D
Dan_Musicant
On Mon, 20 Feb 2006 09:32:05 +1100, "Rod Speed"
:> On Sun, 19 Feb 2006 10:02:19 +1100, "Rod Speed"
:>
:>>> On Fri, 17 Feb 2006 21:07:53 +0000, Mike Tomlinson
:>>>
:>>>> In article <[email protected]>,
:>>>>
:>>>>> I have another HD in the box that's 200 GB and it has one logical
:>>>>> drive only, and is formatted NTFS. That's obviously not the
:>>>>> problem. Also, the drive was working fine for 1.5 years and I made
:>>>>> no changes.
:>>>>
:>>>> Read what Eric wrote. The drive doesn't get trashed _until_ data
:>>>> has been written past the 137GB boundary, so the drive can appear
:>>>> to be working well for quite some time, until you fill the disk up
:>>>> enough.
:>>>>
:>>>>> The logical
:>>>>> drives simply disappeared from it, evidently data corruption.
:>>>>
:>>>> Which is exactly what writing past 137GB does. The write "wraps
:>>>> around" to cylinder 0. What's on cylinder 0? The boot sector and
:>>>> partition tables - bye bye logical drives.
:>>>>
:>>>> See http://www.48bitlba.com/
:>>>
:>>> Thanks. I'm checking out the site.
:>>>
:>>> One thing I didn't mention and it could well have a bearing here:
:>>>
:>>> 2-3 days before the corruption occurred I removed my Promise
:>>> Ultra100 TX2 PCI IDE Controller card from the system in order to
:>>> free up a PCI slot. I had been running over 4 IDE devices, so I
:>>> removed all but 4, including the 3 IDE HD's and my DVD burner. The
:>>> problem drive may well have been on the card and not the MB IDE
:>>> channels.
:>>
:>> Very likely and was getting 48bit LBA support from the driver for
:>> that card.
:>>
:>>> It's currently primary slave.
:>>
:>> Then set the EnableBigLBA value to 1 in the registry.
:
:> Hmm. I guess you are supposed to enter 1, not 0x1 like it says.
:
:Yeah, that's just a formal way of specifying the base with a number.
:
oesnt matter with a value of 1.
:
:> Really cute. Argh.
:
:Yeah, could be clearer.
:
However, if what you said is right, I have to delete my entry and do it
again. I used DWORD, and you say to do binary. Are you sure about that?
:> On Sun, 19 Feb 2006 10:02:19 +1100, "Rod Speed"
:>
:>>> On Fri, 17 Feb 2006 21:07:53 +0000, Mike Tomlinson
:>>>
:>>>> In article <[email protected]>,
:>>>>
:>>>>> I have another HD in the box that's 200 GB and it has one logical
:>>>>> drive only, and is formatted NTFS. That's obviously not the
:>>>>> problem. Also, the drive was working fine for 1.5 years and I made
:>>>>> no changes.
:>>>>
:>>>> Read what Eric wrote. The drive doesn't get trashed _until_ data
:>>>> has been written past the 137GB boundary, so the drive can appear
:>>>> to be working well for quite some time, until you fill the disk up
:>>>> enough.
:>>>>
:>>>>> The logical
:>>>>> drives simply disappeared from it, evidently data corruption.
:>>>>
:>>>> Which is exactly what writing past 137GB does. The write "wraps
:>>>> around" to cylinder 0. What's on cylinder 0? The boot sector and
:>>>> partition tables - bye bye logical drives.
:>>>>
:>>>> See http://www.48bitlba.com/
:>>>
:>>> Thanks. I'm checking out the site.
:>>>
:>>> One thing I didn't mention and it could well have a bearing here:
:>>>
:>>> 2-3 days before the corruption occurred I removed my Promise
:>>> Ultra100 TX2 PCI IDE Controller card from the system in order to
:>>> free up a PCI slot. I had been running over 4 IDE devices, so I
:>>> removed all but 4, including the 3 IDE HD's and my DVD burner. The
:>>> problem drive may well have been on the card and not the MB IDE
:>>> channels.
:>>
:>> Very likely and was getting 48bit LBA support from the driver for
:>> that card.
:>>
:>>> It's currently primary slave.
:>>
:>> Then set the EnableBigLBA value to 1 in the registry.
:
:> Hmm. I guess you are supposed to enter 1, not 0x1 like it says.
:
:Yeah, that's just a formal way of specifying the base with a number.
:
oesnt matter with a value of 1.
:
:> Really cute. Argh.
:
:Yeah, could be clearer.
:
However, if what you said is right, I have to delete my entry and do it
again. I used DWORD, and you say to do binary. Are you sure about that?