[ solved partly ] Hard drive disappears/reappears

[wedgetail]

Conclusion: Physically the hard drive seems to have come right after re-seating cables. The following draklive-install problem in the end almost certainly is caused by having a USB memory stick connected at the wrong time, ie when running draklive-install.



Background
I have/had a master computer running Windows2000 for many years, it has presently one EIDE 80GB Western Digital WD800BB 00JHA0.  I think about 85% full so basically data is stored on 150 GB USB drive which is connected as required.

I think it was mid last year last year booting began to be a bit unreliable on first go but second attempt fine. It was infrequent so I did not think much about it.  Then it became more noticeable and I got a bit concerned and prepared to copy everything of value. To make sure I did not loose anything. (I have quite a number of GHOST copies so I would never loose everything, but since I have started moving to Linux I am getting slacker and probably could be a 12 months loss)

I did manage to get copies of everything though during the last week I lost contact very predictably after a number of hours. re-seating the drive cable both ends then seemed to bring it to live though I might have to wait overnight.  As the OS was Win2000 my local wizard suggested I install XP on a new PC, since he was giving up supporting Win2000 and concentrate on XP.  As I need to keep Windows running for some more years for my wife I have set up a new system and are transferring data as we need it from the backup.

Problem:
I would like to figure what is going on before I thrash the drive contents. To this end I have started running LiveCDs, LXDE, KDE, KDE-mini really what was closest. Initially that worked fine so I thought it was the Windows install. Then on a couple of occasions Linux could not see the drive, the hard drive again disappeared.

I gave up I think it was 30GB of photos and have reformatted this ext3 and wanted to install pclos here but this failed, not sure at what stage.  

Got as small 2500 MB drive out tried to get that running but SMART predicted problems and since the size so small I gave this away.  Fished out a SeaGate 850MB drive and that booted fine into an odd Linux system from long ago, forgotten about it, from probaly aroun year 2000.  ie now the cable system to that EIDE channel has been wiggled.

I have reinstalled the original 80GB and today I have gone to the Western Digital site and entered my drive model and found diagnostic software: DLGDIAG 5.04 this is a DOS version which makes me feel a little bit more comfortable.

Downloaded the iso and burnt a CD and popped it in the tray. I was a bit slow and the system booted nicely into Win2000, so the hard drive has re-appeared again.  Anyway I restarted and the have run the Quick Test and Extended Test.

The next tests is writing zeros to the drive so this is a bit early.

I am not quite sure what to think or expect but since old-polack hardly ever looses a drive I am reluctant to just ditch it without a fight.  So I will also be looking for the "scrub" command he uses sometimes, I think it was dd= something and when I used it last it took 10 or more hours on a small drive but the drive is still running somewhere.

I have done a quick search on Forum but not found what I wanted  

Now the DLGDIAG test is finished I will try and see if I can install KDE-mini again, then install was it smartmontools I have seen mentioned in some topics and see what happens.  

Any suggestions welcome

[wedgetail]

Something ain't right, have now run up KDE-mini  LiveCD and inspected the drive partition scheme

[root@localhost ~]# fdisk -l -u=cylinders

Code: [Select]

Disk /dev/sda: 80.0 GB, 80026361856 bytes
255 heads, 63 sectors/track, 9729 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0xa1dba1db

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *           1         957     7687071    b  W95 FAT32
/dev/sda2             958        5766    38628292+   f  W95 Ext'd (LBA)
/dev/sda3            5767        9729    31832797+   c  W95 FAT32 (LBA)
/dev/sda5             958        2997    16386268+   b  W95 FAT32
/dev/sda6            2998        5766    22241961    b  W95 FAT32

Disk /dev/sdb: 4007 MB, 4007657472 bytes
86 heads, 22 sectors/track, 4137 cylinders
Units = cylinders of 1892 * 512 = 968704 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0xff88a840

To make it a bit easier seeing Labels I used cfdisk:

Code: [Select]

                     cfdisk (util-linux-ng 2.18)

                          Disk Drive: /dev/sda
                    Size: 80026361856 bytes, 80.0 GB
          Heads: 255   Sectors per Track: 63   Cylinders: 9729

   Name       Flags     Part Type FS Type        [Label]       Size (MB)
 ----------------------------------------------------------------------
   sda1       Boot       Primary  vfat          [C_CEL2000]      7871.60
   sda5                  Logical  vfat          [D_CEL2000]     16779.58
   sda6                  Logical  vfat          [E_CEL2000]     22775.81
   sda3                  Primary  ext4           [kde-mini]     32596.79

The labels are correct but the partition scheme with the new ext4 is not 'nice' looking. This must be so long ago I set this up.  /sda3 was a FAT32 partition holding the photos.

Well when I set up the linux partition I must have been thinking wrong choosing primary partition, does not look proper following old-polacks suggestion when partitioning. Since linux will run nicely in an extended partition I will change /sda3 to extended so it goes inside the /sda2 container.

