Lost half of my harddrive...?

Firestone · 20 August 2013, 01:53

Got a really odd problem here with my A4000. It's set up with AmigaOS3.9 around 10 years ago, and it has worked ever since.... until lately.

Two of my partitions suddenly became listed on my Workbench as HD4:unreadable and HD5:unreadable.

After reading some docs around on Aminet, I tried to recover it with RDBrecov from Aminet. I've had success with this tool in the past, and it's great!
Partitions are formatted as PFS2 btw.

Well.... RDBrecov requires that I know how many cylinders there are, bytes pr block and so on.... and to know this it requires me to open HDToolBox and try to read the configuration from the disk, but not saving anything.
HDToolBox sais that the RDB may be corrupted and asks me if I want to define a new rdb, or edit the existing one. I tried to edit it, and wrote down all values before I pressed the read configuration from disk just in case, and to compare the results.
The differences were huge... The disk is a 4.2GB Fujitisudisk.
When I pressed read configuration it says that It's only 1,9GB in HDToolbox....

RDBRecov only recognizes the first three partitions which is inside the 1,9GB limit obviously.....

I THINK I have seen this problem in the last with a disk I had in one of my PCs (Seagate I think) running Windows95 or 98, can't remember exactly. With that this, exactly the same happened. Half the size, and the system recognized as just that, and I was able to partition it and use it just like before, only with that limited space.....

Any tips on how to recover those two mission partitions..?

thomas · 20 August 2013, 11:35

RDBRecov is the wrong utility IMHO.

Please first run check4gb, save the report and post it here.

Then please run pfssalv2 on each of the damaged partitions and check if it finds most of the files. Save the log of each session and attach them here.

Finally you might give RDBRecov another try. First run it with CS=1 and C=10000. When it finds the first partition, press Ctrl-C. Now divide the cylinder number of the first partition by 2 and use the result as CS in another run.

Like this (you have to adjust the numbers for your HDD):

Code:

4> ram:rdbrecov uaehf.device 0 cs 1 c 10000
Trying to open uaehf.device unit 0
SectorSize   = 512
Cylinders    = 1014     (overridden by parameter: 10000)
CylSectors   = 1009     (overridden by parameter: 1)
Allocated partition table for 100 entries.
SFS Bootblock found at cylinder  2018: Size =103927 cyls DosType=53465300 Name="Workbench"  BlockSize=512 
Reading cylinder  4999
*** Break
4> ram:rdbrecov uaehf.device 0 cs 1009 c 10000
Trying to open uaehf.device unit 0
SectorSize   = 512
Cylinders    = 1014     (overridden by parameter: 10000)
CylSectors   = 1009     (overridden by parameter: 1009)
Allocated partition table for 100 entries.
using NSD style commands
SFS Bootblock found at cylinder     2: Size =  103 cyls DosType=53465300 Name="Workbench"  BlockSize=512 
SFS Bootblock found at cylinder   105: Size =  909 cyls DosType=53465300 Name="Work"       BlockSize=512 
Reading cylinder  1015
TD_READ (1) Error = -4 (invalid length)
Checking FFS partitions
4>

Also cut the entire log from the shell window and paste it here please.

And because HDToolbox says the RDB is corrupted, please run ReadRDB and paste its output here, too. It should look similar to this:

Code:

4> ram:readrdb uaehf.device 0
Bytes per block on device: 512
Rigid Disk Block:
   Vendor          : WUAE    
   Product         : hd31.vhd        
   Revision        : 0.3 
   Cylinders       :    1014
   Heads           :       1
   Sectors         :    1009
   Bytes per block :     512
   Last LUN
Partitions:
   WDH0: bootable
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 2
      High Cyl    = 104
      Buffers     = 30
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x1FE00
      Boot prio   = 0
      DOS type    = 0x53465300
   WDH1:
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 105
      High Cyl    = 1013
      Buffers     = 150
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x1FE00
      Boot prio   = 0
      DOS type    = 0x53465300
File systems:
   0x53465300   1.279 
4>

Firestone · 20 August 2013, 17:06

Output from 4GB:

Code:

New Shell process 7
 
Current Shell process initiated at: Tuesday 20-Aug-13 16:45:22 @656
 
This is the 232 day of the year, there are 133 left.

