Large Storage Devices in Linux.....Kernel level support.....

Large Storage Devices in Linux.....Kernel level support.....

[Cyrus]

hi all,

i've got a setup of 2 hard drives (30GB & 40GB) with an Asus a7m266 mobo 
with a VIA Technologies, Inc. VT82C586 IDE [Apollo] (rev 06).

30GB= fujitsu, 40GB= IBM (both are 7200rpm

i've got my cdrw on /dev/hdc, 30GB=/dev/hda, and 40GB=/dev/hdb...

all works alright for a while, but when i keep my computer turned on for 
a couple of days and then reboot. bios sometimes tells me that smart 
array (or something) failed with my primary master (30GB) and i should 
back-up soon.. next reboot it tells me that pri-master fails.. it's 
doing this quite regularly and i don't know how to stop it. i'm running 
kernel 2.4.12 vanilla.

i'm sorry guys if this is the wrong place to ask but i could use some 
hints in probably adding some options to the kernel at compile(config) 
time to aid my ailing box.

Cyrus Santos

Registered Slackware Linux User # 220455
Sydney, Australia

"...the best things in life are free...."

Re: Large Storage Devices in Linux.....Kernel level support.....

[Mike Fedyk]

On Sun, Oct 14, 2001 at 05:24:59PM +1000, Cyrus wrote:
> hi all,
> 
> i've got a setup of 2 hard drives (30GB & 40GB) with an Asus a7m266 mobo 
> with a VIA Technologies, Inc. VT82C586 IDE [Apollo] (rev 06).
> 
> 30GB= fujitsu, 40GB= IBM (both are 7200rpm
> 
> i've got my cdrw on /dev/hdc, 30GB=/dev/hda, and 40GB=/dev/hdb...
> 
> all works alright for a while, but when i keep my computer turned on for 
> a couple of days and then reboot. bios sometimes tells me that smart 
> array (or something) failed with my primary master (30GB) and i should 
> back-up soon.. next reboot it tells me that pri-master fails.. it's 
> doing this quite regularly and i don't know how to stop it. i'm running 

Turn off "S.M.A.R.T." in your bios...  Probably under the advanced bios
config menu.

I know that Compaq has a SMART RAID controller, but does anyone know what
this does?  (I've seen it on old p2 MBs and they didn't have raid...)

Re: Large Storage Devices in Linux.....Kernel level support.....

[Kip Macy]

S.M.A.R.T. and SMART are not the same thing. A quick google search got me
to this page:
 
<http://www.belarc.com/Images/blank.gif>
[About S.M.A.R.T.]
S.M.A.R.T. (Self-Monitoring Analysis and Reporting Technology) is a
diagnostic method originally developed by I.B.M. for their mainframe
drives to give advanced warning of drive failures. Large mainframe data
centers wanted to know in advance if a hard disk drive was going to fail,
because this gave them the opportunity to take steps to protect their
data. Later Compaq announced a diagnostic which operated with a number of
different disk drive manufacturers. These products were submitted to the
ATA/IDE standards committees and the resulting standard was named
S.M.A.R.T. Today all major hard disk drive manufacturers support
S.M.A.R.T., including IBM, Western Digital, Quantum, Seagate, and Fujitsu.
etc.


On Sun, 14 Oct 2001, Mike Fedyk wrote:

> On Sun, Oct 14, 2001 at 05:24:59PM +1000, Cyrus wrote:
> > hi all,
> > 
> > i've got a setup of 2 hard drives (30GB & 40GB) with an Asus a7m266 mobo 
> > with a VIA Technologies, Inc. VT82C586 IDE [Apollo] (rev 06).
> > 
> > 30GB= fujitsu, 40GB= IBM (both are 7200rpm
> > 
> > i've got my cdrw on /dev/hdc, 30GB=/dev/hda, and 40GB=/dev/hdb...
> > 
> > all works alright for a while, but when i keep my computer turned on for 
> > a couple of days and then reboot. bios sometimes tells me that smart 
> > array (or something) failed with my primary master (30GB) and i should 
> > back-up soon.. next reboot it tells me that pri-master fails.. it's 
> > doing this quite regularly and i don't know how to stop it. i'm running 
> 
> Turn off "S.M.A.R.T." in your bios...  Probably under the advanced bios
> config menu.
> 
> I know that Compaq has a SMART RAID controller, but does anyone know what
> this does?  (I've seen it on old p2 MBs and they didn't have raid...)
> -
> 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/
> 

Re: Large Storage Devices in Linux.....Kernel level support.....

