[ale] what exactly does a long smart hdd test do?

Sat May 12 15:35:01 EDT 2012

Oh, LORD, I told myself I was not going to get sucked into another Steve
Gibson cesspool religious discussion, yet here I am...

On Sat, 2012-05-12 at 10:29 -0400, Ron Frazier (ALE) wrote:
> Hi Justin,

> Thanks for the link. I'll take a look at it as well as rummage through
> Steve Gibson's website a bit. We had a discussion about this on the
> ALE list about a year ago. I think most of the docs there are for
> SpinRite 5. I'm much more interested in what it can do for modern
> drives, and that information is hard to find. I will say that are
> probably very few people on these mailing lists, including me, which
> know more about how hard drives work than Mr. Gibson. And, since he's
> the inventor of the SpinRite product, no one knows more about how that
> works than him.

I know a great deal (probably more than I really want to) about rotating
magnetic drives.  Before getting involved with computer security, I was
involved with designing embedded systems and controller boards for
drives including old OLD MFM drives all the way back to Seagate ST501
stepper drives (A WHOLE 5Meg of capacity, what will we do with all of
this?!?! - I still have one on my shelf), and the Phoenix cartridge
drives and even the open head assembly drives with the covers you screw
off when you insert the diskpacks.  I have actually, hands on,
recalibrated ancient "open frame head stack" drives with scopes and
adjusting to the sync patterns on their servo surfaces (we're talking
huge 8" and 14" drives here).  As I mentioned in one of my talks, I have
actually recovered data from damaged drives by removing the cover from
the drive and repairing a damaged actuator pole piece and spinning up
(open) and recovering an entire (whole 140MB Maxtor) drive (you get the
data off, one shot, and your done).

My personal opinion from both the disk storage angle and the security
angle is that Steve Gibson is now little more than snake-oil surviving
on past glories.  I've used SpinRite down through the years and even
mentioned it in my ALE talk a couple years ago on drive and disaster
recovery.  Sometimes it does good.  Sometimes it commits random acts of
terrorism and makes a bad situation worse (I've personally seen both).
My advice then and is, in data recovery, back EVERYTHING off the drive
that you can using dd-rescue before letting SpinRite NEAR that drive.
That was a mistake I made and which I won't make again.

Steve Gibson has a following, just like many other pundants and
commentators but his credibility, especially in the security community,
is incredibly low.  He makes a pronouncement that (to us) naively lame
and the masses go wild and eat it up.  The rest of us in the profession
shake our heads and mumble "give us a break".  Seems like he's been
laying low and quiet lately.  I haven't heard anything truly outrageous
from him in the news in a couple of years now.

SpinRite USE TO do really useful things that really centered around 3
things.

One was data fading due to magnetic reluctance and retention.  Older
drives might have poor or mediocre media coatings were known to
gradually loose data.  Mostly it was some defective runs of the ferric
oxide coatings but I also saw it in some Maxtor metallic film batches.
SpinRite would rewrite the data before it had a chance to become
unreadable and restore the magnetic field strength.  I'll give him that.
That was credible.  That was the real source of this idea of rewriting
the data to make the magnetic field stronger.  But it was only
applicable to really old drives, and I mean really old MFM drives and
ESDI drives.  This is NOT a significant factor with modern drives and
CERTAINLY not with modern high capacity drives with tight tolerance,
vertical recording media.  Last time I personally saw any true evidence
of "data fading due to magnetic reluctance" was over 25 years ago in
Maxtor ESDI drives with metal deposition thin film media surfaces.  They
had a really bad batch that was prone to it.

Another thing was just simply bad block recovery and remapping.  If a
block was marginal, Spinrite would make good efforts to get that data
back and then rewrite the data.  If the drive detected a truly bad error
on write, the block could be remapped to another good block.  This is
possibly still of some value today but dd-rescue and other similar tools
can do just as good if not better than SpinRite on the recovery side.
If you've got blocks going bad, you should probably be using the SMART
tools to see what the drive firmware diagnostics are telling you and
pitch a drive that's starting to go bad like that before it becomes
totally unusable.  SMART uses the firmware on the SMART enabled drives.
They can do things that SpinRite absolutely could not.  For that matter,
with modern IDE, SATA, and SAS drives, who knows...  Maybe SpinRite is
tapping into the SMART data under the hood.  I don't know how it could
through the BIOS and Dos interfaces but it's entirely possible.

