Is this hdd toast???

[Michael Hipp]

Net Llama! wrote:
> On Wed, 28 Jan 2004, Michael Hipp wrote:
> 
>>My main server here at the SOHO developed a problem about 2am. Some cron
>>jobs were reporting i/o errors and several hung. I'm doing fsck -t ext3
>>/dev/hdd1. It has, so far, come up with maybe 150 of these:
>>
>>-----------
>>Error reading block 1632178 (Attempt to read block from filesystem
>>resulted in short read) while reading indirect blocks of inode 803514.
>>Ignore error<y>? yes
>>
>>Force rewrite<y>? yes
>>
>>Inode 803514, i_blocks is 312, should be 104. Fix<y>? yes
>>-----------
>>
>>I took the default answers (yes). Is this hard drive dying, or did some
>>errant process just trounce all over the fs?
> 
> 
> If you run fsck again manually, does it finish cleanly?  Looking in your
> messages log do you see any other errors or problems in the days prior to
> this problem?

Well more than one pass thru the thing continued to get countless errors 
of the above type. So I guess it's a goner.

The problem appears to have started exactly here according to 
/var/log/messages:

-------
Jan 28 02:05:04 linux kernel: hdd: dma_intr: status=0x59 { DriveReady 
SeekComplete DataRequest Error }
Jan 28 02:05:04 linux kernel: hdd: dma_intr: error=0x40 { 
UncorrectableError }, LBAsect=129499311, high=7, low=12058799, 
sector=129499248
Jan 28 02:05:04 linux kernel: end_request: I/O error, dev 16:41 (hdd), 
sector 129499248
Jan 28 02:05:04 linux kernel: EXT3-fs error (device ide1(22,65)): 
ext3_get_inode_loc: unable to read inode block - inode=8093752, 
block=16187406
-------

Anyway I have a new one on order and an RMA approved for the old one.

I'm truly thankful for backups. This pup had some important stuff on it. 
And the night's backups had run not long before.

Michael