#319 Issue closed: Wrong diskrecover on RHEL4 and rear 1.14¶
Labels: bug, documentation
MarkusRoth opened issue at 2013-10-25 09:27:¶
I want to recover a RHEL4 system. Backup with rear on a cifs share was successful. I boot the test system (identical hardware) with the rescue ISO. The recover process stops at generating the disklayout.
Found out, that the parted -s /dev/sda mkpart primary for /dev/sda4 is
wrong. It must be "extended" instead of "primary" and the End Blocks
must be changed.
Although the lines parted -s /dev/sda set /dev/sda3 on and furhter
must be deleted.
The command for creating the filesystem on /dev/sda6 (/root2). The
tune2fs -U <none> ... had no FS-UUID of this partition. I checked on
the original system with tune2fs -l /dev/sda6. There was no FS UUID !
gdha commented at 2013-10-25 10:36:¶
From the mail of Markus:
Log file:
+++ Print 'Creating partitions for disk /dev/sda (msdos)'
+++ test 1
+++ echo -e 'Creating partitions for disk /dev/sda (msdos)'
+++ parted -s /dev/sda mklabel msdos
+++ parted -s /dev/sda mkpart primary 0 5671
+++ parted -s /dev/sda set 1 boot on
+++ parted -s /dev/sda mkpart primary 5671 53599
+++ parted -s /dev/sda mkpart primary 53599 101527
+++ parted -s /dev/sda set /dev/sda3 on
Error: Expecting a partition number.
2013-10-25 08:21:19 An error occured during layout recreation.
2013-10-25 08:22:03 User selected: 6) Abort Relax-and-Recover
2013-10-25 08:22:03 Error detected during restore.
2013-10-25 08:22:03 Restoring backup of /var/lib/rear/layout/disklayout.conf
2013-10-25 08:22:03 ERROR: There was an error restoring the system layout. See /var/log/rear/rear-daytona2.log for details.
The disklayout looks as follows:
#disk /dev/hda 4294965248
disk /dev/sda 291999055872 msdos
part /dev/sda 5946845184 32256 primary boot /dev/sda1
part /dev/sda 50256460800 5946877440 primary none /dev/sda2
part /dev/sda 50256460800 56203338240 primary /dev/sda3
part /dev/sda 1024 106459799040 primary /dev/sda4
part /dev/sda 10075935744 106459831296 logical /dev/sda5
part /dev/sda 5946845184 116535799296 logical /dev/sda6
part /dev/sda 11852596224 122482676736 logical /dev/sda7
part /dev/sda 11720991744 134335305216 logical /dev/sda8
fs /dev/sda1 / ext3 uuid=8f3bc4f0-d605-4029-8cf2-c8fff9dfb95b label= blocksize=4096 reserved_blocks=4% max_mounts=27 check_interval=180d bytes_per_inode=8177 options=rw
fs /dev/sda6 /root2 ext3 uuid=<none> label= blocksize=4096 reserved_blocks=4% max_mounts=25 check_interval=180d bytes_per_inode=8177 options=rw,noexec,nosuid,nodev,noatime
fs /dev/sda5 /var ext3 uuid=79293447-713d-48e9-be5f-dfe2a680e623 label= blocksize=4096 reserved_blocks=4% max_mounts=37 check_interval=180d bytes_per_inode=8187 options=rw,noexec,nosuid,nodev,noatime
fs /dev/sda2 /opt/avaya ext3 uuid=78f75f71-8e71-403e-a98d-798b6212641d label= blocksize=4096 reserved_blocks=4% max_mounts=31 check_interval=180d bytes_per_inode=8179 options=rw
fs /dev/sda3 /opt/dbase ext3 uuid=58b70eff-daf0-4ab0-9fca-e9f31999970b label= blocksize=4096 reserved_blocks=4% max_mounts=25 check_interval=180d bytes_per_inode=8179 options=rw
fs /dev/sda8 /DVD_OPS ext3 uuid=dd8dd625-1c8d-4ed9-83f8-2ba60fe287d3 label= blocksize=4096 reserved_blocks=4% max_mounts=32 check_interval=180d bytes_per_inode=8177 options=rw
swap /dev/sda7 uuid= label=
fdisk -l on the original system looks:
Disk /dev/sda: 291.9 GB, 291999055872 bytes
255 heads, 63 sectors/track, 35500 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/sda1 * 1 723 5807466 83 Linux
/dev/sda2 724 6833 49078575 83 Linux
/dev/sda3 6834 12943 49078575 83 Linux
/dev/sda4 12944 35500 181189102+ 5 Extended
/dev/sda5 12944 14168 9839781 83 Linux
/dev/sda6 14169 14891 5807466 83 Linux
/dev/sda7 14892 16332 11574801 82 Linux swap
/dev/sda8 16333 17757 11446281 83 Linux
Has anybody a hint why the creation of the extended partition /dev/sd4 fails?