Here is what happens if I try to install linux to the existing /sda3 primary partition.

[root@localhost ~]# draklive-install
/media/ user
Backtrace has 11 calls on stack:
  11: /usr/lib/libparted.so.0(ped_assert+0x29) [0xb71888b9]
  10: /usr/lib/libparted.so.0(+0x41435) [0xb71be435]
  9: /usr/lib/libparted.so.0(+0x422b7) [0xb71bf2b7]
  8: /usr/lib/libparted.so.0(+0x436aa) [0xb71c06aa]
  7: /usr/lib/libparted.so.0(+0x1028e) [0xb718d28e]
  6: /usr/lib/libparted.so.0(ped_disk_add_partition+0x271) [0xb7190bf1]
  5: /usr/lib/libparted.so.0(+0x450e3) [0xb71c20e3]
  4: /usr/lib/libparted.so.0(+0x4530e) [0xb71c230e]
  3: /usr/lib/libparted.so.0(ped_disk_new+0x79) [0xb7191a29]
  2: /usr/lib/libDrakX/auto/c/stuff/stuff.so(XS_c__stuff_get_disk_type+0x1ed) [0xb720dfd5]
  1: /usr/lib/perl5/5.10.1/i386-linux-thread-multi/CORE/libperl.so(Perl_pp_entersub+0x56e) [0xb76d51ee]
A bug has been detected in GNU Parted.  Refer to the web site of parted http://www.gnu.org/software/parted/parted.html for more information of what could be useful for bug submitting!  Please email a bug report to bug-parted@gnu.org containing at least the version (2.3) and the following message:  Assertion (head_size <= 63) at dos.c:661 in function probe_partition_for_geom() failed.
Aborted
[root@localhost ~]#

The above was the 'failed to install' I referred to in original post.  Is there anyone that can shed light on what I am being told.

I would like to think it may be something to do with my partitioning scheme?  I don't recall having seen this before when trying to use draklive-install  

Now thinking back when old-polack explains his magic, there is the scary job of doing surgery on the drives partition table.

Let me see if I have got it right:

/sda2 extend this to the end 9729 and then change /sda3 to /sda7 next when running fdisk -l the order should be right. Changes marked in red. This is proposed but how in practice.

  Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *           1         957     7687071    b  W95 FAT32
/dev/sda2             958        9729   38628292+   f  W95 Ext'd (LBA)
/dev/sda7            5767        9729    31832797+   c  W95 FAT32 (LBA)
/dev/sda5             958        2997    16386268+   b  W95 FAT32
/dev/sda6            2998        5766    22241961    b  W95 FAT32

I believe that extending the cylinders to the end 9729 can be carefully done by writing a new table, my luck has it that the present /sda3 is already in the last position physically and even has a filesystem.

Naaahh I am scaring myself,  for example /sda3 is not "Linux" type   help please.  

I would like to keep the Windows system running if the hard drive does not have problems. (not clear yet)

[wedgetail]

I made a bit of a search "change size partition table" in Forum with old-polack as author and came accross this one:

http://www.pclinuxos.com/forum/index.php/topic,100832.msg856795.html#msg856795

This triggered something from long ago about fdisk old / fdisk new I know the linux partition /sda3 I am referring to above was made with fdisk new as I used KDE-mini 2011.09.  Since the partition is empty I have now deleted this and the new table:

Code: [Select]

[root@localhost ~]# fdisk -l

Disk /dev/sda: 80.0 GB, 80026361856 bytes
255 heads, 63 sectors/track, 9729 cylinders, total 156301488 sect
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0xa1dba1db

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *          63    15374204     7687071    b  W95 FAT32
/dev/sda2        15374205    92630789    38628292+   f  W95 Ext'd
/dev/sda5        15374268    48146804    16386268+   b  W95 FAT32
/dev/sda6        48146868    92630789    22241961    b  W95 FAT32
[root@localhost ~]#

This time I let fdisk show sectors, as /sda1 starts on sector 63 this indicates to me old partition method. This would have been 2048 for new:

Code: [Select]

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1   *        2048    62926603    31462278   83  Linux
/dev/sdb2        62926604   146818033    41945715   83  Linux

The above is shown for a 1TB drive, I am not sure if this would have been true if the drive was only 80 GB making an assumption here.

Getting closer, there is now approximately 32 GB unused on the drive as the extended partition does not reach the end. This is exactly the silly situation that old-polack has pointed out a number of times. What I should do is find a way to expand the extended partition to the end, by using the fdisk only?

Since this is fdisk-old territory I had better find an old LiveCD to use.  With this I could delete the extended partition and very carefully with fdisk make a new one starting at the same sector as the one I have just deleted and let it finish at the end (default i think)

When that is done I can make a new logical partition which will be /sda7 and maybe I can install KDE-mini 

Very seldom is old-polack wrong so with a record of my partition tables, then deleting partitions does not loose me any data. Therefore I should be able to experiment.   

