Disks write-back cache b

Lawrence Chen lchen at pyramid.com
Thu Mar 30 10:24:34 PST 1995

On Mar 30,  9:44am, "Jim McGrath" wrote:
> Subject: Re: Disks write-back cache b
>         Reply to:   RE>Disks write-back cache behavior????
> Bill,
> I can answer the hard disk drive questions:
>  Disk drive has write-back cache enabled and a bus reset occurs:
> 	For all completed (cmd comp. msg. sent ) I/O Processes that has
> 	write data in the cache that has not been placed on media.
> 	What happens to that write data, does it get flushed to the bit
> 	bucket, does it flush it to media or does it not do any special
> 	processing of that data and will get to media at some point in
> 	time?
> We (Quantum) have always stated that once we cache data it will be written
> to the media barring a fatal drive problem (i.e. head crash) or power
> loss.  Specifically, we take care to execute the reset on the bus while
> continuing to keep track of the cached data and writing it back to the media.
>   ...what happens if:
> 	A media error is encountered:
> 		Is the media error reported as a deferred error along
> 		with the UNIT ATTENTION for the CA generated by the
> 		reset condition?
> We do not report deferred errors in general.  We found that customer drivers
> did not expect or want this.  Instead, we use the most extensive error
> recovery
> process possible in the drive and back it up with automatic reallocation of
> the
> error to a new portion of the media.  The only drive failure mode that can
> defeat us is the inability to write anywhere for seconds at a time - true in
> a fatal drive failure (i.e. head crash), but very hard to induce otherwise
> (e.g. shock and vibration could do it, but it is very hard to defeat the drive
> for all those retries without also violating the drive shock and vibration
> specification - we have never seen such a case in practice).
> Given this, there is no real use in reporting a deffered error, since there is
> no error recovery operation the system could perform on that drive that
> we have mot already attempted several times.
On some systems that implement RAID or MIRROR disks, deferred error
reporting could be helpful in recovering from this type of fault.


> After all of this work, any
> residue probability of an error is far lower than the probability that we
> simply would write the data incorrectly to the media (something that you
> cannot always detect unless you also do a verify of the media after it is
> written) - such events are the main contributer to the drive's hard error
> rate.
> Jim
>-- End of excerpt from "Jim McGrath"

       -m-------  Lawrence Chen           Pyramid Technology Corporation
     ---mmm-----  lchen at pyramid.com       3860 North First Street
   -----mmmmm---  VOICE: (408) 428-8974   Mail Stop SJ2-2-30
 -------mmmmmmm-  FAX:   (408) 428-7260   San Jose, CA 95134-1702
-- End of excerpt from Lawrence Chen
-- And something to think about ...
How many Zen masters does it take to screw in a light bulb?

None.  The Universe spines the bulb, and the Zen master stays out of
the way.

More information about the T10 mailing list