sync hanging

sync hanging

[Corin Hartland-Swann]

Hi there,

I'm using kernel 2.4.8-ac2 on a Dual PIII-1000 with 4096M RAM, and a
reiserfs filesystem on a RAID-1 mirror of two 76GB UDMA disks, and I'm
experiencing a strange problem after the machine has been running for a
while.

Every now and again, running sync(1) (i.e. the program) seems to hang and
end up in state D (uninterruptible sleep). There is no way to kill it
(even with SIGKILL but I assume that this is typical for state D
processes.

Does anyone have any idea what may be causing this? I have searched the
archives and couldn't find anything similar.

Thanks,

Corin

PS: Please CC any replies to me.

/------------------------+-------------------------------------\
| Corin Hartland-Swann   |    Tel: +44 (0) 20 7491 2000        |
| Commerce Internet Ltd  |    Fax: +44 (0) 20 7491 2010        |
| 22 Cavendish Buildings | Mobile: +44 (0) 79 5854 0027        | 
| Gilbert Street         |                                     |
| Mayfair                |    Web: http://www.commerce.uk.net/ |
| London W1K 5HJ         | E-Mail: cdhs@commerce.uk.net        |
\------------------------+-------------------------------------/

Re: sync hanging

[Hans Reiser]

Corin Hartland-Swann wrote:
> 
> Hi there,
> 
> I'm using kernel 2.4.8-ac2 on a Dual PIII-1000 with 4096M RAM, and a
> reiserfs filesystem on a RAID-1 mirror of two 76GB UDMA disks, and I'm
> experiencing a strange problem after the machine has been running for a
> while.
> 
> Every now and again, running sync(1) (i.e. the program) seems to hang and
> end up in state D (uninterruptible sleep). There is no way to kill it
> (even with SIGKILL but I assume that this is typical for state D
> processes.
> 
> Does anyone have any idea what may be causing this? I have searched the
> archives and couldn't find anything similar.
> 
> Thanks,
> 
> Corin
> 
> PS: Please CC any replies to me.
> 
> /------------------------+-------------------------------------\
> | Corin Hartland-Swann   |    Tel: +44 (0) 20 7491 2000        |
> | Commerce Internet Ltd  |    Fax: +44 (0) 20 7491 2010        |
> | 22 Cavendish Buildings | Mobile: +44 (0) 79 5854 0027        |
> | Gilbert Street         |                                     |
> | Mayfair                |    Web: http://www.commerce.uk.net/ |
> | London W1K 5HJ         | E-Mail: cdhs@commerce.uk.net        |
> \------------------------+-------------------------------------/
> 
> -
> 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/
turn off highmem, known bug, I don't know if it is solved yet.

Hans

Re: sync hanging

[Corin Hartland-Swann]

Hi Hans,

On Tue, 21 Aug 2001, Hans Reiser wrote:
> Corin Hartland-Swann wrote:
> > I'm using kernel 2.4.8-ac2 on a Dual PIII-1000 with 4096M RAM, and a
> > reiserfs filesystem on a RAID-1 mirror of two 76GB UDMA disks, and I'm
> > experiencing a strange problem after the machine has been running for a
> > while.
> > 
> > Every now and again, running sync(1) (i.e. the program) seems to hang and
> > end up in state D (uninterruptible sleep). There is no way to kill it
> > (even with SIGKILL but I assume that this is typical for state D
> > processes.
> > 
> > Does anyone have any idea what may be causing this? I have searched the
> > archives and couldn't find anything similar.
>
> turn off highmem, known bug, I don't know if it is solved yet.

Is 2G RAM OK (by manually changing __PAGE_OFFSET ?)

Is this directly related to reiserfs (i.e. will it go away if I go back to
using ext2 or switch to xfs?)

How serious is the problem? Can I get away with using it as is (with
highmem) on a production server or will that cause serious repercussions?

Thanks,

Corin

/------------------------+-------------------------------------\
| Corin Hartland-Swann   |    Tel: +44 (0) 20 7491 2000        |
| Commerce Internet Ltd  |    Fax: +44 (0) 20 7491 2010        |
| 22 Cavendish Buildings | Mobile: +44 (0) 79 5854 0027        | 
| Gilbert Street         |                                     |
| Mayfair                |    Web: http://www.commerce.uk.net/ |
| London W1K 5HJ         | E-Mail: cdhs@commerce.uk.net        |
\------------------------+-------------------------------------/

Re: sync hanging

[Hans Reiser]

Corin Hartland-Swann wrote:
> 
> Hi Hans,
> 
> On Tue, 21 Aug 2001, Hans Reiser wrote:
> > Corin Hartland-Swann wrote:
> > > I'm using kernel 2.4.8-ac2 on a Dual PIII-1000 with 4096M RAM, and a
> > > reiserfs filesystem on a RAID-1 mirror of two 76GB UDMA disks, and I'm
> > > experiencing a strange problem after the machine has been running for a
> > > while.
> > >
> > > Every now and again, running sync(1) (i.e. the program) seems to hang and
> > > end up in state D (uninterruptible sleep). There is no way to kill it
> > > (even with SIGKILL but I assume that this is typical for state D
> > > processes.
> > >
> > > Does anyone have any idea what may be causing this? I have searched the
> > > archives and couldn't find anything similar.
> >
> > turn off highmem, known bug, I don't know if it is solved yet.
> 
> Is 2G RAM OK (by manually changing __PAGE_OFFSET ?)
> 

surf through lkm postings for more info on this bug, it is not in my area of
expertise.

> Is this directly related to reiserfs (i.e. will it go away if I go back to
> using ext2 or switch to xfs?)

it is completely unrelated to reiserfs.

> 
> How serious is the problem? Can I get away with using it as is (with
> highmem) on a production server or will that cause serious repercussions?

you will be screwed if you use it on a production server, as best I understand
it.

> 
> Thanks,
> 
> Corin
> 
> /------------------------+-------------------------------------\
> | Corin Hartland-Swann   |    Tel: +44 (0) 20 7491 2000        |
> | Commerce Internet Ltd  |    Fax: +44 (0) 20 7491 2010        |
> | 22 Cavendish Buildings | Mobile: +44 (0) 79 5854 0027        |
> | Gilbert Street         |                                     |
> | Mayfair                |    Web: http://www.commerce.uk.net/ |
> | London W1K 5HJ         | E-Mail: cdhs@commerce.uk.net        |
> \------------------------+-------------------------------------/

Re: sync hanging

[Alan Cox]

> Every now and again, running sync(1) (i.e. the program) seems to hang and
> end up in state D (uninterruptible sleep). There is no way to kill it
> (even with SIGKILL but I assume that this is typical for state D
> processes.

Can you replicate this in 2.4.8ac8, and if so where is it sleeping according
to ps wait queue data

I've had a couple of similar reports against 2.4.8 and early 8ac. I'm
curious if its still a problem as Im not aware of any reason it should have
been fixed by the vm fixups