[Old-Polack]

I made a bit of a search "change size partition table" in Forum with old-polack as author and came accross this one:

http://www.pclinuxos.com/forum/index.php/topic,100832.msg856795.html#msg856795

This triggered something from long ago about fdisk old / fdisk new I know the linux partition /sda3 I am referring to above was made with fdisk new as I used KDE-mini 2011.09.  Since the partition is empty I have now deleted this and the new table:

Code: [Select]

[root@localhost ~]# fdisk -l

Disk /dev/sda: 80.0 GB, 80026361856 bytes
255 heads, 63 sectors/track, 9729 cylinders, total 156301488 sect
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0xa1dba1db

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *          63    15374204     7687071    b  W95 FAT32
/dev/sda2        15374205    92630789    38628292+   f  W95 Ext'd
/dev/sda5        15374268    48146804    16386268+   b  W95 FAT32
/dev/sda6        48146868    92630789    22241961    b  W95 FAT32
[root@localhost ~]#

This time I let fdisk show sectors, as /sda1 starts on sector 63 this indicates to me old partition method. This would have been 2048 for new:

Code: [Select]

  Device Boot      Start         End      Blocks   Id  System
/dev/sdb1   *        2048    62926603    31462278   83  Linux
/dev/sdb2        62926604   146818033    41945715   83  Linux

The above is shown for a 1TB drive, I am not sure if this would have been true if the drive was only 80 GB making an assumption here.

Getting closer, there is now approximately 32 GB unused on the drive as the extended partition does not reach the end. This is exactly the silly situation that old-polack has pointed out a number of times. What I should do is find a way to expand the extended partition to the end, by using the fdisk only?

Since this is fdisk-old territory I had better find an old LiveCD to use.  With this I could delete the extended partition and very carefully with fdisk make a new one starting at the same sector as the one I have just deleted and let it finish at the end (default i think)

When that is done I can make a new logical partition which will be /sda7 and maybe I can install KDE-mini  

Very seldom is old-polack wrong so with a record of my partition tables, then deleting partitions does not loose me any data. Therefore I should be able to experiment.  

Before you get carried away, try creating a new partition. When it asks for type, chose logical. If it accepts that, and the default start sector is above 92630789, accept that and then accept the end sector default as well. Use the p command to check what that looks like. The extended partition may actually then show that it now extends to the end of the drive, and the last partition should show that it's a Linux native partition type 83.

Post your results.

[wedgetail]

old-polack
Thank you for coming along. I took a break so as I would not rush this too fast but I am back on full steam.

I have started up pclos 2009 LiveCD and tried your suggestion.

Code: [Select]

[root@localhost guest]# fdisk -l

Disk /dev/hda: 80.0 GB, 80026361856 bytes
255 heads, 63 sectors/track, 9729 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xa1dba1db

   Device Boot      Start         End      Blocks   Id  System
/dev/hda1   *           1         957     7687071    b  W95 FAT32
/dev/hda2             958        5766    38628292+   f  W95 Ext'd (LBA)
/dev/hda5             958        2997    16386268+   b  W95 FAT32
/dev/hda6            2998        5766    22241961    b  W95 FAT32

[root@localhost guest]# fdisk /dev/hda
The number of cylinders for this disk is set to 9729.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
   (e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): m
Command (m for help): n
Command action
   l   logical (5 or over)
   p   primary partition (1-4)
l
No free sectors available

Command (m for help):                                  

One give away that fdisk-old is in use is the change of drive designation to /hda as well as back in cylinders as default unit.

I assume that fdisk-new would come up with the same answer.  

Go ahead delete the extended partion?

[Old-Polack]

old-polack
Thank you for coming along. I took a break so as I would not rush this too fast but I am back on full steam.

I have started up pclos 2009 LiveCD and tried your suggestion.

Code: [Select]

[root@localhost guest]# fdisk -l

Disk /dev/hda: 80.0 GB, 80026361856 bytes
255 heads, 63 sectors/track, 9729 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xa1dba1db

   Device Boot      Start         End      Blocks   Id  System
/dev/hda1   *           1         957     7687071    b  W95 FAT32
/dev/hda2             958        5766    38628292+   f  W95 Ext'd (LBA)
/dev/hda5             958        2997    16386268+   b  W95 FAT32
/dev/hda6            2998        5766    22241961    b  W95 FAT32

[root@localhost guest]# fdisk /dev/hda
The number of cylinders for this disk is set to 9729.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
   (e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): m
Command (m for help): n
Command action
   l   logical (5 or over)
   p   primary partition (1-4)
l
No free sectors available

Command (m for help):                                   

One give away that fdisk-old is in use is the change of drive designation to /hda as well as back in cylinders as default unit.

