[lug] Data integrity problems on new machine

[Alan Robertson]

Archer Sully wrote:
> 
> Alan Robertson wrote:
> >
> > But disks that are dying almost always complain.  I think it's some kind of
> > compatibility issue.
> >
> 
> I have a 9G IBM SCSI drive in my Solaris box that clearly
> has a bad spot on it that got through testing.  No real
> complaints, it just silently gives I/O errors.  It happens.
> The unit is brand-spanking-new.

Does it silently corrupt data, or does it give I/O errors?  Silently giving
I/O errors is an oxymoron.

This is a HUGE difference.  Most of the time, silent data corruption is NOT
the drive's fault.  It's somewhere else in the system (like the DMA
subsystem, disk controller, memory system, etc.).

> I have another 18G SCSI drive that is waiting for me to
> get an RMA for it that has a case temperature of well
> over 100 deg F right after power up.  It too gives no
> complaints.

Does it corrupt data, or just get hot?

Some drives are designed to get hot.

	-- Alan Robertson
	   alanr at suse.com