[rear-users] Several errors with NETFS recovery from ISO

Jeremy Laidman jlaidman at rebel-it.com.au
Mon Jul 14 08:35:40 CEST 2014


I'm using OUTPUT=ISO and BACKUP=NETFS so I can recover entirely from one
ISO.  However on recovery, I get several errors that I don't understand, so
I don't know if I'm doing something wrong or if it's a bug.  I'm using ReAR
v1.16.1 (installed from RPM) running on RHELv6.

>From the recovery log:

2014-07-14 15:02:36 fifo creation succesful !
2014-07-14 15:02:36 Including restore/NETFS/default/40_restore_backup.sh
mkdir: cannot create directory
`/tmp/rear.PQ1W8GIAWo8CUMA/outputfs/localhost': Read-only file system
2014-07-14 15:02:36 Decrypting disabled
2014-07-14 15:02:36 Restoring tar archive ''
/usr/share/rear/restore/NETFS/default/40_restore_backup.sh: line 42: [: ==:
unary operator expected

The error in line 42 of 40_restore_backup.sh appears to be expecting
$BACKUP_TYPE to be non-empty.  I think the default is "full" and so I
suspect this can be fixed by enclosing "$BACKUP_TYPE" in double quotes.  If
I set BACKUP_TYPE to "full" in my site.conf then this message doesn't show
up.  I can't tell if this is causing other problems like...

The second error:

2014-07-14 15:02:36 dd if=/tmp/rear.PQ1W8GIAWo8CUMA/tmp/tar_fifo | cat |
tar --block-number --totals --verbose --gzip -C /mnt/local/ -x -f -
2014-07-14 15:02:37 Processing backup.tar.gz.00 ...
gzip: stdin: invalid compressed data--format violated
tar: Unexpected EOF in archive
tar: Unexpected EOF in archive
tar: Error is not recoverable: exiting now
rm: cannot remove `/tmp/rear.PQ1W8GIAWo8CUMA/tmp/backup.md5': No such file
or directory
2014-07-14 15:08:42 WARNING !
There was an error (Nr. 2) while restoring the archive.
Please check '/var/log/rear/rear-localhost.log' for more information. You
should also
manually check the restored system to see whether it is complete.

I'm guessing that this error is related to the fifo, which seems to be
providing a corrupted stream that gzip can't parse.  Weirdly, this was all
after the partitions were all created and a tar restore was completed, and
so I'm wondering if it's because ReAR thinks that the backup is
multi-volume when it isn't.

Possibly related is an error near the end of the log that says:

2014-07-14 15:16:22 Running exit tasks.
2014-07-14 15:16:22 The following jobs are still active:
[2]  17901 Exit 1                  ( if [[ -f "${TMP_DIR}/backup.splitted"
]]; then
    Print ""; while read file; do
        name=${file%% *}; vol_name=${file##* };
file_path="${opath}/${name}"; touch ${TMP_DIR}/wait_dvd; while ! [[ -f
$file_path ]]; do
            umount "${BUILD_DIR}/outputfs"; ProgressInfo "Please insert the
media called $vol_name in your CD-ROM drive..."; sleep 2; drive=$(cat
/proc/sys/dev/cdrom/info | grep -i "drive name:" | awk '{print $3 " "
$4}'); for dev in $drive

and then

/usr/share/rear/lib/_input-output-functions.sh: line 65: kill: (17901) - No
such process

The active job code is from restore/NETFS/default/40_restore_backup.sh.

As I said, the restore appears to work OK, despite the error from tar and
the lingering job, but I'm not entirely sure.  The restore log doesn't give
me any hints.  Can anyone help?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://pikachu.3ti.be/pipermail/rear-users/attachments/20140714/ca71724c/attachment.html>

More information about the rear-users mailing list