I assume that fdisk-new would come up with the same answer.  

Go ahead delete the extended partion?

No, just create a new partition and when asked make it a logical partition.

[wedgetail]

old-polack
Sorry for some time the post had verbal description about no free sectors, then I moved the facts via a USB drive and this camouflaged the result. I pondered your reply, but realised you had not seen the bottom of the fdisk result in the code section.

No free sectors available

I can only create a primary partition. 

[Old-Polack]

old-polack
Sorry for some time the post had verbal description about no free sectors, then I moved the facts via a USB drive and this camouflaged the result. I pondered your reply, but realised you had not seen the bottom of the fdisk result in the code section.

No free sectors available

I can only create a primary partition.  

You're right, I missed the lower half of that code box. OK, go ahead and delete the extended partition. This will also remove the two existing logical partitions. Create a new extended partition starting at the same start cylinder as the original and extend it to the end of the drive. Recreate the hda5 and hda6 partitions with the same start and end cylinders as now, then use the p command to see the results. The new partitions will be identified as Linux type 83 partitions. Use the t command on each of those partitions to change the id type to type b W95 FAT32. Last, create the new hda7 partition, which again will by default be listed as type 83 Linux, which in this case is what you want.

[wedgetail]

As I happened to have the KDE-mini LiveCD running on the system when old-polack gave the green light plus clear guideline of what to do I proceeded using fdisk-new, I think I was expecting it to work, but it did not quite make it. The extended partition re-created with the correct start sector it seemed and final sector on the drive, which was what I was after.

The logical partition /sda5 I failed to get correct starting point, I tried entering the start sector but no go, outside range. I accepted and because I did not write the changes to the drive, absolutely nothing happened.  This is when it really hit home to me how much one can delete partitions and recreate etc, no harm done until one hits the w=write changes to disk.

I fell back on the KDE 2009.1 LivdCD and repeated the procedure.

[root@localhost guest]# fdisk /dev/hda

The number of cylinders for this disk is set to 9729.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
   (e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): d
Partition number (1-6): 2    <-------------------------------------------- deleting extended partition

Command (m for help): n
Command action
   e   extended
   p   primary partition (1-4)
e
Partition number (1-4): 2            <--------------------------------------------   re-creating extended partition
First cylinder (958-9729, default 958):
Using default value 958
Last cylinder or +size or +sizeM or +sizeK (958-9729, default 9729):
Using default value 9729              <--------------------------------------------   notice both default start and default finish correct

Command (m for help): n
Command action
   l   logical (5 or over)
   p   primary partition (1-4)
l      <--------------------------------------------   selecting l that is L lower case for logical partition 5 and I screw up
First cylinder (958-9729, default 958):
Using default value 958
Last cylinder or +size or +sizeM or +sizeK (958-9729, default 9729): ^[[B^[[5~^[      <--- screw up                 
Using default value 9729

Command (m for help): d       <--------------------------------------------  deleting the partition again
Partition number (1-5): 5

Command (m for help): n       <--------------------------------------------   restarting logical partition recovery
Command action
   l   logical (5 or over)
   p   primary partition (1-4)
l
First cylinder (958-9729, default 958):
Using default value 958
Last cylinder or +size or +sizeM or +sizeK (958-9729, default 9729): 2997      <-------------- needed from notes

Command (m for help): n
Command action
   l   logical (5 or over)
   p   primary partition (1-4)
l
First cylinder (2998-9729, default 2998):
Using default value 2998
Last cylinder or +size or +sizeM or +sizeK (2998-9729, default 9729): 5766      <-------- again needed from notes

Command (m for help): n             <--------------------------------------------   setting up /sda7
Command action
   l   logical (5 or over)
   p   primary partition (1-4)
l
First cylinder (5767-9729, default 5767):
Using default value 5767
Last cylinder or +size or +sizeM or +sizeK (5767-9729, default 9729):
Using default value 9729                <--------------------------------------------  this is end of drive

Command (m for help): p

Disk /dev/hda: 80.0 GB, 80026361856 bytes
255 heads, 63 sectors/track, 9729 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xa1dba1db

   Device Boot      Start         End      Blocks   Id  System
/dev/hda1   *           1         957     7687071    b  W95 FAT32
/dev/hda2             958        9729    70461090    5  Extended
/dev/hda5             958        2997    16386268+  83  Linux  <--------------------------------------------  should be FAT32
/dev/hda6            2998        5766    22241961   83  Linux    <--------------------------------------------   same here 
/dev/hda7            5767        9729    31832766   83  Linux

Command (m for help): t
Partition number (1-7): 5
Hex code (type L to list codes): b
Changed system type of partition 5 to b (W95 FAT32)

Command (m for help): t
Partition number (1-7): 6
Hex code (type L to list codes): b
Changed system type of partition 6 to b (W95 FAT32)

Command (m for help): p

Disk /dev/hda: 80.0 GB, 80026361856 bytes
255 heads, 63 sectors/track, 9729 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xa1dba1db

   Device Boot      Start         End      Blocks   Id  System
/dev/hda1   *           1         957     7687071    b  W95 FAT32
/dev/hda2             958        9729    70461090    5  Extended
/dev/hda5             958        2997    16386268+   b  W95 FAT32
/dev/hda6            2998        5766    22241961    b  W95 FAT32
/dev/hda7            5767        9729    31832766   83  Linux

Command (m for help):

Looks jolly good

Command (m for help): w         <--------------------------------------------  moment of truth write changes to drive
The partition table has been altered!