[Kip Macy]

On second thought they most likely are the same thing. And, just in case
it is not obvious, it is telling you to back up your data because it
thinks the drive is about to fail.

				-Kip

On Sun, 14 Oct 2001, Kip Macy wrote:

> S.M.A.R.T. and SMART are not the same thing. A quick google search got me
> to this page:
>  
> <http://www.belarc.com/Images/blank.gif>
> [About S.M.A.R.T.]
> S.M.A.R.T. (Self-Monitoring Analysis and Reporting Technology) is a
> diagnostic method originally developed by I.B.M. for their mainframe
> drives to give advanced warning of drive failures. Large mainframe data
> centers wanted to know in advance if a hard disk drive was going to fail,
> because this gave them the opportunity to take steps to protect their
> data. Later Compaq announced a diagnostic which operated with a number of
> different disk drive manufacturers. These products were submitted to the
> ATA/IDE standards committees and the resulting standard was named
> S.M.A.R.T. Today all major hard disk drive manufacturers support
> S.M.A.R.T., including IBM, Western Digital, Quantum, Seagate, and Fujitsu.
> etc.
> 
> 
> On Sun, 14 Oct 2001, Mike Fedyk wrote:
> 
> > On Sun, Oct 14, 2001 at 05:24:59PM +1000, Cyrus wrote:
> > > hi all,
> > > 
> > > i've got a setup of 2 hard drives (30GB & 40GB) with an Asus a7m266 mobo 
> > > with a VIA Technologies, Inc. VT82C586 IDE [Apollo] (rev 06).
> > > 
> > > 30GB= fujitsu, 40GB= IBM (both are 7200rpm
> > > 
> > > i've got my cdrw on /dev/hdc, 30GB=/dev/hda, and 40GB=/dev/hdb...
> > > 
> > > all works alright for a while, but when i keep my computer turned on for 
> > > a couple of days and then reboot. bios sometimes tells me that smart 
> > > array (or something) failed with my primary master (30GB) and i should 
> > > back-up soon.. next reboot it tells me that pri-master fails.. it's 
> > > doing this quite regularly and i don't know how to stop it. i'm running 
> > 
> > Turn off "S.M.A.R.T." in your bios...  Probably under the advanced bios
> > config menu.
> > 
> > I know that Compaq has a SMART RAID controller, but does anyone know what
> > this does?  (I've seen it on old p2 MBs and they didn't have raid...)
> > -
> > 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/
> > 
> 
> -
> 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/
> 

Re: Large Storage Devices in Linux.....Kernel level support.....

[Mike Fedyk]

> > On Sun, 14 Oct 2001, Mike Fedyk wrote:
> > 
> > > On Sun, Oct 14, 2001 at 05:24:59PM +1000, Cyrus wrote:
> > > > hi all,
> > > > 
> > > > i've got a setup of 2 hard drives (30GB & 40GB) with an Asus a7m266 mobo 
> > > > with a VIA Technologies, Inc. VT82C586 IDE [Apollo] (rev 06).
> > > > 
> > > > 30GB= fujitsu, 40GB= IBM (both are 7200rpm
> > > > 
> > > > i've got my cdrw on /dev/hdc, 30GB=/dev/hda, and 40GB=/dev/hdb...
> > > > 
> > > > all works alright for a while, but when i keep my computer turned on for 
> > > > a couple of days and then reboot. bios sometimes tells me that smart 
> > > > array (or something) failed with my primary master (30GB) and i should 
> > > > back-up soon.. next reboot it tells me that pri-master fails.. it's 
> > > > doing this quite regularly and i don't know how to stop it. i'm running 
> > > 
> > > Turn off "S.M.A.R.T." in your bios...  Probably under the advanced bios
> > > config menu.
> > > 
> > > I know that Compaq has a SMART RAID controller, but does anyone know what
> > > this does?  (I've seen it on old p2 MBs and they didn't have raid...)
> > > -
> > > 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/
> > > 
> > 

