[lkml]   [1997]   [Feb]   [11]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: IDE Disk Problems

On Tue, 11 Feb 1997, Gary Vinson wrote:

> Hello,
> We are running Linux Redhat 3.0.3 (1.2.13 kernel) using Cyrix 586.
> We have lost one Western Digital Caviar 31600 (1.6GB) IDE drive and
> one WD Caviar (700MB) IDE drive. Now, when I say lost, I mean that when
> the system is rebooted, the disk sounds like it is physcially broke. There
> is a periodic clanging noise, as if one of the drive components is

Really loud, ugly sounding clanking, yes?

> actually damaged. At this point, we don't get past the BIOS.
> These two drives ran for approximately one month each
> before dieing. Prior to this, we noticed in our system messages:
> kernel: hdb: read_intr: status=0x4b { DriveRead DataRequest Index Error }
> " " read_intr: error=0x04 { Drive Status Error}
> " ide0: do_ide_reset: success

I recently posted nearly identical symptoms to the list. It was
also a WD Caviar 1.5 gig drive. The above messages began appearing
with varying but increasing frequency.

> Now, my question is, can (or is it remotely possible) linux be
> writing something out to the WD IDE drives to break the drive. The
> analogy being, that you can write to a video controller and cause
> damage to a monitor. One of our people have suggested that Linux is
> writing to sector 0 of the WD IDE drive.

Some one graced my mailbox with a suggestion to check out Western
Digital's web site (, I think) and look at the firmware
patch that they have for these drives. Read it carefully, it was
written carefully by some marketdroid. I remember a couple of years
ago some guy was claiming he could destroy WD ide drives with
software but (nicely) would not release his code. Perhaps this
was the mechanism he was using. I doubt that Linux is the cuplrit
here because I have used many WD Caviars with Linux and the drive
failure rate is the same as other OS's.

> Sorry to make this so long, but to add to this woe, we have had two other
> drives go bad on other Linux boxes, both running slackware with kernel of
> (1.2.8). These drives ran for approximately 6 months before expiring.
> On one, the conditions very similar to description above. The other system
> had a seagate ST5108A (1GB) where we did not get the clanging noise but
> received the "hda: read_intr ..." messages after Linux has loaded and
> during file check. It happened to be an inode where the first bad
> spot showed up. We changed out the drive and tried to format the
> seagate with DOS but format failed (just to see if DOS could
> read/write to the drive).

A few days after my 1.5 gig Caviar started doing this I decided to
replace it. I installed Slackware on a new 2.5 gig non-Caviar drive
and slaved the Caviar to it, copied /usr/src, /usr/local/, /root,
parts of /etc, /home and so on to the new drive. Then, unfortunately,
I went home without taking the Caviar out of the system. In the
morning the filesystems on the new drive were quite hosed. I repeated
the above procedure, removed the Caviar and I've been sailing
smoothly ever since.

I noticed that I would tend to get more of these type of errors
after the Caviar had been running for a while, i.e. perhaps the
Caviar is a little more sensitive to heat.

I won't be buying any more Caviar drives.


 \ /
  Last update: 2005-03-22 13:38    [W:0.069 / U:2.016 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site