I thought that maybe something willbe correlated on how the /dev/hda6 is mounted!amanda-2.4.4 compiled from scratch! linuxDistrib:mandrake9.2
The problem has I saw it now, is that the user/groups defined in the amanda's xinet file cannot run the dump program correctly under the backuped host (that what i've seen in the /tmp/amanda/sendsize.xx file):
cat /tmp/amanda/sendsize.xx
....
sendsize[30425]: time 2.180: running "/sbin/dump 0Ssf 1048576 - /export/diskA1/pittelou"
sendsize[30425]: time 2.180: running /unige/amanda/libexec/killpgrp
sendsize[30425]: time 2.228: DUMP: Cannot open /dev/hda6
What does 'ls -l /dev/hda6' say?
# more /etc/fstab | grep hda6
/dev/hda6 /export/diskA1 ext3 defaults 1 2
# grep hda6 /etc/mtab
/dev/hda6 /export/diskA1 ext3 rw 0 0
but it looks fine! :(
ls -l /dev/hda6
lr-xr-xr-x 1 root root 33 oct 28 11:40 /dev/hda6 -> ide/host0/bus0/target0/lun0/part6
# ls -l /dev/ide/host0/bus0/target0/lun0/part6
brw------- 1 root root 3, 6 jan 1 1970 /dev/ide/host0/bus0/target0/lun0/part6
I've just created an amanda/disk user/group, and as the amanda's user > Hmm. On my RH systems, I put amanda in the 'disk' group, which owns the
raw disk devices.and I chown root:disk on /dev/ide/host0/bus0/target0/lun0/part6 without succes backuping up under the amanda user!
any other idea?
C�dric Briner