regards,
Markus
gdha commented at 2013-10-25 10:37:¶
This issue is linked with another issue #263
To further investigate we would like to see:
- parted -v
- parted -m -s /dev/sda print
- parted -s /dev/sda print
- sfdisk -l /dev/sda
- disklayout.conf
- rear -d savelayout output (you may attach it via Gist)
MarkusRoth commented at 2013-10-25 10:49:¶
GNU Parted 1.6.19
Output of parted -s /dev/sda print is:
Disk geometry for /dev/sda: 0.000-278472.000 megabytes
Disk label type: msdos
Minor Start End Type Filesystem Flags
1 0.031 5671.384 primary ext3 boot
2 5671.384 53599.680 primary ext3
3 53599.680 101527.976 primary ext3
4 101527.976 278470.458 extended
5 101528.007 111137.167 logical ext3
6 111137.199 116808.552 logical ext3
7 116808.583 128112.099 logical linux-swap
8 128112.130 139290.139 logical ext3
Output of sfdisk -l /dev/sda is
Disk /dev/sda: 35500 cylinders, 255 heads, 63 sectors/track
Units = cylinders of 8225280 bytes, blocks of 1024 bytes, counting from
0
Device Boot Start End #cyls #blocks Id System
/dev/sda1 * 0+ 722 723- 5807466 83 Linux
/dev/sda2 723 6832 6110 49078575 83 Linux
/dev/sda3 6833 12942 6110 49078575 83 Linux
/dev/sda4 12943 35499 22557 181189102+ 5 Extended
/dev/sda5 12943+ 14167 1225- 9839781 83 Linux
/dev/sda6 14168+ 14890 723- 5807466 83 Linux
/dev/sda7 14891+ 16331 1441- 11574801 82 Linux swap
/dev/sda8 16332+ 17756 1425- 11446281 83 Linux
See also Gist: https://gist.github.com/MarkusRoth/7192428/
gdha commented at 2013-10-28 12:46:¶
In the debug output I clearly see the following:
+++ sfdisk -c /dev/sda 4
++ declare partition_id=5
++ [[ 5 = @(5|f|85) ]]
++ sed -i '/^4 /s/ primary / extended /' /tmp/rear.fdgeeRpDvd29698/tmp/partitions
++ read partition_nr size start type flags junk
The question is now why a bit further down in the logging I see (reading
input from $TMP_DIR/partitions:
++ echo 'part /dev/sda 1024 106459799040 primary /dev/sda4'
@jhoekx Jeroen - can you see it? Why wasn't it saved properly in partitions (extended keyword) - don't get it. If I do it by hand it works.
gdha commented at 2013-11-15 11:47:¶
@jhoekx could you do me a favor and have a quick look at the following piece of code change - does it makes sense or was there a specific reason you wrote it as it is?
diff ./20_partition_layout.sh ~/projects/rear/usr/share/rear/layout/save/GNU/Linux/20_partition_layout.sh
99c99,103
< if (( $partition_nr < 4 )) ; then
---
> if (( $partition_nr > 4 )) ; then
> ### logical
> has_logical=1
> sed -i /^$partition_nr\ /s/$/\ logical/ $TMP_DIR/partitions
> else
103,109d106
< elif (( $partition_nr = 4 )) ; then
< declare type="extended"
< sed -i /^$partition_nr\ /s/$/\ extended/ $TMP_DIR/partitions
< else
< ### logical for $partition_nr > 4
< has_logical=1
< sed -i /^$partition_nr\ /s/$/\ logical/ $TMP_DIR/partitions
jhoekx commented at 2013-11-15 12:06:¶
The reason is that /dev/sd*4 can be a primary partition. The first four can be primary or extended.
gdha commented at 2013-11-15 13:22:¶
@jhoekx Ok thanks for the clarification - the code mentioned just above
makes no sense then. Agreed.
However, 3 text boxes above it was shown it should have changed it, but
it didn't and I couldn't find the reason.
jhoekx commented at 2013-11-15 13:25:¶
I'm also puzzled why it fails.
I don't have a RHEL 4 system anymore to test... I really should spend some more time on Rear :-)
gdha commented at 2013-11-25 07:56:¶
@MarkusRoth Sorry, 'till now we didn't find the reason why it was not changed properly. We don't have a RHEL4 anymore to try to reproduce this behavior and more recent versions seem to work fine. As said we're puzzled!
gdha commented at 2013-12-03 15:25:¶
will make a note in the release notes about this.
gdha commented at 2014-04-02 08:27:¶
added to release notes
[Export of Github issue for rear/rear.]