> On Sun, 14 Oct 2001, Kip Macy wrote:
> 
> > S.M.A.R.T. and SMART are not the same thing. A quick google search got me
> > to this page:
> >  
> > <http://www.belarc.com/Images/blank.gif>
> > [About S.M.A.R.T.]
> > S.M.A.R.T. (Self-Monitoring Analysis and Reporting Technology) is a
> > diagnostic method originally developed by I.B.M. for their mainframe
> > drives to give advanced warning of drive failures. Large mainframe data
> > centers wanted to know in advance if a hard disk drive was going to fail,
> > because this gave them the opportunity to take steps to protect their
> > data. Later Compaq announced a diagnostic which operated with a number of
> > different disk drive manufacturers. These products were submitted to the
> > ATA/IDE standards committees and the resulting standard was named
> > S.M.A.R.T. Today all major hard disk drive manufacturers support
> > S.M.A.R.T., including IBM, Western Digital, Quantum, Seagate, and Fujitsu.
> > etc.

On Sun, Oct 14, 2001 at 08:08:35PM -0700, Kip Macy wrote:
> On second thought they most likely are the same thing. And, just in case
> it is not obvious, it is telling you to back up your data because it
> thinks the drive is about to fail.
> 
> 				-Kip

Yes, some BIOSes have "S.M.A.R.T." and some "SMART".

Thanks for the history lesson and research...

Cyrus, replace that drive, or at least test it with badblocks.  One thing
that I've noticed about badblocks is that there are times when part of a
drive is about to go bad, but doesn't give an error message.  It just takes
longer, and uses several reads but finally completes successfully.  The
only time I've noticed this was on
a drive where badblocks *did* find an error, but there were parts that took
much too long to read, and probably should've been marked faulty.  (This was
with badblocks in e2fsprogs 1.18).  Though, taking a long time to read could
be caused by other users on a multi-user system, but if there is a way to
detect if a drive is retrying to read, that would be a good way to check...

BTW Cyrus, your mail bounced, so if you take the time to read the archive,
you'll know...

----- The following addresses had permanent fatal errors -----  
<cyrus@linuxmail.org>                                               
    (reason: 521 sorry, user inactive)         
       
Mike

Re: Large Storage Devices in Linux.....Kernel level support.....

[Mike Fedyk]

On Mon, Oct 15, 2001 at 07:08:50PM +1000, Cyrus wrote:
>    hi,
> 
>    thanks for the reply guys. this hdrive is actually quite new i'm wondering
>    what it's doing.. it doesn't warn me anymore of that Smart thing... :-)
>     but, i noticed the longer my system is turned on X just slows down a
>    bit... like moving my mouse to a certain position takes ages. as in, very
>    less responsive.

This could mean that the system is trying to access one drive harder, and
not letting the other respond to requests.  This is even more reason to test
the drive...  While you're at it, test both drives.

>I just got this drives about a week ago the IBM 40GB one.
>    the fujitsu 30GB one is probably just three weeks now. but they're all
>    brand new.... i don't know really, i'm quite confused. i don't know how to
>    prove to that guy in the shop that his drives are faulty.... anyway, so do
>    you think Mike that replacing it would be the only solution?

read-only badblocks will tell you if you have a drive with sectors that are
unreadable.  read-write badblocks will tell you if there are any sectors
that are unwritable.  Be careful, older versions of badblocks will erase all
of your data.  Newer versions have a mode to preserve the data, and still do
a write test, while still keeping the other write mode.  Read the manual page.