Calling ioctl() to re-read partition table.

WARNING: If you have created or modified any DOS 6.x
partitions, please see the fdisk manual page for additional
information.
Syncing disks.
[root@localhost guest]#               

Well it worked I booted up in win2000 and all looked ok.  Next step will Linux now install???     
 

[wedgetail]

No Linux did not install, same problem as in the beginning, but partition order is nice   

Code: [Select]

[root@localhost ~]# cfdisk

                           cfdisk (util-linux-ng 2.18)

                          Disk Drive: /dev/sda
                    Size: 80026361856 bytes, 80.0 GB
          Heads: 255   Sectors per Track: 63   Cylinders: 9729

   Name       Flags     Part Type FS Type        [Label]       Size (MB)
 ----------------------------------------------------------------------
   sda1       Boot       Primary  vfat          [C_CEL2000]      7871.60
   sda5                  Logical  vfat          [D_CEL2000]     16779.58
   sda6                  Logical  vfat          [E_CEL2000]     22775.81
   sda7                  Logical  Linux                         32596.79


     
Here I did format /sda7 to ext4 before trying to install PCLOS KDE-mini

Code: [Select]

[root@localhost ~]# mke2fs -t ext4 -L KDE-mini /dev/sda7
mke2fs 1.41.14 (22-Dec-2010)
Filesystem label=KDE-mini
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
Stride=0 blocks, Stripe width=0 blocks
1990656 inodes, 7958191 blocks
397909 blocks (5.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=0
243 block groups
32768 blocks per group, 32768 fragments per group
8192 inodes per group
Superblock backups stored on blocks:
        32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,
        4096000

Writing inode tables: done
Creating journal (32768 blocks): done
Writing superblocks and filesystem accounting information: done

This filesystem will be automatically checked every 29 mounts or
180 days, whichever comes first.  Use tune2fs -c or -i to override.
[root@localhost ~]#
Trying to install KDE-mini

Code: [Select]

[root@localhost ~]# draklive-install
/media/ user
Backtrace has 11 calls on stack:
  11: /usr/lib/libparted.so.0(ped_assert+0x29) [0xb71f08b9]
  10: /usr/lib/libparted.so.0(+0x41435) [0xb7226435]
  9: /usr/lib/libparted.so.0(+0x422b7) [0xb72272b7]
  8: /usr/lib/libparted.so.0(+0x436aa) [0xb72286aa]
  7: /usr/lib/libparted.so.0(+0x1028e) [0xb71f528e]
  6: /usr/lib/libparted.so.0(ped_disk_add_partition+0x271) [0xb71f8bf1]
  5: /usr/lib/libparted.so.0(+0x450e3) [0xb722a0e3]
  4: /usr/lib/libparted.so.0(+0x4530e) [0xb722a30e]
  3: /usr/lib/libparted.so.0(ped_disk_new+0x79) [0xb71f9a29]
  2: /usr/lib/libDrakX/auto/c/stuff/stuff.so(XS_c__stuff_get_disk_type+0x1ed) [0xb7275fd5]
  1: /usr/lib/perl5/5.10.1/i386-linux-thread-multi/CORE/libperl.so(Perl_pp_entersub+0x56e) [0xb773d1ee]
A bug has been detected in GNU Parted.  Refer to the web site of parted http://www.gnu.org/software/parted/parted.html for more information of what could be useful for bug submitting!  Please email a bug report to bug-parted@gnu.org containing at least the version (2.3) and the following message:  Assertion (head_size <= 63) at dos.c:661 in function probe_partition_for_geom() failed.
Aborted
[root@localhost ~]#

Is it possible that using fdisk-old to partition drive that problem will occur when trying to install latest pclos systems?  Hmmm interesting but beyond me at the moment.

I am out of ideas??  Something about geometry?

