Re: got "swap_pager: indefinite wait buffe" under -current

On Wed, 10 Mar 2004, Scott Long wrote:

> Date: Wed, 10 Mar 2004 02:42:05 -0700
> From: Scott Long <scottl_at_freebsd.org>
> To: Poul-Henning Kamp <phk_at_phk.freebsd.dk>
> Cc: Michael Reifenberger <mike_at_Reifenberger.com>, freebsd-current_at_freebsd.org
> Subject: Re: got "swap_pager: indefinite wait buffe" under -current
>
> Poul-Henning Kamp wrote:
> > In message <20040310103020.Y69923_at_fw.reifenberger.com>, Michael Reifenberger wr
> > ites:
> >
> >>Hi,
> >>under -current as of yesterday I got during the fixation of an audio CD
> >>using cdrecord via atapicam the following console message:
> >>
> >>swap_pager: indefinite wait buffer: device: ad0s2a, blkno: 34107, size: 16384
> >>
> >>What does that mean?
> >
> >
> > That a disk-I/O request did not complete.
> >
> > Likely cause is disk hardware or media error.
> >
> > Check for console messages from ata driver.
> >

No, no ata failures.
Just two-times the above-statet failure.

> > You can try to run gstat(8) the leftmost column which shows number
> > of outstanding I/O requests.
> >
> >
>
> Actually, this is more likely due to bad error recovery in ATAPI causing
> all of the other I/O to be held hostage.  It's hard to say if this is
> purely the fault of the ata driver, or if there is a more insideous
> interaction between ata and GEOM.  In any case, you might try using
> burncd(8) to see if the problem persists.
>

Yes, burncd did survive.

Bye/2
---
Michael Reifenberger, Business Development Manager SAP-Basis, Plaut Consulting
Comp: Michael.Reifenberger_at_plaut.de | Priv: Michael_at_Reifenberger.com
      http://www.plaut.de           |       http://www.Reifenberger.com