If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
I formatted that bad drive
I'm speaking about that bad drive (partition) that caused me to lose a
lot of data and could not be repaired. I finally re-downloaded as much of the data as I could after saving what was salvagable from the drive. I have now recreated the original drive, minus around 15 or 20 files that can not be replaced. Since I no longer needed to read that partition, and it contained a few folders that would not allow me to delete them, I decided to format it, before I toss it in the garbage. I did a FULL format on it, and it formatted just fine, said there were no errors or bad segments. I ran Scandisk on it (win98), and that too said "no problems". (I only used the quick scandisk, not the one that takes around 12 hours). After all that trouble, I cant believe this drive checks out good. But I wont trust it, so I am going to just trash it. But I did want to see what a Format would do to it anyhow. I never expected it to work, in fact I was waiting for the format to fail mid stream..... |
Ads |
#2
|
|||
|
|||
I formatted that bad drive
|
#3
|
|||
|
|||
I formatted that bad drive
|
#4
|
|||
|
|||
I formatted that bad drive
In message , mike
writes: On 12/10/2017 9:19 AM, Paul wrote: wrote: I'm speaking about that bad drive (partition) that caused me to lose a lot of data and could not be repaired. I finally re-downloaded as much of the data as I could after saving what was salvagable from the drive. Thanks for keeping us informed. Glad you finally reached an endpoint. I have now recreated the original drive, minus around 15 or 20 files that can not be replaced. Since I no longer needed to read that partition, and it contained a few folders that would not allow me to delete them, I decided to format it, before I toss it in the garbage. I think I would have too, just out of curiosity. I did a FULL format on it, and it formatted just fine, said there were no errors or bad segments. I ran Scandisk on it (win98), and that too said "no problems". (I only used the quick scandisk, not the one that takes around 12 hours). Perhaps do that one, overnight or something? After all that trouble, I cant believe this drive checks out good. But I wont trust it, so I am going to just trash it. But I did want to see what a Format would do to it anyhow. I never expected it to work, in fact I was waiting for the format to fail mid stream..... Paul and others: can you think of what might have been the reason for James' experiences? I'd been thinking he definitely has a physical bad patch; the fact that it was confined to one partition, with the other two continuing to work fine, supported that. But (with reservations depending on the outcome of the "long" Scandisk if he does it) it looks like it might not be that after all. _Could_ it have been just some weird form of data corruption? [] Paul I'm partial to HDDscan. Gives a graph of speed vs address. I've got drives that chkdsk says are clean, but HDDscan shows a bunch of places where the response time is very slow, presumably due to multiple reads of nearly bad sectors. Or, I think Paul has said, to where the drive's internal firmware has "swapped in" a good sector for one that's been detected as bad; the increase in response time being because extra head movements are required to access those sectors, compared to the minimal amount needed to just go through the drive in sequence as HDDscan does. And it's usually points where the speed drop has a width and a flat bottom, and obviously where it remains the same over successive runs; because of the nature of the Windows OS, there are times during a run where Windows decides to go off and do its own thing, but those should produce downward spikes in the speed graph that are narrow, and not in the same place on successive runs. (Just out of interest - I've followed these discussions, but never actually run it - how long does a scan take to produce one of these graphs? [I presume it doesn't vary _that_ much with capacity, as bigger drives tend to be faster - or?]) -- J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf Bread is lovely, don't get me wrong. But it's not cake. Or it's rubbish cake. I always thought that bread needed more sugar and some icing. - Sarah Millican (Radio Times 11-17 May 2013) |
#5
|
|||
|
|||
I formatted that bad drive
On 12/10/2017 7:49 PM, J. P. Gilliver (John) wrote:
