[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#616716: No crypted volume identification for existing volumes



Package: debian-installer
Severity: normal
Tags: d-i

Lates debian installer on Squeeze Iso disk debian-6.0.0-amd64-BD-1.iso used to
install new computer.

I've created on sda and sdb the same partition configuration:
xen:~# fdisk -l /dev/sda /dev/sdb

Disk /dev/sda: 320.0 GB, 320072933376 bytes
255 heads, 63 sectors/track, 38913 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1               2          12       88357+  fd  Linux raid autodetect
/dev/sda2              13        2444    19530752   fd  Linux raid autodetect
/dev/sda3            8520       38914   244140032   fd  Linux raid autodetect

Disk /dev/sdb: 320.0 GB, 320072933376 bytes
255 heads, 63 sectors/track, 38913 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1               2          12       88357+  fd  Linux raid autodetect
/dev/sdb2              13        2444    19530752   fd  Linux raid autodetect
/dev/sdb3            8520       38914   244140032   fd  Linux raid autodetect

At the moment the above two hds are available for installation. After migrating
something from  the old two existing
hard drives with the same geometry I will add them in the system, too.
So I want to configure all md's with 4 drives where 2 are missing.
This is not configurable in debian installer at the moment so I created the
needed md's manually with mdadm command.

In short the command I used must be

mdadm --bitmap=internal --name=bootmd --level=1 --raid-devices=3 --spare-
devices=1 /dev/md125 /dev/sda1 /dev/sdb1 missing missing
mdadm --bitmap=internal --name=bootmd --level=1 --raid-devices=3 --spare-
devices=1 /dev/md126 /dev/sda2 /dev/sdb2 missing missing
mdadm --bitmap=internam --name=datamd --level=6 --raid-devices=4 /dev/md127
/dev/sda3 /dev/sdb3 missing missing

Here the output of the examine command after starting the old lenny:
xen:~# mdadm --examine /dev/sda1
/dev/sda1:
          Magic : a92b4efc
        Version : 01
    Feature Map : 0x1
     Array UUID : ffc6e33d:438fe377:28767059:ae3356d9
           Name : xen:bootmd  (local to host xen)
  Creation Time : Wed Mar  2 13:03:11 2011
     Raid Level : raid1
   Raid Devices : 3

    Device Size : 176691 (86.29 MiB 90.47 MB)
     Array Size : 176690 (86.29 MiB 90.47 MB)
      Used Size : 176690 (86.29 MiB 90.47 MB)
    Data Offset : 24 sectors
   Super Offset : 8 sectors
          State : clean
    Device UUID : af12b571:217d90d7:7dc7b5ee:850ef865

Internal Bitmap : 8 sectors from superblock
    Update Time : Sun Mar  6 02:28:30 2011
       Checksum : 79726af5 - correct
         Events : 370


    Array Slot : 0 (0, 1)
   Array State : Uu_
xen:~# mdadm --examine /dev/sda2
/dev/sda2:
          Magic : a92b4efc
        Version : 01
    Feature Map : 0x1
     Array UUID : 48d33aa8:997c6d99:ac8f68c1:a1a4a73e
           Name : xen:rootmd  (local to host xen)
  Creation Time : Wed Mar  2 13:03:45 2011
     Raid Level : raid1
   Raid Devices : 3

    Device Size : 39059456 (18.63 GiB 20.00 GB)
     Array Size : 39059312 (18.62 GiB 20.00 GB)
      Used Size : 39059312 (18.62 GiB 20.00 GB)
    Data Offset : 2048 sectors
   Super Offset : 8 sectors
          State : clean
    Device UUID : ec97cdb6:dfa61e5a:134643a1:f7d45645

Internal Bitmap : 8 sectors from superblock
    Update Time : Sun Mar  6 02:28:33 2011
       Checksum : 30d293a4 - correct
         Events : 3808


    Array Slot : 0 (0, 1)
   Array State : Uu_
xen:~# mdadm --examine /dev/sda3
/dev/sda3:
          Magic : a92b4efc
        Version : 01
    Feature Map : 0x1
     Array UUID : 45d7a2d4:628276ee:28b1e996:c44bc3c6
           Name : xen:datamd  (local to host xen)
  Creation Time : Sat Mar  5 02:23:00 2011
     Raid Level : raid6
   Raid Devices : 4

    Device Size : 488278016 (232.83 GiB 250.00 GB)
     Array Size : 976553984 (465.66 GiB 500.00 GB)
      Used Size : 488276992 (232.83 GiB 250.00 GB)
    Data Offset : 2048 sectors
   Super Offset : 8 sectors
          State : clean
    Device UUID : b5f9456b:6b379c3f:0837a437:044b2aa8

Internal Bitmap : 8 sectors from superblock
    Update Time : Sat Mar  5 22:28:58 2011
       Checksum : c59fd9de - correct
         Events : 168

     Chunk Size : 512K

    Array Slot : 0 (0, 1)
   Array State : Uu__

All the rest I did with the installer:
The /dev/md125 is configured as /boot with ext3.

Then I created on /dev/md126 and /dev/md127 the crypted devices without
changing the parameters.

After the crypting devices are created (named md126_crypt and md127_crypt ?)
I've set up both for LVM
and created in md126_crypt the xenrootdg. md127_crypt I left alone for the
installation.

On xenrootdg I created an varvol for /var, homevol for /home, rootvol for / all
with ext3. I've used varfs, homefs
and rootfs for logical name of the filesystem. I created swapvol for swap use.

Here the graphical layout:
/boot   ext3 bootfs on md125 with sda1|sdb1|missingdisk|missingspare
/               ext3    rootfs on xenrootdg/rootvol in md126_crypt on md126
with sda2|sdb2|missingdisk|missingspare
/home   ext3    homefs on xenrootdg/rootvol in md126_crypt on md126 with
sda2|sdb2|missingdisk|missingspare
/var            ext3    varfs on xenrootdg/rootvol in md126_crypt on md126 with
sda2|sdb2|missingdisk|missingspare

After installation was done I'couldn't boot the system because there couldn't
find the root filesystem. But this is here
not the point of the bug.

So I tried to start the rescue system option with the debian squeeze bd. It has
been startet and at the time of asking
where the root fs is, I cannot see it in the list. There was much partitions
found incl. the started md devices, but
there has'nt been unlocked the encrypted devices (here md126). That is the
point which I try to explain.

So I started with Alt-F2 an console and made the following manually:

modprobe dm_crypt, cbc, sha256_generic, aes_generic.
Now it is possible to check all device (incl. the md devices 126) with
        cryptsetup isLuks /dev/md126
so on the above there is an crypted device is found. Now I unlock the device
with
        cryptsetup luksOpen /dev/md126 md126_crypt.

Now I go back with cancel and then try again if the warnings are shown. Now the
rootdg and there volumes are
shown so I can give xenrootdg/rootvol as the root device for the rescue system.

So there are two points the problem should fix:

After generell testing and starting the md devices it should been checked for
existend crypt devices with cryptsetup isLuks.
If found the user should been ask for unlock the device.
After this all volumes should been checked for lvm volumes which should been
checked again for encrypted devices and so on ;-)

I hope this could been fixed easily.

The identification should been started in installation mode, two. Here there is
the same problem so without manually parts it is not possible to install
squeeze on an existing system.



-- System Information:
Debian Release: 6.0
Architecture: amd64 (x86_64)

Kernel: Linux 2.6.32-5-amd64 (SMP w/8 CPU cores)

Attachment: signature.asc
Description: OpenPGP digital signature


Reply to: