[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: consensus on softupdates


On Wed, Apr 04, 2001 at 03:06:30PM +1000, Christopher Biggs wrote:
> I later activated soft updates on a few test systems (2.8-stable) I
> had installed, to see what would happen.
> <several weeks pass>
> Some of the systems had since been cvs-updated to -current, and much
> compiling and other file-tossing had occured with *never* a problem.

... similar story here ..., but:

> Then there was a power failure.  <ominous chord>

I regularly reboot because of updating to -current w/o problems as
of the last few months, AND

> I have never *seen* such horribly scrambled filesystems.  Mulitple
> passes of fsck just kept finding more and *more* errors; duplicated
> inodes and horribly mangled directories.   Fsck would declare it
> needed to be re-run, then find a bunch more problems.  Rinse, repeat.

had one machine crash due to hardware failure (power supply :( ) hard
in my absence and once while doing a make world. Both times the machine
came back up sane. OTOH on a 2.8 release and shortly thereafter I had
severe problems which forced me to newfs the system.

> \---------veni vidi nuclei deceiri --- I came, I saw, I dumped core----------/


Best Regards,