7.Work:> ram:
7.Ram Disk:> Check4GB scsi.device
scsi.device unit 0 supports NSD commands
scsi.device unit 0 supports SCSI commands
scsi.device unit 1 supports NSD commands
7.Ram Disk:> Check4GB hd0:
Name     Volume       Size Device       Unit Version  DosType  Version   Big Check
HD0      System       149M scsi            0 43.43    PFS2     17.8       no ok
7.Ram Disk:> Check4GB hd1:
Name     Volume       Size Device       Unit Version  DosType  Version   Big Check
HD1      Work        1008M scsi            0 43.43    PFS2     17.8       no ok
7.Ram Disk:> Check4GB hd2:
Name     Volume       Size Device       Unit Version  DosType  Version   Big Check
HD2      Music       1003M scsi            0 43.43    PFS2     17.8       no ok
7.Ram Disk:> Check4GB hd3:
Name     Volume       Size Device       Unit Version  DosType  Version   Big Check
HD3      CFS       999M scsi            0 43.43    PFS2     17.8       no ok
7.Ram Disk:> Check4GB hd4:
Name     Volume       Size Device       Unit Version  DosType  Version   Big Check
HD4      CFS       595M scsi            0 43.43    PFS2     17.8       no ok
7.Ram Disk:> Check4GB mhd:
Name     Volume       Size Device       Unit Version  DosType  Version   Big Check
MHD      CFS       319M scsi            0 43.43    RESV                no ok
7.Ram Disk:> info

Mounted disks:
Unit       Size     Used     Free Full Errs   Status   Name
ENV:        27K       23        4  85%   0  Read/Write Env
RAM:        48K       48        0 100%   0  Read/Write Ram Disk
HD0:       138M   276110     7038  97%   0  Read/Write System
DF0:       837K      377     1381  21%   0  Read/Write Empty
HD1:       945M  1785959   150713  92%   0  Read/Write Work
HD2:       940M  1821895   105024  95%   0  Read/Write Music
HD3:      Unreadable disk
HD4:      Unreadable disk

Volumes available:
Ram Disk [Mounted]
Music [Mounted]
Work [Mounted]
Empty [Mounted]
System [Mounted]
7.Ram Disk:>

Output from PFSSalv2: (Goes on like this forever with one of the damaged partitions.....)

Code:

opening scsi.device unit 0
partition HD4
low cylinder 6860
high cylinder 8152
blocks per cyl 945
drive block size 512 bytes
file system block size 512 bytes
first block 6482700
no. of blocks 1221885
max transfer 128 kb (0x20000)
first reserved 2
last reserved 47472
Pass 1 of 2: Scanning for anode blocks...
*** read error 3 in block 2
*** read error 3 in block 256
*** read error 3 in block 258
*** read error 3 in block 260
*** read error 3 in block 262
*** read error 3 in block 264
*** read error 3 in block 266
*** read error 3 in block 268
*** read error 3 in block 270
found 0 anode blocks

Output from ReadRDB:

Code:

7.Ram Disk:> ReadRDB scsi.device 0
Bytes per block on device: 512
Rigid Disk Block:
   Vendor          : FUJITSU 
   Product         : MPD3043AT       
   Revision        : DD-0
   Cylinders       :    8940
   Heads           :      15
   Sectors         :      63
   Bytes per block :     512
   Last disk
   Last LUN
   Last ID
Partitions:
   HD0: bootable
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 2
      High Cyl    = 325
      Buffers     = 30
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x20000
      Boot prio   = 0
      DOS type    = 0x50465302
   HD1:
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 326
      High Cyl    = 2513
      Buffers     = 30
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x20000
      Boot prio   = 0
      DOS type    = 0x50465302
   HD2:
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 2514
      High Cyl    = 4690
      Buffers     = 30
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x20000
      Boot prio   = 0
      DOS type    = 0x50465302
   HD3:
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 4691
      High Cyl    = 6859
      Buffers     = 30
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x20000
      Boot prio   = 0
      DOS type    = 0x50465302
   HD4:
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 6860
      High Cyl    = 8152
      Buffers     = 30
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x20000
      Boot prio   = 0
      DOS type    = 0x50465302
   MHD:
      Block size  = 512
      Reserved    = 2
      Low Cyl     = 8153
      High Cyl    = 8846
      Buffers     = 5
      Mask        = 0x7FFFFFFE
      MaxTransfer = 0x20000
      Boot prio   = 0
      DOS type    = 0x52455356
File systems:
   0x444F5303  40.1   
   0x50465302  17.8   
7.Ram Disk:>

Output from RDBRecov (two passes, last one with cylindernumber devided by 2 as instructed

Code:

8.Ram Disk:> RDBrecov scsi.device 0 CS 1 c 10000
Trying to open scsi.device unit 0
SectorSize   = 512
Cylinders    = 4092     (overridden by parameter: 10000)
CylSectors   = 1008     (overridden by parameter: 1)
Allocated partition table for 100 entries.
PFS Bootblock found at cylinder  1890: Size =306180 cyls DosType=50465301 Name="System"
Reading cylinder 10000
Checking FFS partitions
Complete.

