All of lore.kernel.org
 help / color / mirror / Atom feed
* Invalidate: busy buffer + MD RAID 1
@ 2002-12-18 18:45 Igmar Palsenberg
  2002-12-19 15:35 ` Teodor Iacob
  0 siblings, 1 reply; 3+ messages in thread
From: Igmar Palsenberg @ 2002-12-18 18:45 UTC (permalink / raw)
  To: linux-kernel


Hi,

I get a 'invalidate: busy buffer' about 20 times at reboot. Only at 
reboot however.

Setup :

linux-2.4.19 + grsecurity-1.9.7d + acl+xattr 0.8.53 + freeswan (inc. aes 
and that kind of stuff)

The machine (Compaq ML350) has 2 scsi devices (sda, sdb) and a RAID 1 
setup :

md0 : sda1 + sdb1
md1 : sda3 + sdb3
swap is done on sda2 + sdb2, using default prio's.

Triggering the 'invalidate: busy buffer' is easiest done by letting squid 
create it's cache dirs and then rebooting.

No data corruption is occuring (at last not any that a force fsck can 
detect), but I removed the RAID1 setup to make sure I sleep well tonight 
:)

Looking at the md.c code, line 1708 :

    ITERATE_RDEV(mddev,rdev,tmp) {
        if (rdev->faulty)
            continue;
        invalidate_device(rdev->dev, 1);
        if (get_hardsect_size(rdev->dev)
            > md_hardsect_sizes[mdidx(mddev)])
            md_hardsect_sizes[mdidx(mddev)] =
                get_hardsect_size(rdev->dev);
    }

Looks like it is invalidating the underlying devices (sda[13], sdb[13] in 
my case.

Since my RAID array doesn't get screwed I suspect that the md code does 
the above again on a do_md_stop(), but I can't find it.

Anyone got any comments on this ?? 



	Regards,


		Igmar


Please CC all responses.



-- 

Igmar Palsenberg
JDI Media Solutions

Helhoek 30
6923PE Groessen
Tel: +31 (0)316 - 596695
Fax: +31 (0)316 - 596699
The Netherlands

mailto: i.palsenberg@jdimedia.nl
PGP/GPG key : http://www.jdimedia.nl/formulier/pgp/igmar


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: Invalidate: busy buffer + MD RAID 1
  2002-12-18 18:45 Invalidate: busy buffer + MD RAID 1 Igmar Palsenberg
@ 2002-12-19 15:35 ` Teodor Iacob
  2002-12-22 16:40   ` Igmar Palsenberg
  0 siblings, 1 reply; 3+ messages in thread
From: Teodor Iacob @ 2002-12-19 15:35 UTC (permalink / raw)
  To: Igmar Palsenberg; +Cc: linux-kernel

I get the same behaviour on several machines .. but I sleep well at night :P

On Wed, Dec 18, 2002 at 07:45:00PM +0100, Igmar Palsenberg wrote:
> 
> Hi,
> 
> I get a 'invalidate: busy buffer' about 20 times at reboot. Only at 
> reboot however.
> 
> Setup :
> 
> linux-2.4.19 + grsecurity-1.9.7d + acl+xattr 0.8.53 + freeswan (inc. aes 
> and that kind of stuff)
> 
> The machine (Compaq ML350) has 2 scsi devices (sda, sdb) and a RAID 1 
> setup :
> 
> md0 : sda1 + sdb1
> md1 : sda3 + sdb3
> swap is done on sda2 + sdb2, using default prio's.
> 
> Triggering the 'invalidate: busy buffer' is easiest done by letting squid 
> create it's cache dirs and then rebooting.
> 
> No data corruption is occuring (at last not any that a force fsck can 
> detect), but I removed the RAID1 setup to make sure I sleep well tonight 
> :)
> 
> Looking at the md.c code, line 1708 :
> 
>     ITERATE_RDEV(mddev,rdev,tmp) {
>         if (rdev->faulty)
>             continue;
>         invalidate_device(rdev->dev, 1);
>         if (get_hardsect_size(rdev->dev)
>             > md_hardsect_sizes[mdidx(mddev)])
>             md_hardsect_sizes[mdidx(mddev)] =
>                 get_hardsect_size(rdev->dev);
>     }
> 
> Looks like it is invalidating the underlying devices (sda[13], sdb[13] in 
> my case.
> 
> Since my RAID array doesn't get screwed I suspect that the md code does 
> the above again on a do_md_stop(), but I can't find it.
> 
> Anyone got any comments on this ?? 
> 
> 
> 
> 	Regards,
> 
> 
> 		Igmar
> 
> 
> Please CC all responses.
> 
> 
> 
> -- 
> 
> Igmar Palsenberg
> JDI Media Solutions
> 
> Helhoek 30
> 6923PE Groessen
> Tel: +31 (0)316 - 596695
> Fax: +31 (0)316 - 596699
> The Netherlands
> 
> mailto: i.palsenberg@jdimedia.nl
> PGP/GPG key : http://www.jdimedia.nl/formulier/pgp/igmar
> 
> -
> 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/

-- 
      Teodor Iacob,
Network Administrator
Astral TELECOM Internet

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: Invalidate: busy buffer + MD RAID 1
  2002-12-19 15:35 ` Teodor Iacob
@ 2002-12-22 16:40   ` Igmar Palsenberg
  0 siblings, 0 replies; 3+ messages in thread
From: Igmar Palsenberg @ 2002-12-22 16:40 UTC (permalink / raw)
  To: Teodor Iacob; +Cc: linux-kernel


Hi,

> I get the same behaviour on several machines .. but I sleep well at night :P

It didn't eat your FS ?? I'll probably remove the message from the kernel 
source, it confuses people around here.



	Igmar

-- 

Igmar Palsenberg
JDI Media Solutions

Helhoek 30
6923PE Groessen
Tel: +31 (0)316 - 596695
Fax: +31 (0)316 - 596699
The Netherlands

mailto: i.palsenberg@jdimedia.nl
PGP/GPG key : http://www.jdimedia.nl/formulier/pgp/igmar


^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2002-12-22 16:32 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-12-18 18:45 Invalidate: busy buffer + MD RAID 1 Igmar Palsenberg
2002-12-19 15:35 ` Teodor Iacob
2002-12-22 16:40   ` Igmar Palsenberg

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.