Re: [OT] Howl of soul...

Re: [OT] Howl of soul...

[Samium Gromoff]

> > > Does anyone know what this bug actually is, and > > whether there's a possible
> > > workaround without disabling udma entirely?
> >    so if i will disable udma and switch to
> >   -X34 == multiword dma2, then corruptions will go away????
> >
> >    i have a VX chipset/Zida 5DVX with PIIX3...
>
> there's nothing wrong with the chipset/controller; isn't this thread
> about the well-known DTLA problem?  if so, then what mode you use
> is completely irrelevant, since the physical media is degrading.

     i feel like the media isn`t downgrading because
 the badblocks _arent_ physical: low-level drive
 reformat doesnt show anything.
     how i think it may be related to chipset/controller?
 when data is transferred in udma mode, there may be
 checksumming errors, and the wrong crc is being wrote
 to the disk.
     so when data is being accessed, bah - you are in
 shit...

 although, this is only a theory...

---


cheers,


   Samium Gromoff

Re: [OT] Howl of soul...

[Barry K. Nathan]

> > there's nothing wrong with the chipset/controller; isn't this thread
> > about the well-known DTLA problem?  if so, then what mode you use
> > is completely irrelevant, since the physical media is degrading.
> 
>      i feel like the media isn`t downgrading because
>  the badblocks _arent_ physical: low-level drive
>  reformat doesnt show anything.

A low-level format (using IBM DFT) is going to *silently* remap bad
parts of the disk. It's only going to complain once it's no longer
possible to remap the bad sectors. So, just because the low-level format
doesn't complain does not mean that there is no media degradation!

Also, many people seem to be correlating it to heat, and possibly to
flaws in IBM's implementation of glass platters in these drives. For
example, see
http://www.storagereview.com/welcome.pl?/http://www.storagereview.com/jive/sr/thread.jsp?forum=1&thread=16057

-Barry K. Nathan <barryn@pobox.com> 

Re: [OT] Howl of soul...

[Marco Colombo]

On Sun, 26 Aug 2001, Barry K. Nathan wrote:

> > > there's nothing wrong with the chipset/controller; isn't this thread
> > > about the well-known DTLA problem?  if so, then what mode you use
> > > is completely irrelevant, since the physical media is degrading.
> >
> >      i feel like the media isn`t downgrading because
> >  the badblocks _arent_ physical: low-level drive
> >  reformat doesnt show anything.
>
> A low-level format (using IBM DFT) is going to *silently* remap bad
> parts of the disk. It's only going to complain once it's no longer
> possible to remap the bad sectors. So, just because the low-level format
> doesn't complain does not mean that there is no media degradation!
>
> Also, many people seem to be correlating it to heat, and possibly to
> flaws in IBM's implementation of glass platters in these drives. For
> example, see
> http://www.storagereview.com/welcome.pl?/http://www.storagereview.com/jive/sr/thread.jsp?forum=1&thread=16057
>
> -Barry K. Nathan <barryn@pobox.com>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

Just to add something to the noise, I'm so happy with DTLA-307030's
that I'm not even considering buying something different. As I see it,
there's simply no competing product, full stop.  Recently I've bought
a 41Gb 60GXP, and I'm happy with that too.  It is true that I usually
stick with the same HW for a long while, so maybe I'm just being lucky.

.TM.
-- 
      ____/  ____/   /
     /      /       /			Marco Colombo
    ___/  ___  /   /		      Technical Manager
   /          /   /			 ESI s.r.l.
 _____/ _____/  _/		       Colombo@ESI.it

Re: [OT] Howl of soul...

[Samium Gromoff]

>  A low-level format (using IBM DFT) is going to *silently* remap bad
>  parts of the disk. It's only going to complain once it's no longer
>  possible to remap the bad sectors. So, just because the low-level format
>  doesn't complain does not mean that there is no media degradation!
    1. how to find problematic blocks?
           - just read, and if read fails goto 2.
	     (i.e. we found no new bad sectors)
	   - goto 2 on the sectors reported before as bad.
             (i.e. drive remembers sectors on which he had failures)
    2. what to do when i found problematic sector?
           - just see if it still usable.
         2a. i write to the sector, and after that i read crap.
               - sector is bad! should remap it!
         2b. i can write data to the sector, then reads goes ok.
               - hmm, i think that was kinda magnetetic storm, sector
 is still usable. do not remapping.

        in my case there was just magnetic storm, so the sector can be safely
     read/written again.

cheers,
Sam

Re: [OT] Howl of soul...