Last thing, which was really his claim to fame, was head drift
correction and re-calibration.  Older drives had a separate surface and
head as the tracking or servo surface on which they wrote "servo
tracks".  These were written with special field patterns at the vendor
and where used by the drive servos to determine the head positioning of
the entire stack of heads (old drives you use to be able to do a "low
level format" and it actually meant something then).  It was a great
idea and a hell of an improvement over the old stepper motors in both
speed and reliability.  But, they had a few problems.  Over a long time
of heating up and cooling down, the heads would "drift" in relation to
the servo tracks on the servo surface so the data no longer aligned
properly.  Eventually, this would cause soft errors and you could
correct for some by seeking in the correct directions and with different
timing.  This IS where SpinRite would shine, no argument.  All the data
was periodically read and rewritten and the new tracks aligned with the
servo heads and servo tracks long before they could drift out of
calibration and start generating hard errors.

Modern drives use embedded servo patterns.  The servo "quadrature
signals" (unless they've invented something better by now) are actually
embedded in the tracks between each sector.  As you leave one sector and
before you reach another, you pass over a region of data that is
actually recorded on sort of "half tracks" with a signal that is out of
phase between these two half tracks by 90 degrees (quadrature).  The
heads sense this and the analog electronics adjusts the servos so this
quadrature signal is even.  That keeps you centered on the track.  You
no longer need to be calibrated to a servo head or servo track because
the servo signals are embedded in the tracks themselves.  That, largely,
is what made the concept of TRUE "low level formatting" pretty much
irrelevant for modern drives.  It also eliminated the entire issue of
"head drift".

So...  Modern drives are NOT subject to head drift or data fade through
magnetic reluctance and retention.  That leaves what?  Bad block
detection?  You can manage that through dd.  Anything else, the SMART
tools can beat anything SpinRite can come up with.

Regards,
Mike

> I see no need for a flame war at all. Hard drives are going to be
> around for a long time. I sometimes tell people I could power my house
> with solar power for the cost of the house. (Analogy follows.) Well,
> at $ 1 - $ 2 / GB, I could SSD my entire computer for the cost of the
> computer. Not happening. So, I think any material discussion we can
> have about how to make hard drives last longer and be more reliable is
> beneficial.
> 
> I say, if you want to hang around in the discussion, if there is more
> discussion, then please do.
> 
> Sincerely,
> 
> Ron
> 
> 
> --
> 
> Sent from my Android Acer A500 tablet with bluetooth keyboard and K-9 Mail.
> Please excuse my potential brevity.
> 
> (To whom it may concern. My email address has changed. Replying to former
> messages prior to 03/31/12 with my personal address will go to the wrong
> address. Please send all personal correspondence to the new address.)
> 
> (PS - If you email me and don't get a quick response, you might want to
> call on the phone. I get about 300 emails per day from alternate energy
> mailing lists and such. I don't always see new email messages very quickly.)
> 
> Ron Frazier
> 770-205-9422 (O) Leave a message.
> linuxdude AT techstarship.com
> 
> 
> Justin Goldberg <justgold79 at gmail.com> wrote:
> 
> But exactly how does it do its disabling of the drives internal ec and do its defect analysis? What kind of voodoo does it do on the drive? This has never been documented. It makes sense to use it on older drives (I keep 5.0 in my utility toolchest for the odd mfm/rll drive I encounter) but with the advanced error correction in newer drives it doesn't make sense.
> 
> I dug up this discussion from the comp.sys.ibm.pc.hardware.storage group:
> 
> http://groups.google.com/group/comp.sys.ibm.pc.hardware.storage/browse_thread/thread/958d7e095d15c3fc/ce0dec6a189c5eae?lnk=gst&q=Spinrite#ce0dec6a189c5eae
> 
> Here's one quote from Arno:
> 
> Their product used to have some merit in the old MFM times. Today, they are just cashing in on the remainder of that reputation.
> 
> 
> I don't want to engage in a flame-war, and the vagaries of low level magnetic storage isn't really my thing, so I'll drop out of the discussion. The discussion I linked to above goes quite in-depth.
> 
> 
> On Fri, 11 May 2012 21:09:45 -0400, ale-bounces at ale.org wrote:
> >
> _____________________________________________
> 
> Ale mailing list
> Ale at ale.org
> http://mail.ale.org/mailman/listinfo/ale
> See JOBS, ANNOUNCE and SCHOOLS lists at
> http://mail.ale.org/mailman/listinfo
> 
> 
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://mail.ale.org/mailman/listinfo/ale
> See JOBS, ANNOUNCE and SCHOOLS lists at
> http://mail.ale.org/mailman/listinfo

-- 
Michael H. Warfield (AI4NB) | (770) 985-6132 |  mhw at WittsEnd.com
   /\/\|=mhw=|\/\/          | (678) 463-0932 |  http://www.wittsend.com/mhw/
   NIC whois: MHW9          | An optimist believes we live in the best of all
 PGP Key: 0x674627FF        | possible worlds.  A pessimist is sure of it!
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 482 bytes
Desc: This is a digitally signed message part
Url : http://mail.ale.org/pipermail/ale/attachments/20120512/373600bc/attachment-0001.bin 