Heavy fs corruption with 9.0-RELEASE

[David Thiel]

On Fri, Mar 09, 2012 at 01:53:59AM -0800, Adrian Chadd wrote:
> PR ?

The original thread was here:

http://marc.info/?t=132502460000001&r=1&w=2

Ignore the part where it takes me a while to figure out fsck isn't 
softupdate-aware.

There was further followup off-list with the associated SUJ developers 
with clearer test results, but no definitive resolution as of yet, as 
far as I know. I didn't get a chance to test the write cache disabling 
approach or do further testing before I had to turn off SUJ.

Arnaud: Would you be able to record a few test cases with SUJfsck and 
then regular fsck, both using the -v flag (with output piped to some 
other fs)? Can you also see if:

kern.cam.ada.write_cache=0
hw.ata.wc=0

while leaving SUJ enabled helps at all?

> On 8 March 2012 11:07, David Thiel <lx at freebsd.org> wrote:
> > On Mon, Mar 05, 2012 at 05:12:16PM -0500, Arnaud Lacombe wrote:
> >> I've been running a couple of system with 9.0-RELEASE since it is out.
> >> All the system were installed through the standard installation
> >> procedure. After unclean reboot, either crash or power-failure, I get
> >> a huge amount of really bad filesystem corruption (read: "silent",
> >> fs-wide, corruptions). This happens with either i386 or amd64 build.
> >> Systems involved use compact flash as their system permanent storage
> >> medium.
> >
> > I have had this same behavior on every SUJ system I've built, both on
> > SSDs and otherwise, on i386, PPC and amd64. Remove SUJ and revert to
> > plain softupdates, and I strongly suspect your problems will disappear.
> >
> > _______________________________________________
> > freebsd-stable at freebsd.org mailing list
> > http://lists.freebsd.org/mailman/listinfo/freebsd-stable
> > To unsubscribe, send any mail to "freebsd-stable-unsubscribe at freebsd.org"