[Pavel Zaitsev]

Samium Gromoff(_deepfire@mail.ru)@Wed, Sep 26, 2001 at 03:01:38AM +0000:
> >  A low-level format (using IBM DFT) is going to *silently* remap bad
> >  parts of the disk. It's only going to complain once it's no longer
> >  possible to remap the bad sectors. So, just because the low-level format
> >  doesn't complain does not mean that there is no media degradation!

My drive has suffered simiar failure. After being fixed or replaced is
all fine. No problems and fast access times. If it fails again, I will
not buy IBM anymore. One of companies I worked for, wanted to buy a mainframe,
but technical staff they were sending us, were might as well been Futureshop/Frys
salesment types to help with understanding how loaner worked. Service for
rack equipment we ordered at the time was terrible. There were awfully fast,
but we had to figure all the stuff ourselves. Sort of like during MITS times,
were you put your computer together. Duh.
p.


-- 
Research causes cancer in rats.
110461387
http://gpg.md5.ca
http://perlpimp.com

Re: [OT] Howl of soul...

[Samium Gromoff]

> I am confused, you and your friend both had problems, and you used IBM and he
> used quantum?  Do they have similar electronics?
>
> Take a look at www.namesys.com, there is a badblocks > patch, god@namesys.com wil tell you details about it.
>
> hans
    no, sorry for being unclear: after having two
  ibm drives (75gxp, then 60gxp) which caused troubles 
  he replaced 60 gxp to quantum AS... (quantum AS is
  quite okay now... :)

    i meant that we had these drives simultaneously:
  i bought 75gxp, so he did. he replaced it to 60gxp
  due to failures, so did i. and now he replaced his
  60gxp to quantum-AS. and i still own 60gxp
  (and hoping that noudma will help)...

---


cheers,


   Samium Gromoff

Re: [OT] Howl of soul...

[Hans Reiser]

Samium Gromoff wrote:
> 
>      dear people!
> 
>    Sorry for OT, but i want only good to you, lkml
>  people...
>    If you are to buy a new ide drive, do not buy
>  recent IBM 7200 drives!!!
>    The story begins when at February of this year i`ve bought
>  perfectly shining ever-fast IBM DTLA-307045...
>    After 3 months i`ve hardly regreted about such
>  decision: drive started to covers himself with a thick
>  layer of logical-not-physical badblocks (ie lowlevel
>  reformat doesnt show anything).
>    So i went to storagereview and lerned about the matter.
>  As i found, these drives had an internal controller bug.
> 
>    So i said okay, while restoring 3rd time my reiserfs
>  and dumping data to spare drive, and went to
>  replace the drive to perfectly new and shining
>  IBM IC35L040AVER07-0 also known as 60GXP.
> 
>    Now i`am heavily punished for that.
>  Badblocks are reapperaing on runtime.
>    After they appeared first time i`ve attached large
>  fan to the drive, so it was cold(!) to touch. Also
>  i stopped to transport the drive between boxes.
> 
>    Nevertheless these fscking logical badblocks
>  appeared again twice.
> 
>    The fact is, that we had bought these drives
>  with my friend synchronusly, and now he owns
>  quantum drive, after 75gxp and 60gxp...
>    Ofcourse he had similar problems.... (btw he use windoze)
> 
>    I am _tired_ fixing my poor reiserfs root partition.
>    I can say that now i`am expert on how to restore
>  badblocked reiserfs partiotions... ;(
> 
>    Beware.
> 
> ---
> 
> cheers,
> 
>    Samium Gromoff
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/


I am confused, you and your friend both had problems, and you used IBM and he
used quantum?  Do they have similar electronics?

Take a look at www.namesys.com, there is a badblocks patch, god@namesys.com will
tell you details about it.

hans

Re: [OT] Howl of soul...

[Samium Gromoff]

> Does anyone know what this bug actually is, and whether there's a possible
> workaround without disabling udma entirely?
   so if i will disable udma and switch to 
  -X34 == multiword dma2, then corruptions will go away????

   i have a VX chipset/Zida 5DVX with PIIX3...


---


cheers,


   Samium Gromoff

Re: [OT] Howl of soul...

[Mike Castle]

On Fri, Aug 24, 2001 at 09:00:40PM +0100, Alan Cox wrote:
> There have been multiple reports of problems with drive failures on recent
> IBM drives with lots of OS's.

