logs if I haven’t just rebooted after being in Windows.
While I don’t much care for Windows, I could lose it without a care, I was going to swap my hard disks around and put /home on the nice 160GB disk as I have about 5 or 6 GB spare on my 60GB drive that currently houses /home. I got down to about 800MB and had to delete a few big files that I don’t use much. I would have preferred to keep them though.
Now I have to try to convince the supplier that it’s a faulty disk and get them to replace it which means I have to pay to return it. Pain in the arse. Not only that, but this disk was the replacement for an 18 month old disk that I lost in November and now I’m having to replace it again, while I have another disk right next to it that I’ve had for around 3 years without ever having a problem (touch wood).
Fucking flaky hardware. Thankfully I’ve never been cursed myself with an identifiable hardware problem, possibly due to the fact that I’m great at diagnosing the cause of hardware faults, though I was responsible for someone elses machine that had such problems. It turned out to be flaky BIOS power management on an SiS motherboard. Turned it off and the problems went away.
Check out the SMART tools – If you’re using Debian try :
apt-get install smartmontools
They can monitor disks, and report problems…. this might let you know if the disk is about knackered or there are just fs consistency problems, or just a couple of bad sectors from you banging around.
David.