Help! Nedd to recover files

[cbagger01]

PC partitions can be set up in 2 different ways.

1) As a primary partition, or a "Real" partition. You can only have 4 primary partitions on a hard drive
2) As an "extended" or "logical" partition. This is basically a partition inside a "Real" partition. You can have MANY logical partitions.. more than a dozen. But all of these extended partitions are stored inside a "Real" partition.

In Linux, primary partitions are identified by /dev/devicenameX , where X = 1 through 4
The extended partition names begin at X = 5

So my guess is that your Disk director created partition table looks like this:

sda1 (Primary, FAT/FAT32)
sda2 (Primary, NTFS most likely your "C:" drive)
sda3 (placeholder for extended partition, not a "Real" partition)
and inside sda3:

sda5 (Extended, NTFS, most likely your "E:" drive)

I would not trust the gpart generated information because it looks incorrect. For some reason it saw your "E:" drive as a primary partition and not an extended one. This explains the gap in the last sector of partition #2 and the first sector of partition #3

Chances are that your "Disk Director" has already fixed your partition table correctly, but if you wish to use a different program to try and recover your partition table, I would try "TESTDISK" and not gpart.

Ironically, you can get a version of TESTDISK from the bootable "Gparted" livecd. Just Right-Click on an empty part of the desktop and then choose TESTDISK from the menu. Another option is to burn a copy of the ULTIMATE BOOT CD. I am pretty sure that it contains TESTDISK.

But I would not use gpart to WRITE out a new partition table because the information looks inconsistent to me.

Good Luck.

[xpcomputers]

What is the best way to check the data that Disk Director has already written to the partition table? Basically I want to check that the partition data in the partition table isn't screwy already (like the stuff that gpart was suggesting), in case that is messing up my file system.

You are right about the extended partition. That is how it is laid out. Thanks for giving me the info on the numbering system, it makes much more sense now.

Thanks for taking the time to reply, I'm learning heaps here!

Mike

[xpcomputers]

If I was to try to write another new partition table, would it ignore the existing one, and look for the possible partition start and end points itself, or would it see the existing partition table and just use that?

I had no joy getting Testdisk to work. Via Knoppix it said the libntfs.so.9 (or something like that) wasn't found. I had it on another bootcd, but it wouldn't run there either.

Am trying DiskPatch v2.1.1 it has listed a whole load of potential partitions, but it is difficult to figure out which 2 of the bottom 6 are the correct ones....

SEQ TYPE START--- (GB) LENGTH---(GB) END-----(GB) PRI RAT
01 FAT32 63 (0) 9221247 (4.4) 9221309 (4.4) PRI #
02 NTFS 9221310 (4.4) 165019680 (78.69) 174240989 (83.08) PRI -
03 NTFS 9221310 (4.4) 165019673 (78.69) 174240982 (83.08) PRI +
04 NTFS 9221317 (4.4) 165019673 (78.69) 174240989 (83.08) PRI +
05 NTFS 174241053 (83.08) 145918332 (69.58) 320159384 (152.66) LOG -
06 NTFS 174241053 (83.08) 145918332 (69.58) 320159381 (152.66) LOG +
07 NTFS 174241056 (83.08) 145918329 (69.58) 320159384 (152.66) LOG +

Obviously the 63 figure in the gpart readout means something (is it the partition ending on a sector boundary?). Is there an easy way that I can use a calculator to get an equivalent reading from the above data? Am I just adding 1, then dividing the numbers by 64 to see if I get no remainder?

What is the best way to get that data off the current set of partition in the recovered partition table that Disk Director gave me?

I would rather use TestDisk on Knoppix though, if someone can tell me how to overcome the lib error. I am using Knoppix 5.1.1 DVD version.

Thanks

[Harry Kuhman]

If I was to try to write another new partition table, would it ignore the existing one, and look for the possible partition start and end points itself, or would it see the existing partition table and just use that?....

The partition table is a fixed size and position, always the last part of sector 0 track 0 cylinder 0 of the hard disk. So if you write a new partition table, the old one is no longer there. The is no chance for any doubt if the old one will be used; it's gone.

[xpcomputers]

If I was to try to write another new partition table, would it ignore the existing one, and look for the possible partition start and end points itself, or would it see the existing partition table and just use that?....

The partition table is a fixed size and position, always the last part of sector 0 track 0 cylinder 0 of the hard disk. So if you write a new partition table, the old one is no longer there. The is no chance for any doubt if the old one will be used; it's gone.

Thanks Harry,

I actually meant, does the scan search the partition starts/ends or will it just use the partition table if it exists. You actually indirectly answered me, as it can't write a new one if it hasn't created something different! So it must scan for partition start/ends and writing this info to a new partition table, which obviously will obliterate the old one.

Mike