[KLUG Members] Re: SLES 9 not showing free disk space

[Stuart Gillis]

Seemed like good idea but I got an error.

projsrv2:/ # blockdev --rereadpt /dev/cciss/c0d0p2
BLKRRPART: Invalid argument
projsrv2:/ # df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/cciss/c0d0p2      16G   15G   28M 100% /
tmpfs                 1.5G   12K  1.5G   1% /dev/shm
/dev/cciss/c0d0p1      63M   39M   25M  62% /boot
/dev/cciss/c0d0p7     648G  326G  323G  51% /data
/dev/cciss/c0d0p3     5.1G  1.1G  4.0G  21% /tmp
/dev/cciss/c0d0p5      11G  144M  9.9G   2% /var

Any ideas?

On 9/7/05, Marr <marr at flex.com> wrote:
> On Wednesday 07 September 2005 10:43am, Stuart Gillis wrote:
> > I have a server running SLES 9, apache and oracle 10g.
> >
> > Oracle is installed in /opt which is part of the same partition as
> > root. Data files are in a separate partion. Oracle created a 6.7GB
> > temp01.dbf file in /opt which filled the partition which is causing
> > havoc. I copied the temp01.dbf file to the /data partition and oracle
> > was reconfigured to look for it there. I then deleted the temp01.dbf
> > file from /opt.
> >
> > The problem is after deleting this 6.7GB file, df -h still shows that
> > partition as full.
> >
> > dev/cciss/c0d0p2      16G   15G   28M 100% /
> > tmpfs                 1.5G   12K  1.5G   1% /dev/shm
> > /dev/cciss/c0d0p1      63M   39M   25M  62% /boot
> > /dev/cciss/c0d0p7     648G  326G  323G  51% /data
> > /dev/cciss/c0d0p3     5.1G  1.1G  4.0G  21% /tmp
> > /dev/cciss/c0d0p5      11G  141M  9.9G   2% /var
> >
> > Is there a way to force the system to rescan the partitions?
> 
> Try this:
> 
>    blockdev --rereadpt
> 
> I've used that in the past for (unmounted) external hard disk drives whereby
> 
> 'fdisk -l' is showing stale (i.e. shows info from the previously-mounted 
> drive) partition size data.
> 
> I don't know if it'll work in your particular case though -- seems like your
> 
> problem is too dissimilar to mine.
> 
> HTH....
> 
> Bill Marr
>