Recently I've published the summary I'd like to correct.
In my original message I wrote that in trying to mount another root file
system to /mnt I got "Dirty file system" even if that system was accurately
shutdowned before.
The thing to correct is that in my comment (in the summary) I didn't attach
all due importance to the KEY fact mentioned in the response by Alan Rollow
and quoted in the summary (and pointed also by Marco Luchini), that was:
>The root can never be unmounted cleanly, so it always has to be checked.
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Besides, I'd like to note - if I'm allowed - that in my opinion this fact
would be worth mentioning in the example of mounting another root file
system to /mnt in "Installation Guide", Section 7.6, Mounting File Systems.
Sorry for being not accurate enough.
The previous summary and original posting follow my signature.
Thanks,
Irene
*************************************************************************
* *
* Irene A. Shilikhina e-mail: irene_at_alpha.iae.nsk.su *
* System administrator, *
* Institute of Automation & Electrometry, *
* Siberian Branch of Russian Academy of Sciences, *
* Novosibirsk, Russia *
*
http://www.iae.nsk.su/~irene *
*************************************************************************
* * *
* The road to hell is paved with * Every cloud has a silver lining. *
* good intentions. * *
* * *
*************************************************************************
On Thu, 6 May 1999, Irene A. Shilikhina wrote:
>
> Hello Managers,
>
> yesterday I wrote that my attempt in mounting file systems created under
> another version of the system had failed (with the message "Dirty file system").
>
> Thanks to everybody who replied:
>
> Olle Eriksson <olle_at_cb.uu.se>
> Udo Grabowski <grabow_at_imk.fzk.de>
> Cy Dewhurst <cy.dewhurst_at_rbch-tr.swest.nhs.uk>
> Dr Marco Luchini <m.luchini_at_ic.ac.uk>
> MC.Vialatte_at_custsv.univ-bpclermont.fr
> "David A. Massaro - Unix, Linux, VMS, TCP/IP..."
> <MASSARDA_at_mail.suny.edu>
> Sysadmin <sysadmin_at_tatatel.co.in>
> Gyula Szokoly <szgyula_at_tarkus.pha.jhu.edu>
> "Alan Rollow - Dr. File System's Home for Wayward Inodes."
> <alan_at_nabeth.cxo.dec.com>
> "Picard, Roger" <Roger.Picard_at_compaq.com>
> "Dr. Tom Blinn, 603-884-0646" <tpb_at_doctor.zk3.dec.com>
> Jim Bostwick <Jim_Bostwick_at_cargill.com>
>
> The answer by Dr. Tom Blinn deserves to be awarded a prise for laconism:
> ***********************************
> * Read the "fsck" reference page. *
> ***********************************
>
> My special thanks to Alan Rollow for an extra information:
>
> >Run fsck on the file system before trying to mount it. When a file
> >system is mounted on any version, a field is changed in the super-
> >block to indicate that the file system has changed. When the file
> >system is unmounted, the field is cleared and the file system is
> >clean. Mount and fsck use this flag to determine whether the file
> >system needs to be checked or that it be corrupt; if the system
> >crashed before a reboot, the field won't have been cleared. The
> >root can never be unmounted cleanly, so it always has to be
> >checked.
>
> My comment:
> since there is such a string in the /sbin/rc0 file:
>
> /sbin/umount -a 2> /dev/null ,
>
> I'll try "init s" instead of shutdown (in order to fight against the cause
> rather than a consequence, and keeping in mind that the / file system
> is always mentioned in /etc/fstab). I have to delay the experiment until
> weekend.
>
> So, the matter does not concern the version of the system.
> All of those who answered mentioned fsck. My original mail is at the end of
> the message.
>
> Thanks again,
> Irene
> *************************************************************************
>
> On Wed, 5 May 1999, Irene A. Shilikhina wrote:
>
> >
> > Hello Managers,
> >
> > I've just installed 4.0D from scratch, ang now I'm in the phase of settling
> > all necessary things before commitment it to the users. So both systems are
> > able to boot from different disks now. To make this process easier, I'd like
> > to mount the file systems created under 3.2c when running the new system.
> > I'm only able to make it with one of the two disks - that with /usr2 file
> > system (this drive is an IBM's and with no disklabel at all). As to the
> > other one, with / and /usr file systems (a DEC's drive, originally shipped
> > with 3.2c FIS, and of course with a disklabel), neither of that file systems
> > won't do it with such an error in mounting: Dirty file system. In fact,
> > everything is all right with this disk since under 3.2c it has been working
> > well.
> >
> > On the other hand, when I try to mount the disk with the newinstalled 4.0D
> > (non-DEC's drive) under 3.2c, it's a success if only it concerns /usr on
> > /dev/rz1g but fails with /dev/rz1a - with the same message: Dirty file system.
> >
> > Neither Release Notes nor the archive have helped me out.
> >
> > Thanks,
> > Irene
>
>
Received on Sat May 08 1999 - 04:33:12 NZST