RFC 2005 - mounting two slices from 2 different disks to same
mount point ;)
bob89 at eng.ufl.edu
Thu Mar 24 07:51:05 PST 2005
Odhiambo Washington wrote:
>On FreeBSD 5.3-RELEASE, we have mounted slices as follows:
>/dev/ad0s3a on / (ufs, local)
>devfs on /dev (devfs, local)
>/dev/ad0s3e on /tmp (ufs, local, soft-updates)
>/dev/ad0s3f on /usr (ufs, local, soft-updates)
>/dev/ad0s3d on /var (ufs, local, soft-updates)
>devfs on /var/named/dev (devfs, local)
>/dev/ad0s5 on /drive_d (msdosfs, local)
>/dev/ad3s1d on /tmp (ufs, local, soft-updates)
>/dev/ad0s3e was 500MB
>/dev/ad3s1d was 40GB
>In /etc/fstab, we specified the mount point for both to /tmp.
>We then wrote a 1GB file to /tmp and there was no error. The
>file was successfully written.
>Can an expert with disks tell me where the file was written?
>How did the system decide where to write the file? How did it
>"select" /tmp in this case?
It was written to the last disk you mounted. When you mount a disk, it
replaces whatever was mounted there previously. At least, that's what
the mount docs say, I didn't actually try to test it. If you do a df it
show that /tmp has the capacity of the last drive you mounted there.
Try mounting the 40GB drive first and then the 500 MB drive and
run your test again.
>Linux guys have talked about "disk spanning" using the method
>we just tested. I am wondering the dangers I face doing such
If you want to span disks, take a look at gconcat(8) or vinum(8). gconcat
replaces an older utility I don't remember the name of at the moment.
>Happy Easter to ye all!!
More information about the freebsd-questions