Code: [Select]

grub> geometry (hd0)
drive 0x80: C/H/S = 9729/255/63, The number of sectors = 156301488, /dev/sda
   Partition num: 0,  Filesystem type is fat, partition type 0xb
   Partition num: 4,  Filesystem type is fat, partition type 0xb
   Partition num: 5,  Filesystem type is fat, partition type 0xb
   Partition num: 6,  Filesystem type is ext2fs, partition type 0x83

grub>


[Old-Polack]

No Linux did not install, same problem as in the beginning, but partition order is nice  


Trying to install KDE-mini

Code: [Select]

[root@localhost ~]# draklive-install
/media/ user
Backtrace has 11 calls on stack:
  11: /usr/lib/libparted.so.0(ped_assert+0x29) [0xb71f08b9]
  10: /usr/lib/libparted.so.0(+0x41435) [0xb7226435]
  9: /usr/lib/libparted.so.0(+0x422b7) [0xb72272b7]
  8: /usr/lib/libparted.so.0(+0x436aa) [0xb72286aa]
  7: /usr/lib/libparted.so.0(+0x1028e) [0xb71f528e]
  6: /usr/lib/libparted.so.0(ped_disk_add_partition+0x271) [0xb71f8bf1]
  5: /usr/lib/libparted.so.0(+0x450e3) [0xb722a0e3]
  4: /usr/lib/libparted.so.0(+0x4530e) [0xb722a30e]
  3: /usr/lib/libparted.so.0(ped_disk_new+0x79) [0xb71f9a29]
  2: /usr/lib/libDrakX/auto/c/stuff/stuff.so(XS_c__stuff_get_disk_type+0x1ed) [0xb7275fd5]
  1: /usr/lib/perl5/5.10.1/i386-linux-thread-multi/CORE/libperl.so(Perl_pp_entersub+0x56e) [0xb773d1ee]
A bug has been detected in GNU Parted.  Refer to the web site of parted http://www.gnu.org/software/parted/parted.html for more information of what could be useful for bug submitting!  Please email a bug report to bug-parted@gnu.org containing at least the version (2.3) and the following message:  Assertion (head_size <= 63) at dos.c:661 in function probe_partition_for_geom() failed.
Aborted
[root@localhost ~]#

Is it possible that using fdisk-old to partition drive that problem will occur when trying to install latest pclos systems?  Hmmm interesting but beyond me at the moment.

I am out of ideas??  Something about geometry?

Code: [Select]

grub> geometry (hd0)
drive 0x80: C/H/S = 9729/255/63, The number of sectors = 156301488, /dev/sda
   Partition num: 0,  Filesystem type is fat, partition type 0xb
   Partition num: 4,  Filesystem type is fat, partition type 0xb
   Partition num: 5,  Filesystem type is fat, partition type 0xb
   Partition num: 6,  Filesystem type is ext2fs, partition type 0x83

grub>


There is a bug in parted, and I believe you are seeing it. It's supposed to be backwards compatible with existing partitions from older versions, but seems to do a wally over the first partition starting at sector 63, cylinder 1. I've not encountered this myself, so far. I have the following for my /dev/sda;

[root@fatman ~]# fdisk -l /dev/sda

Disk /dev/sda: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders, total 1953525168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0000d21c

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1              63      626534      313236   83  Linux                 <-- Note the start at sector 63
/dev/sda2          626535    16820054     8096760   82  Linux swap / Solaris
/dev/sda3        16820055   114495254    48837600   83  Linux
/dev/sda4       114495255  1953520064   919512405    5  Extended
/dev/sda5       114495318   329332499   107418591   83  Linux
/dev/sda6       329332563   534434354   102550896   83  Linux
/dev/sda7       534434418   596943269    31254426   83  Linux
/dev/sda8       596943333   659452184    31254426   83  Linux
/dev/sda9       659452248   721961099    31254426   83  Linux
/dev/sda10      721961163   784470014    31254426   83  Linux
/dev/sda11      784470078   994198589   104864256   83  Linux
/dev/sda12      994198653  1623368249   314584798+  83  Linux
/dev/sda13     1623368313  1685893229    31262458+  83  Linux
/dev/sda14     1685893293  1749366044    31736376   83  Linux
/dev/sda15     1749366108  1816485614    33559753+  83  Linux
/dev/sda16     1816487663  1889887982    36700160   83  Linux
/dev/sda17     1889892080  1953520064    31813992+  83  Linux

[root@fatman ~]# fdisk -l /dev/sda -u=cylinders

Disk /dev/sda: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0000d21c

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1               1          39      313236   83  Linux                    <-- This is the equivalent of sector 63; cylider 1, first sector
/dev/sda2              40        1047     8096760   82  Linux swap / Solaris
/dev/sda3            1048        7127    48837600   83  Linux
/dev/sda4            7128      121601   919512405    5  Extended
/dev/sda5            7128       20500   107418591   83  Linux
/dev/sda6           20501       33267   102550896   83  Linux
/dev/sda7           33268       37158    31254426   83  Linux
/dev/sda8           37159       41049    31254426   83  Linux
/dev/sda9           41050       44940    31254426   83  Linux
/dev/sda10          44941       48831    31254426   83  Linux
/dev/sda11          48832       61886   104864256   83  Linux
/dev/sda12          61887      101050   314584798+  83  Linux
/dev/sda13         101051      104942    31262458+  83  Linux
/dev/sda14         104943      108893    31736376   83  Linux
/dev/sda15         108894      113071    33559753+  83  Linux
/dev/sda16         113072      117641    36700160   83  Linux
/dev/sda17         117641      121601    31813992+  83  Linux