The following makes me feel all warm and fuzzy:

On Wed, Aug 08, 2001 at 06:12:40PM +0000, Linus Torvalds wrote:
> [ Damn, maybe I should get one of those nice big 7200 rpm IBM drives ]

http://www.uwsg.indiana.edu/hypermail/linux/kernel/0108.1/0120.html
   
mrc
-- 
     Mike Castle      dalgoda@ix.netcom.com      www.netcom.com/~dalgoda/
    We are all of us living in the shadow of Manhattan.  -- Watchmen
fatal ("You are in a maze of twisty compiler features, all different"); -- gcc

Re: [OT] Howl of soul...

[Paul Jakma]

On Fri, 24 Aug 2001, Alan Cox wrote:

> There have been multiple reports of problems with drive failures on recent
> IBM drives with lots of OS's.

hmmm... tell me about it. and not even neccessarily recent.

i had 2 IBM DDRS and 1 DGHS 9GB SCSI drives fail on me all in the
space of a month. All part of one 3disk raid5 array -> 3/3 failure 
rate. 

wasn't bad power, cause the seagates that were in the same box are
still chugging along now, nearly 2 years later. the seagates and 1
quantum that replaced the 3 failed IBMs are also still chugging
along.

> Alan

regards,
-- 
Paul Jakma	paul@clubi.ie	paul@jakma.org	Key ID: 64A2FF6A
Fortune:
You can measure a programmer's perspective by noting his attitude on
the continuing viability of FORTRAN.
		-- Alan Perlis

Re: [OT] Howl of soul...

[David Huen]

On Fri, 24 Aug 2001, Alan Cox wrote:

> 
> There have been multiple reports of problems with drive failures on recent
> IBM drives with lots of OS's.
> 
Frinstance this horror story:-
http://www.hexus.net/article.php?review=209

David H.

Re: [OT] Howl of soul...

[Alan Cox]

> I *think* the i8xx N/Sbridge chipsets all work fine with the drives.  Mine
> do.  A quick search shows complaints with: BX, amd 751, kt 133a, 
> chipsets.

BX and 810 are basically identical IDE. I dont think controller has anything
to do with drive hardware failures

> Does anyone know what this bug actually is, and whether there's a possible
> workaround without disabling udma entirely?

There have been multiple reports of problems with drive failures on recent
IBM drives with lots of OS's.

Alan

Re: [OT] Howl of soul...

[Justin Guyett]

On Fri, 24 Aug 2001, Samium Gromoff wrote:

>    If you are to buy a new ide drive, do not buy
>  recent IBM 7200 drives!!!
<snip>
>    Ofcourse he had similar problems.... (btw he use windoze)

This is a known issue that allegedly has nothing to do with the drive, and
everything to do with a bug in many ide controllers.  I have my 307030
running just fine in windows 2k with an intel i815 board.

I *think* the i8xx N/Sbridge chipsets all work fine with the drives.  Mine
do.  A quick search shows complaints with: BX, amd 751, kt 133a, highpoint
chipsets.

Does anyone know what this bug actually is, and whether there's a possible
workaround without disabling udma entirely?


justin

[OT] Howl of soul...

[Samium Gromoff]

     dear people!

   Sorry for OT, but i want only good to you, lkml
 people...
   If you are to buy a new ide drive, do not buy 
 recent IBM 7200 drives!!!
   The story begins when at February of this year i`ve bought
 perfectly shining ever-fast IBM DTLA-307045...
   After 3 months i`ve hardly regreted about such
 decision: drive started to covers himself with a thick
 layer of logical-not-physical badblocks (ie lowlevel
 reformat doesnt show anything).
   So i went to storagereview and lerned about the matter.
 As i found, these drives had an internal controller bug.

   So i said okay, while restoring 3rd time my reiserfs
 and dumping data to spare drive, and went to
 replace the drive to perfectly new and shining
 IBM IC35L040AVER07-0 also known as 60GXP.

   Now i`am heavily punished for that.
 Badblocks are reapperaing on runtime.
   After they appeared first time i`ve attached large
 fan to the drive, so it was cold(!) to touch. Also
 i stopped to transport the drive between boxes.

   Nevertheless these fscking logical badblocks
 appeared again twice.

   The fact is, that we had bought these drives
 with my friend synchronusly, and now he owns
 quantum drive, after 75gxp and 60gxp...
   Ofcourse he had similar problems.... (btw he use windoze)

   I am _tired_ fixing my poor reiserfs root partition.
   I can say that now i`am expert on how to restore
 badblocked reiserfs partiotions... ;(

   Beware.

---


cheers,


   Samium Gromoff