>yeah, by the
>    way, i'm using reiserfs how can i check this harddrive for bad blocks i
>    think the program you suggested was for ext2, or was it?
>

Badblocks doesn't care what is on the drive, it just deals with block
devices, that means hard drives, floppy, etc.

I don't know if the new block device in page cache in 2.4.10+ will affect
this, but it looks plausible.  To be sure, use a -ac kernel or 2.4.9 or
older...

>    thanks a lot!
> 
>    cyrus
> 
>    by the way this is my real email address :-) .. i think linuxmail.org
>    closed my account due to inactivity.
> 
>    cheers!

Make sure that you cc the linux-kernel list...

Mike

Re: Large Storage Devices in Linux.....Kernel level support.....

[Cyrus]

Cyrus wrote:

 > hi all,
 >
 > i've got a setup of 2 hard drives (30GB & 40GB) with an Asus a7m266 mobo
 > with a VIA Technologies, Inc. VT82C586 IDE [Apollo] (rev 06).
 >
 > 30GB= fujitsu, 40GB= IBM (both are 7200rpm
 >
 > i've got my cdrw on /dev/hdc, 30GB=/dev/hda, and 40GB=/dev/hdb...
 >
 > all works alright for a while, but when i keep my computer turned on for
 > a couple of days and then reboot. bios sometimes tells me that smart
 > array (or something) failed with my primary master (30GB) and i should
 > back-up soon.. next reboot it tells me that pri-master fails.. it's
 > doing this quite regularly and i don't know how to stop it. i'm running
 > kernel 2.4.12 vanilla.
 >
 > i'm sorry guys if this is the wrong place to ask but i could use some
 > hints in probably adding some options to the kernel at compile(config)
 > time to aid my ailing box.
 >
 > Cyrus Santos
 >
 > Registered Slackware Linux User # 220455
 > Sydney, Australia
 >
 > "...the best things in life are free...."
 >
 >
 >

hi all,

thanks for the feedback! i'll try all those recommendations now. by the
way i tried reiserfsck --check /dev/hda2, etc.... and it seems that
everything is ok... i suspect it's either the power supply or the
cooling... power supply because, i've got a 300W one but, almost all the
jumper power cables are used... i've got 2 chassis fans one for ducting
are out at the back below the power supply and one in front to drive the
air in.. i've separated my hard drives now they're not on top of each
other anymore... i'll try connecting the other hdisk to another power
cable and try if it does work.... i'll give apmd a go as well...

thanks guys!

btw, this is my real email address now Mike... cheers!


cyrus

-- 


Cyrus Santos

Registered Linux User # 220455
Sydney, Australia

"...the best things in life are free...."

Re: Large Storage Devices in Linux.....Kernel level support.....

[Tim Moore]

>  > i've got a setup of 2 hard drives (30GB & 40GB) with an Asus a7m266 mobo
>  > with a VIA Technologies, Inc. VT82C586 IDE [Apollo] (rev 06).
>  >
>  > 30GB= fujitsu, 40GB= IBM (both are 7200rpm
>  >
>  > i've got my cdrw on /dev/hdc, 30GB=/dev/hda, and 40GB=/dev/hdb...

You also want to move the IBM to IDE1 to avoid 1/2 speed throttle
situations.  Don't forget to update /dev/fstab and the /dev/cdrom and/or
/dev/cdr links before you reboot.

Current
-------
IDE0/M	/dev/hda	30GB
IDE0/S	/dev/hdb	40GB
IDE1/M	/dev/hdc	CD-RW

Better disk I/O
---------------
IDE0/M	/dev/hda	30GB
IDE1/M	/dev/hdc	40GB
IDE1/S	/dev/hdd	CD-RW

	or

IDE0/M	/dev/hda	30GB
IDE0/S	/dev/hdd	CD-RW
IDE1/M	/dev/hdc	40GB


rgds,
tim.
--