All but the last two partitions were created with old fdisk, and I'm currently running the 64 bit test release from /dev/sda9, with two other 64 bit installations on /dev/sda15 and /dev/sda17, with an older 32 bit MiniMe installation on /dev/sda16. I had no problem installing to any of the partitions, whether created with old fdisk or new fdisk. I have not tried to install the newest MiniMe anywhere on this drive, so don't know whether the change in parted is reflected only there. I've had previous versions of LXDE and Phoenix installed on various partitions on this drive without issue, so I suspect this is a quite new problem.

If you want to try an experiment, and have a spare drive, you could see if you can install MiniMe to a single partition on it, then rsync copy the installation to the sda7 partition. If that works, you'd need to edit the /etc/fstab / partition entry on the /dev/sda7 copy to reflect the new location, and also edit the /boot/grub/menu.lst of the copy, for the same reason. Once that's done, from the liveCD, you can then use the grub native install method to install grub to the MBR of /dev/sda.

If there's nothing wrong with the partition or the format of /dev/sda7, it should then boot normally. This would eliminate any actual problems with the drive involved, and place the fault on the installer, and more particularly, the version of parted involved on the MiniMe liveCD.

[wedgetail]

old-polack
I will go hunting for a drive.  On your /dev/sda I did a quick interpartition sector step calculation. Your primary partitions have step of 1, the logical partition steps are 64 but for the two last partitions the 2048 step.  This brings back memories of similar situation on one of my drives, not giving any install problems so far. I do wonder about my CD now.

As I had a recent download of LXDE I just tried to make an install from that apart from it seemed very slow I actually got right through the process. (This may be a slower machine than I expected, only 1GB memory and 2GHz cpu).

I am going to insert the install details in code brackets for my own record (I make an archive of the thread when I finish with it). I don't know if there is anything hinting at problems in this.

Code: [Select]

[root@localhost guest]# draklive-install
/media/ user
umount: /dev/sda7: not mounted
tune2fs 1.41.14 (22-Dec-2010)
tar: ./home/guest/tmp/keyring-zlQdha/pkcs11: socket ignored
tar: ./home/guest/tmp/keyring-zlQdha/control: socket ignored
tar: ./home/guest/tmp/keyring-DCEaNk/gpg: socket ignored
tar: ./home/guest/tmp/keyring-DCEaNk/ssh: socket ignored
tar: ./home/guest/tmp/keyring-DCEaNk/control: socket ignored
tar: ./home/guest/tmp/.pcmanfm-socket--0-guest: socket ignored
tar: ./home/guest/tmp/ssh-PNidlQHY3189/agent.3189: socket ignored
tar: ./home/guest/.gvfs: Cannot stat: Permission denied
tar: ./var/run/gdm_socket: socket ignored
tar: ./var/run/acpid.socket: socket ignored
tar: ./var/run/uuidd/request: socket ignored
tar: ./var/run/saslauthd/mux: socket ignored
tar: ./var/run/dbus/system_bus_socket: socket ignored
tar: ./var/run/cups/cups.sock: socket ignored
tar: ./var/run/avahi-daemon/socket: socket ignored
tar: Exiting with failure status due to previous errors
error: package kde4-logininfo is not installed
Warning: syslog is needed by crond in runlevel 2
Warning: syslog is needed by crond in runlevel 3
Warning: syslog is needed by crond in runlevel 4
Warning: syslog is needed by crond in runlevel 5
Warning: syslog is needed by crond in runlevel 2
Warning: syslog is needed by crond in runlevel 3
Warning: syslog is needed by crond in runlevel 4
Warning: syslog is needed by crond in runlevel 5
Warning: syslog is needed by crond in runlevel 2
Warning: syslog is needed by crond in runlevel 3
Warning: syslog is needed by crond in runlevel 4
Warning: syslog is needed by crond in runlevel 5
Warning: syslog is needed by crond in runlevel 2
Warning: syslog is needed by crond in runlevel 3
Warning: syslog is needed by crond in runlevel 4
Warning: syslog is needed by crond in runlevel 5
Warning: syslog is needed by crond in runlevel 2
Warning: syslog is needed by crond in runlevel 3
Warning: syslog is needed by crond in runlevel 4
Warning: syslog is needed by crond in runlevel 5
Warning: syslog is needed by crond in runlevel 2
Warning: syslog is needed by crond in runlevel 3
Warning: syslog is needed by crond in runlevel 4
Warning: syslog is needed by crond in runlevel 5
Gtk-CRITICAL **: gtk_cell_view_set_cell_data: assertion `cell_view->priv->displayed_row != NULL' failed at /usr/lib/libDrakX/interactive/gtk.pm line 522.
Gtk-CRITICAL **: gtk_cell_view_set_cell_data: assertion `cell_view->priv->displayed_row != NULL' failed at /usr/lib/libDrakX/interactive/gtk.pm line 522.


