Date: Fri, 24 Aug 2007 01:07:51 +1000 (EST) From: Ian Smith <smithi@nimnet.asn.au> To: Karol Kwiatkowski <karol.kwiat@gmail.com> Cc: Chris <chrcoluk@gmail.com>, freebsd-questions@freebsd.org, Bill Moran <wmoran@potentialtech.com> Subject: Re: fsck strangeness Message-ID: <Pine.BSF.3.96.1070823202242.26941E-100000@gaia.nimnet.asn.au> In-Reply-To: <46CD5295.4050406@gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 23 Aug 2007, Karol Kwiatkowski wrote:
> Ian Smith wrote:
> > On Thu, 23 Aug 2007, Karol Kwiatkowski wrote:
> > > Ian Smith wrote:
> > > > On Wed, 22 Aug 2007, Chris wrote:
> > > > > If its bad to run fsck on a mounted read,write then why does
> > > > > background fsck do it? or you talking about foreground fsck only?
> > > >
> > > > Well I was referring to foreground fsck, and I still don't know why
> > > > running it on a mounted fs is 'bad' when fsck runs in 'NO WRITE' mode
> > > > anyway when it finds a fs is mounted, hence my query above.
> > >
> > > Here's my understanding:
> > >
> > > Mounted fs (rw) isn't in stable state, there may be some writes to it -
> > > daemons, buffers flushes, etc. In this condition fsck can report
> > > inconsistency. And fsck running in 'NO WRITE' won't help anyway :)
> >
> > a) Absolutely.
> >
> > b) Indeed it usually does, fairly consistently, especially on /var.
> >
> > c) No it won't help (except where it can help locate problems in a real
> > mess like bad blocks), but the assertion in question was, can it hurt?
>
> Ah sorry, I missed that. With 'NO WRITE' one can suppose it shouldn't
> hurt anything except performance ;)
>
> I made a quick scan through the source and it looks like it won't:
> - in src/sbin/fsck_ffs/setup.c
> if fs is mounted rw fswritefd is set to -1
Ah, the source, who would have thought .. so it does ..
if (bkgrdflag == 0 &&
(nflag || (fswritefd = open(dev, O_WRONLY)) < 0)) {
fswritefd = -1;
if (preen)
pfatal("NO WRITE ACCESS");
printf(" (NO WRITE)");
}
.. which explore answered the flip side of my query, I think: fsck (in
fg mode) _will_ update an fs mounted readonly, unless -n is specified.
> - in src/sbin/fsck_ffs/fsutil.c
> blwrite(), flush() and ckfini() won't write anything if fswritefd<0
>
> Unless, of course, I'm missing something.
I'll keep using -n to be sure for 'casual' fsck, never failed me, and
one day I may figure out how bg fsck works. Thanks for the tute :)
Cheers, Ian
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.96.1070823202242.26941E-100000>