In message , mike writes: On 12/10/2017 9:19 AM, Paul wrote: wrote: I'm speaking about that bad drive (partition) that caused me to lose a lot of data and could not be repaired. I finally re-downloaded as much of the data as I could after saving what was salvagable from the drive. Thanks for keeping us informed. Glad you finally reached an endpoint. I have now recreated the original drive, minus around 15 or 20 files that can not be replaced. Since I no longer needed to read that partition, and it contained a few folders that would not allow me to delete them, I decided to format it, before I toss it in the garbage. I think I would have too, just out of curiosity. I did a FULL format on it, and it formatted just fine, said there were no errors or bad segments. I ran Scandisk on it (win98), and that too said "no problems". (I only used the quick scandisk, not the one that takes around 12 hours). Perhaps do that one, overnight or something? After all that trouble, I cant believe this drive checks out good. But I wont trust it, so I am going to just trash it. But I did want to see what a Format would do to it anyhow. I never expected it to work, in fact I was waiting for the format to fail mid stream..... Paul and others: can you think of what might have been the reason for James' experiences? I'd been thinking he definitely has a physical bad patch; the fact that it was confined to one partition, with the other two continuing to work fine, supported that. But (with reservations depending on the outcome of the "long" Scandisk if he does it) it looks like it might not be that after all. _Could_ it have been just some weird form of data corruption? [] Paul I'm partial to HDDscan. Gives a graph of speed vs address. I've got drives that chkdsk says are clean, but HDDscan shows a bunch of places where the response time is very slow, presumably due to multiple reads of nearly bad sectors. Or, I think Paul has said, to where the drive's internal firmware has "swapped in" a good sector for one that's been detected as bad; the increase in response time being because extra head movements are required to access those sectors, compared to the minimal amount needed to just go through the drive in sequence as HDDscan does. And it's usually points where the speed drop has a width and a flat bottom, and obviously where it remains the same over successive runs; because of the nature of the Windows OS, there are times during a run where Windows decides to go off and do its own thing, but those should produce downward spikes in the speed graph that are narrow, and not in the same place on successive runs. I'm talking about access times over a second. Unlikely that those are seek times to replaced sectors. (Just out of interest - I've followed these discussions, but never actually run it - how long does a scan take to produce one of these graphs? [I presume it doesn't vary _that_ much with capacity, as bigger drives tend to be faster - or?]) I never worried about it. Just turn it on and take a nap. It's only one drive. Doesn't matter how long it takes. |
#6
|
|||
|
|||
I formatted that bad drive
On Mon, 11 Dec 2017 03:49:43 +0000, "J. P. Gilliver (John)"
wrote: (Just out of interest - I've followed these discussions, but never actually run it - how long does a scan take to produce one of these graphs? [I presume it doesn't vary _that_ much with capacity, as bigger drives tend to be faster - or?]) 4 1/2 hours for a 1GB drive (last time I did it)... but it depends on hardware specs so take that with a pinch of salt. []'s -- Don't be evil - Google 2004 We have a new policy - Google 2012 |
#7
|
|||
|
|||
I formatted that bad drive
On Mon, 11 Dec 2017 03:49:43 +0000, "J. P. Gilliver (John)"
wrote: In message , mike writes: On 12/10/2017 9:19 AM, Paul wrote: wrote: I'm speaking about that bad drive (partition) that caused me to lose a lot of data and could not be repaired. I finally re-downloaded as much of the data as I could after saving what was salvagable from the drive. Thanks for keeping us informed. Glad you finally reached an endpoint. Yep, all I have to do now is copy my rebuilt partition to the computer it belongs on. Because of Win98 not accessing many of these large external hard drives, and that computer only having USB 1.1, I did the whole rebuild on a portable drive. Now I'll plug the new drive and the portable one into my XP machine to transfer everything over. I have now recreated the original drive, minus around 15 or 20 files that can not be replaced. Since I no longer needed to read that partition, and it contained a few folders that would not allow me to delete them, I decided to format it, before I toss it in the garbage. I think I would have too, just out of curiosity. I planned to do that, just for the heck of it. I did a FULL format on it, and it formatted just fine, said there were no errors or bad segments. I ran Scandisk on it (win98), and that too said "no problems". (I only used the quick scandisk, not the one that takes around 12 hours). Perhaps do that one, overnight or something? I thought of plugging it into a spare computer to run it. I cant use this one meaning cant use the internet with that thing running, it slows down the computer so much it's not usable. But if I run the tests on an XP machine, I cant run scandisk, I have to use Chkdsk, or I could run Norton Disk Doctor (if that program works on XP.... I never tried it). Or, I suppose I could just install Win98 on a spare machine, thats easy enough to do, and I have 5 or 6 small HDDS to install it on. But this all seems senseless since I am going to trash that drive anyhow. I just cant see myself trusting it again.... After all that trouble, I cant believe this drive checks out good. But I wont trust it, so I am going to just trash it. But I did want to see what a Format would do to it anyhow. I never expected it to work, in fact I was waiting for the format to fail mid stream..... Paul