8.Ram Disk:> RDBrecov scsi.device 0 CS 945 c 10000
Trying to open scsi.device unit 0
SectorSize   = 512
Cylinders    = 4092     (overridden by parameter: 10000)
CylSectors   = 1008     (overridden by parameter: 945)
Allocated partition table for 100 entries.
using NSD style commands
PFS Bootblock found at cylinder     2: Size =  324 cyls DosType=50465301 Name="System"
PFS Bootblock found at cylinder   326: Size = 2188 cyls DosType=50465301 Name="Work"
PFS Bootblock found at cylinder  2514: Size = 2177 cyls DosType=50465301 Name="Music"
Reading cylinder  4365
TD_READ (1) Error = 3
Checking FFS partitions
8.Ram Disk:>

To me this seems to be a hardwareproblem rather than a softwareproblem, don`t you think?

thomas · 20 August 2013, 18:18

Yes, it seems so. But this error 3 is suspicious. Error numbers usually start at 20 or are negative. I cannot find any reference to error 3.

Please try this:

- switch off the Amiga
- hold down both mouse buttons and switch the Amiga on again
- when the early startup menu appears, click on "boot without startup sequence"
- in the CLI window enter version scsi.device to make sure that it's V40 and not V43.
- then try RDBrecov again.

If it's still limited to 1.9GB then it seems like the firmware of the drive has reduced its own size for some reason. This would also explain why HDtoolbox said that the RDB is corrupted: it defines the drive bigger than it actually is (now).

BTW, you could have run Check4GB without arguments. Then a GUI would have opened with a save button which allows you to save a complete report easily.

Firestone · 21 August 2013, 12:18

Tried to boot with no startup sequence now and verified that SCSI.device is version 40.12 before any rom-patches are applied.

Ran rdbrecov scsi.device 0 cs 1 c 10000:

Code:

Trying to open scsi.device unit 0
TD_GETGEOMETRY Error = -3 (unsupported command)

thomas · 21 August 2013, 14:46

Add S 512 to the command line. And you can start with CS 945 immediately. This result should never change.

Firestone · 23 August 2013, 01:13

Got same error.....

I'm thinking about if it is possible to change the electronics containing the firmware on that drive with one identical one if I were lucky finding one..?

Firestone · 23 August 2013, 03:58

Tonight I thought I would try to experiment more with this shitty drive...
Placed it in an external enclosure via USB and connected it to the Mac just for fun to see what size it was recognized at. 4,3GB! Whey!

Started WinUAE via VMWare on my Mac and tried to boot with no startup-sequence, and woala! Those two missing partions were readable and all content were in perfect shape!

Well, now is the question : Why on earth is the disk detected as an 1,9GB drive in my A4000 and 4,32GB on my Mac?

Thanks for all your help thomas btw!

thomas · 23 August 2013, 08:05

Quote:

Originally Posted by Firestone

Got same error.....

Please use the version from my homepage: http://thomas-rapp.homepage.t-online...s/rdbrecov.lha
The one on Aminet seems to be old.

Quote:

Originally Posted by Firestone

Started WinUAE via VMWare on my Mac and tried to boot with no startup-sequence, and woala! Those two missing partions were readable and all content were in perfect shape!

What about starting *with* startup-sequence?
Did you connect the drive to IDE0 or UAE controller?
Were the missing partitions available on the real Amiga with scsi.device V40 without startup-sequence?

Quote:

Well, now is the question : Why on earth is the disk detected as an 1,9GB drive in my A4000 and 4,32GB on my Mac?

I still think it could be the IDE driver. Did you install something related recently? IDEfix? SpeedUP? PIO2 patch?

Firestone · 27 August 2013, 17:57

Nothing.... Got it all working a few days ago, even on the A4000.
The only change I did to make it work again was to move the jumper to Master again, everything then just showed up fine.....

thomas · 27 August 2013, 18:06

Perhaps you accidentally plugged the jumper into the capacity limit place instead of Slave.

Firestone · 27 August 2013, 23:46

Yes, thats probably it!
Why did they make this capacity-jumper in the first place?

thomas · 28 August 2013, 08:05

Because a manufacturer cannot afford to keep supporting ancient drives. It's cheaper to add compatibility features to the new drives.

