3
1
I have a failed IDE hard drive. It shows his size as 8GB, but the drive was a 200GB. It allows to read data from first 8 GB, but most valued data is at 10..200 GB. What does this symptom mean? How can I restore full access to drive?
I need to recover data from it, not to use it again.
I also have the same model hard drive which was produced at same time and works ok for all 200GB.
UPDATE: the failed harddrive is in ATA0 PIO0 mode only.
osgx
Posted 2011-06-04T21:45:36.393
Reputation: 5 419
3
It almost sounds to me like logical block addressing (LBA) isn't working. Back in the late Pentium era, 8GB was the limit for disk size unless the disk and BIOS supported LBA or INT 13H. I suppose if the controller is damaged the disk might failsafe to CHS mode still. Sectors are still 512 bytes, after all (unless you've got a very new drive using Advanced Format technology).
You could possibly recover the data by replacing the disk controller from a disk drive of the same make and model (and lot, if possible).
Bacon Bits
Posted 2011-06-04T21:45:36.393
Reputation: 6 125
It is a very recent linux; machine is P4, but I have two disks of the same model, both are plugged into the same ide controller (SiI0680A), linux uses one disk fully and another - only for 8 gb. – osgx – 2011-06-04T23:20:48.153
Can it be a failure of pci ide controller Sil0680? – osgx – 2011-06-06T15:04:19.323
It could be. The obvious way to check that is to try the disk in a known-good machine (with a different data cable for good measure). – Bacon Bits – 2011-06-07T03:39:00.587
I'll try onboard ICH5. – osgx – 2011-06-07T03:59:23.923
I don't mean the motherboard's disk controller has failed. I mean the controller card on the disk has possibly failed. The PCB on the disk drive itself. Obviously you should try the disk in another computer first, but it is possible to replace the controller card on a disk drive if you use the same make and model. You sacrifice the second disk to do this, and you should only do it long enough to recover the data. It may not help at all, however, if the failed component is internal to the disk. If this is too risky, then I would shop around for a data recovery service. – Bacon Bits – 2011-06-07T23:26:28.767
I understand that you said about hdd onboard controller, but is it possible, that hdd's controller is good and bad is PCI-IDE Raid controller or ide cable. – osgx – 2011-06-07T23:47:38.530
I changed ide controller - the LBA is still not available. – osgx – 2011-06-13T09:28:45.353
Sorry that didn't work. It's always possible the problem could be with the motherboard controller, but the fact that other disks work fine suggests that is not the case. I'm not sure what you could try next. The data is likely only recoverable by a professional recovery service. – Bacon Bits – 2011-06-13T23:43:39.773
I still want to replace onboard hard drive controller myself from other drive. So far, I backuped all accessible information from two similar drives and can do this. The bad drive have very slow reading speed for first 8 Gb (~150-100 kb/s). Also, It reports LBA sizes in low-level IDE protocol as begin=0 - end=0. – osgx – 2011-06-14T11:52:40.410
I checked - linux detects my bad hard drive with ATA-0 PIO0 and uses a CHS-styled access. – osgx – 2011-06-15T21:39:53.237
is this what you're looking for?:http://www.howtogeek.com/howto/15761/recover-data-like-a-forensics-expert-using-an-ubuntu-live-cd/
– kalaracey – 2011-06-04T22:02:01.817Not. I know about filesystems (especially usf1 i have on drive) and mbr and gpt a lot, and the partitions are untouched, data is correct. Harddrive give full access to first 8gb (I sucesfully mounted an 256-MB filesystem from this area), But harddrive give no access to the data with address more than 8gb (I have a partition with size 10 gb just after the first, i can mount it in 'ro', but not to access any data, which physically is at 8+gb, even
lsof fs root have 6 i/o error fails). – osgx – 2011-06-04T22:11:29.380