On 24/01/2011 21:22, STeve Andre' wrote:
On 01/24/11 18:41, Martin Foster wrote:
I was curious as to what would be the expected behaviour when
attempting to read or write to a bad sector on a drive?
I have been encountering hard lock-ups on an MacPPC 4.7 machine when
running rsync against another host. Whenever it reaches a certain file
the system would lock up to anything that requires access to the
physical drive. Hence pf would work, but Apache, future logins or even
attempts to cancel the rsync process would simply do nothing.
Hence is this an expected behaviour, or should I be looking at
something else? And if it is a bad sector, any way to map and disable
them for use?
Thanks.
Martin Foster
[email protected]
With the cost of disks these days, it's insane to do anything other than
copy the data to another disk, and not look back.
I've had a variety of disk write errors, including a hang from the disk
when one of some 10's of bad sectors was written do. If you are "lucky"
you'll see a soft error in /var/log/messages. Reading your post again I
see that you've possibly crashed into one of my problems. The bottom
line is that the disk is dying, and I sure wouldn't trust it.
--STeve Andre'
The drive is returning a fair amount of errors at the console level:
> wd0j: device timeout writing fsbn...
> retrying
> wd0: soft error connected
Even in these cases the lock up is perhaps a 15 second pause. Nothing
like the hard lock-ups that I was seeing prior to the format of that
partition.
At first, I thought it was perhaps a driver issue (associated with high
IO), but certainly seems to be a problem with the drive itself.
Unfortunately, I chose the machine for its small form factor to serve as
a home server (Mac Mini) so its limited to one drive and its not the
easiest to work on.
I'll try to configure an alternate machine using another machine and see
about replacing the antiquated (ATA) though new Seagate drive.