Gtk-CRITICAL **: gtk_cell_view_set_cell_data: assertion `cell_view->priv->displayed_row != NULL' failed at /usr/lib/libDrakX/interactive/gtk.pm line 522.
Gtk-CRITICAL **: gtk_cell_view_set_cell_data: assertion `cell_view->priv->displayed_row != NULL' failed at /usr/lib/libDrakX/interactive/gtk.pm line 522.
Gtk-CRITICAL **: gtk_cell_view_set_cell_data: assertion `cell_view->priv->displayed_row != NULL' failed at /usr/lib/libDrakX/interactive/gtk.pm line 522.
Gtk-CRITICAL **: gtk_cell_view_set_cell_data: assertion `cell_view->priv->displayed_row != NULL' failed at /usr/lib/libDrakX/interactive/gtk.pm line 522.
error reading information on service acpi: No such file or directory
Stopping acpi daemon:                                          [FAILED]
Starting acpi daemon:                                          [  OK  ]


    GNU GRUB  version 0.97  (640K lower / 3072K upper memory)

 [ Minimal BASH-like line editing is supported.  For the first word, TAB
   lists possible command completions.  Anywhere else TAB lists the possible
   completions of a device/filename. ]
grub> root (hd0,6)
 Filesystem type is ext2fs, partition type 0x83
grub> setup --stage2=/boot/grub/stage2 (hd0)
 Checking if "/boot/grub/stage1" exists... yes
 Checking if "/boot/grub/stage2" exists... yes
 Checking if "/boot/grub/e2fs_stage1_5" exists... yes
 Running "embed /boot/grub/e2fs_stage1_5 (hd0)"...  17 sectors are embedded.
succeeded
 Running "install --stage2=/boot/grub/stage2 /boot/grub/stage1 (hd0) (hd0)1+17 p (hd0,6)/boot/grub/stage2 /boot/grub/menu.lst"... succeeded
Done.
grub> quit

umount: /mnt/install: device is busy.
        (In some cases useful info about processes that use
         the device is found by lsof(8) or fuser(1))
umount: /mnt/install: device is busy.
        (In some cases useful info about processes that use
         the device is found by lsof(8) or fuser(1))

What I am going to try is to install KDE-mini to an empty partition on my present big machine, this is a quick process.

[wedgetail]

Oooops, that was quick. No go, so the problem is not with my drive.
I am using a CD-RW and on occasions I have had to do a deep format. I had better check the disk.

I decided to cut a new real CD, confident that would solve the problem, but no I can not instal KDE-mini from LiveCD, seems to run fine but as soon as I start trying install. The indicated failure.  I did the proper md5sum check during cut to make sure.

Will go back and try on a CD-RW using KDE-mini 2011.07 instead. Also I may have to look at the aarnet mirror perhaps. (Funny not in LXDE iso so I will install it myself)

What should the md5sum be for the absolute latest KDE-mini 2011.9 ?  I assume this must be a safe indicator

Right now I am using Firefox in LXDE on the hard drive that was supposedly on the blink.  Not updated yet though.

Edit:

KDE-mini, iso's 2011.07 and 2011.09 when burnt to CD will run fine but not install for me.  I am at a loss as I feel sure that I installed at least from the 2011.07 but I am now getting doubtful.

How can I tell what draklive-install I am using? Some rpm wizardry?

Well it seems Synaptic can be very helpful, I searched for draklive-install then clicked on prperties.

Version 1.30  4pclos2011

The same one was used LXDE

[wedgetail]

I accidentally found an unmarked disk with a KDE version 2011 and tried to install this on top of LXDE.  It was rather slow but given time it succeeded.

KDE 2011.06   <-------- CD was unmarked, do not recognise the desktop either, so this must be KDE see below in as post how I further determined what OS must have been on this unmarked CD

While live CD was running I checked in Synaptic for draklive-install
 
draklive-install   1.30   1pclos2011

Guess it raises the question, how can I use this version installer on the KDE-mini ISO  

This is written using Firefox, on Installed but not updated KDE on the "faulty" hard drive.    


This is written using Firefox, on Installed but not updated KDE on the

[AS]

I accidentally found an unmarked disk with a KDE version 2011 and tried to install this on top of LXDE.  It was rather slow but given time it succeeded.

KDE   <---- CD was unmarked, do not recognise the desktop

while running the liveCD, try:
ls -l /boot

should give you an hint about the remaster date ...