determining the space used in / partition
Jerry McAllister
jerrymc at msu.edu
Tue Oct 2 07:50:10 PDT 2007
On Tue, Oct 02, 2007 at 07:23:30AM +0200, Zbigniew Szalbot wrote:
> Hello again,
>
> > > Through df I realized my / partiotion is out of space:
> > > Filesystem 1K-blocks Used Avail Capacity Mounted on
> > > /dev/ad0s1a 198126 196070 -13794 108% /
> > > devfs 1 1 0 100% /dev
> > > /dev/ad0s1e 44511308 4217762 36732642 10% /usr
> > > /dev/ad0s1d 30462636 3210580 24815046 11% /var
> > > devfs 1 1 0 100% /var/named/dev
> > > /dev/da0s1c 75685352 34308200 35322324 49% /mnt/usbck
> > >
> > > How can I determine what occupies the space in it? That is, it is not
> > > big as you can see. So I issued:
> > > du -hs /
> > > but it was taking ages (I am not sure but maybe du -hs counts all
> > > directories on the HD?
> > >
> > > Anyway, I do not really know where to look what has eaten the / space.
> > > Were it for /usr or /var, it would be obvious to me where to look for
> > > information.
> > >
> > > Many thanks!
> >
> > I don't see you have defined a /tmp partition. Perhaps /tmp is taking up
> > all the space. Try:
> >
> > du -h /tmp
> >
> > and see how much /tmp is taking up.
> du -hs /tmp
> 1.4M /tmp
>
> du -hs /
> 40GB
>
> One thing that comes to my mind. Each Sunday I have a script which
> makes a full dump of the HD to a back-up USB drive. Last weekend
> someone cleaining the computer room, must have accidentally powered
> off the USB drive. As a result, the dump has not been completed
> because the USB drive was not mounted at that time. I use cron for
> this task. Does it matter could have caused this?
It probably then wrote a large dump file at the mount point
you usually use for the USB drive. It looks like /mnt/usbck now
has 34 GB in it. Is /mnt/usbck where the USB is normally mounted?
Maybe, if you unmounted the USB and then looked at it, you would
find the used up space.
But, a previous poster could also be correct that it might be
that you have filled up /tmp, maybe with error writes or something.
When I use du I like to do the following:
cd /directory_of_interest
du -sk *
That gets the summary of each directory and file in
that directory_of_interest. I like the 'k' better than 'h' because
the 'h' doesn't use the same divider for each displayed file or directory.
It uses the biggest for each with a letter appended to tell which. This
is a little difficult to quickly compare with a visual scan. With the 'k'
it is always 1,000 and then I can run my eye down the list and easily see
which file is bigger/smaller, etc.
////jerry
>
> Thanks!
>
> zbigniew szalbot
> _______________________________________________
> freebsd-questions at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-questions
> To unsubscribe, send any mail to "freebsd-questions-unsubscribe at freebsd.org"
More information about the freebsd-questions
mailing list