We have been using gnu tar with amanda to backup our ZFS filesystems on Solaris (both SPARC and x86) for more than 10 years. We take a ZFS snapshot (destroying the previous snapshot if it exists), run amanda pointing to the snapshot directory. We are in the process of replacing our old Solaris x86 file server with one running CentOS and ZFS. Unfortunately, that project has now stalled because the same process that we've been using no longer works. Every night we end up backing up the full filesystem (only three at the moment, but that's still hundreds of gigabytes). I did add the --no-check-device switch, but that didn't help.
Here is an example set of commands (I forgot to save the "Totals" line from the first run, but it's approximately the same as from the second run): # Do full backup, creating the list incremental file tar --create --file - --directory /export/home/gnu/.zfs/snapshot/AMANDA --one-file-system --no-check-device --listed-incremental /tmp/gnu_0 --ignore-failed-read --totals . | cat > /dev/null # Wait a day, destroying and creating a new snapshot cp /tmp/gnu_0 /tmp/gnu_0.new tar --create --file - --directory /export/home/gnu/.zfs/snapshot/AMANDA --one-file-system --no-check-device --listed-incremental /tmp/gnu_0.new --ignore-failed-read --totals . | cat > /dev/null Total bytes written: 81944340480 (77GiB, 75MiB/s) While examining the two listed incremental files, the timestamp at the beginning is obviously changed, but then a bit into the file there are directories that are in a different order, though the contents don't seem to have changed. Any ideas on what I should be looking for to try to resolve this issue? I know this is also affecting other people. Thanks for any help you can provide. Pieter