20 August 2013, 01:53	#1
Firestone Registered User Join Date: Apr 2013 Location: Norway Posts: 258	Lost half of my harddrive...? Got a really odd problem here with my A4000. It's set up with AmigaOS3.9 around 10 years ago, and it has worked ever since.... until lately. Two of my partitions suddenly became listed on my Workbench as HD4:unreadable and HD5:unreadable. After reading some docs around on Aminet, I tried to recover it with RDBrecov from Aminet. I've had success with this tool in the past, and it's great! Partitions are formatted as PFS2 btw. Well.... RDBrecov requires that I know how many cylinders there are, bytes pr block and so on.... and to know this it requires me to open HDToolBox and try to read the configuration from the disk, but not saving anything. HDToolBox sais that the RDB may be corrupted and asks me if I want to define a new rdb, or edit the existing one. I tried to edit it, and wrote down all values before I pressed the read configuration from disk just in case, and to compare the results. The differences were huge... The disk is a 4.2GB Fujitisudisk. When I pressed read configuration it says that It's only 1,9GB in HDToolbox.... RDBRecov only recognizes the first three partitions which is inside the 1,9GB limit obviously..... I THINK I have seen this problem in the last with a disk I had in one of my PCs (Seagate I think) running Windows95 or 98, can't remember exactly. With that this, exactly the same happened. Half the size, and the system recognized as just that, and I was able to partition it and use it just like before, only with that limited space..... Any tips on how to recover those two mission partitions..?

21 August 2013, 12:18	#5
Firestone Registered User Join Date: Apr 2013 Location: Norway Posts: 258	Tried to boot with no startup sequence now and verified that SCSI.device is version 40.12 before any rom-patches are applied. Ran rdbrecov scsi.device 0 cs 1 c 10000: Code: Trying to open scsi.device unit 0 TD_GETGEOMETRY Error = -3 (unsupported command)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
THROW IN 2nd HALF: Announced	TheFoxSoft	Retrogaming General Discussion	4	24 April 2012 10:37
Problem formatting upper half of CF card	Canyon Matt	support.Hardware	3	25 February 2010 21:47
Lost Snowmen, excellent Lost Vikings clone	s2325	Retrogaming General Discussion	3	11 November 2008 11:24
RS-232 and half-duplex PPP connections	pkillo	support.Other	7	12 October 2008 19:21
Can only view half a scanned page	tim_calladine	AMR data problems	9	01 October 2008 17:22

20 August 2013, 18:18	#4
thomas Registered User Join Date: Jan 2002 Location: Germany Posts: 7,020	Yes, it seems so. But this error 3 is suspicious. Error numbers usually start at 20 or are negative. I cannot find any reference to error 3. Please try this: - switch off the Amiga - hold down both mouse buttons and switch the Amiga on again - when the early startup menu appears, click on "boot without startup sequence" - in the CLI window enter version scsi.device to make sure that it's V40 and not V43. - then try RDBrecov again. If it's still limited to 1.9GB then it seems like the firmware of the drive has reduced its own size for some reason. This would also explain why HDtoolbox said that the RDB is corrupted: it defines the drive bigger than it actually is (now). BTW, you could have run Check4GB without arguments. Then a GUI would have opened with a save button which allows you to save a complete report easily.

21 August 2013, 14:46	#6
thomas Registered User Join Date: Jan 2002 Location: Germany Posts: 7,020	Add S 512 to the command line. And you can start with CS 945 immediately. This result should never change.

23 August 2013, 01:13	#7
Firestone Registered User Join Date: Apr 2013 Location: Norway Posts: 258	Got same error..... I'm thinking about if it is possible to change the electronics containing the firmware on that drive with one identical one if I were lucky finding one..?

23 August 2013, 03:58	#8
Firestone Registered User Join Date: Apr 2013 Location: Norway Posts: 258	Tonight I thought I would try to experiment more with this shitty drive... Placed it in an external enclosure via USB and connected it to the Mac just for fun to see what size it was recognized at. 4,3GB! Whey! Started WinUAE via VMWare on my Mac and tried to boot with no startup-sequence, and woala! Those two missing partions were readable and all content were in perfect shape! Well, now is the question : Why on earth is the disk detected as an 1,9GB drive in my A4000 and 4,32GB on my Mac? Thanks for all your help thomas btw!

27 August 2013, 17:57	#10
Firestone Registered User Join Date: Apr 2013 Location: Norway Posts: 258	Nothing.... Got it all working a few days ago, even on the A4000. The only change I did to make it work again was to move the jumper to Master again, everything then just showed up fine.....

27 August 2013, 18:06	#11
thomas Registered User Join Date: Jan 2002 Location: Germany Posts: 7,020	Perhaps you accidentally plugged the jumper into the capacity limit place instead of Slave.

27 August 2013, 23:46	#12
Firestone Registered User Join Date: Apr 2013 Location: Norway Posts: 258	Yes, thats probably it! Why did they make this capacity-jumper in the first place?

28 August 2013, 08:05	#13
thomas Registered User Join Date: Jan 2002 Location: Germany Posts: 7,020	Because a manufacturer cannot afford to keep supporting ancient drives. It's cheaper to add compatibility features to the new drives.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)