and others: can you think of what might have been the reason for James' experiences? I'd been thinking he definitely has a physical bad patch; the fact that it was confined to one partition, with the other two continuing to work fine, supported that. But (with reservations depending on the outcome of the "long" Scandisk if he does it) it looks like it might not be that after all. _Could_ it have been just some weird form of data corruption? I really think Scandisk screwed it up more than anything. Scandisk seems to be helpful "Most of the time", but I am sure it can screw up too. However, I also question my IDE data cable. I initially replaced that HDD with a 160gb drive, but that was before I learned that Win98 can not support a drive larger than 120gb. (Actually 132gb). That drive repeatedly kept running scandisk, and there was very little data on it yet. I was only using it to save downloads at that point. I ordered some of the 80 wire IDE cables from ebay, just to upgrade the cables, and when I changed that cable, the problems seemed to stop/ However I also got my 120gb drive in the mail about that same time and changed to that drive instead. So, it's hard to know if the problem was the cable, or having a drive too big for the OS. [] Paul I'm partial to HDDscan. Gives a graph of speed vs address. I've got drives that chkdsk says are clean, but HDDscan shows a bunch of places where the response time is very slow, presumably due to multiple reads of nearly bad sectors. Or, I think Paul has said, to where the drive's internal firmware has "swapped in" a good sector for one that's been detected as bad; the increase in response time being because extra head movements are required to access those sectors, compared to the minimal amount needed to just go through the drive in sequence as HDDscan does. And it's usually points where the speed drop has a width and a flat bottom, and obviously where it remains the same over successive runs; because of the nature of the Windows OS, there are times during a run where Windows decides to go off and do its own thing, but those should produce downward spikes in the speed graph that are narrow, and not in the same place on successive runs. (Just out of interest - I've followed these discussions, but never actually run it - how long does a scan take to produce one of these graphs? [I presume it doesn't vary _that_ much with capacity, as bigger drives tend to be faster - or?]) I've never run across that HDD SCAN program, I may give it a try, but probably not on that bad drive. I think that drive and also that old data cable earned a place in my garbage.I like to salvage old stuff, but not something that caused me this much trouble... By the way, I mentioned losing the 2011 NEC (Electric code book). I did find that again, it's on archive.org. |
#8
|
|||
|
|||
I formatted that bad drive
On Mon, 11 Dec 2017 01:28:11 -0800, mike wrote:
(Just out of interest - I've followed these discussions, but never actually run it - how long does a scan take to produce one of these graphs? [I presume it doesn't vary _that_ much with capacity, as bigger drives tend to be faster - or?]) I never worried about it. Just turn it on and take a nap. It's only one drive. Doesn't matter how long it takes. That's a long nap.... Running scandisk recently on a different 40gb partition took 12 to 13 hours. In the meantime I could not use the computer for much anything, because it ran so slow. But since I have other computers, there is no reason to run it on the computer I use. This is the only computer that will connect to the internet properly, since I can only get a good connection with Win98. All my other computers run XP. (But I have a few computers in the closet that I dont use at all, and installing Win98 is quite easy). I always wonder why it takes so frikkin long to run those tests??? Of course when Win98 was created, the largest HDDs were something like 6 ot 8 gb, from what I can recall. |
#9
|
|||
|
|||
I formatted that bad drive
|
#10
|
|||
|
|||
I formatted that bad drive
J. P. Gilliver (John) wrote:
In message , writes: On Mon, 11 Dec 2017 01:28:11 -0800, mike wrote: (Just out of interest - I've followed these discussions, but never actually run it - how long does a scan take to produce one of these graphs? [I presume it doesn't vary _that_ much with capacity, as bigger drives tend to be faster - or?]) I never worried about it. Just turn it on and take a nap. It's only one drive. Doesn't matter how long it takes. That's a long nap.... Running scandisk recently on a different 40gb partition took 12 to 13 hours. In the meantime I could not use the [] I always wonder why it takes so frikkin long to run those tests??? [] Well, accessing every sector on a disc - doing some actual tests on it - is going to take a while, and you can't get round it with tricks like compression and so on. "Verifying" a disk, doesn't even rely on checking the data. The disk drive itself, has a Fire polynomial it uses for error correction (a CRC check over each sector individually). And that polynomial can tell the engine in there, that a sector is unreadable. The response is, the disk tries *thousands* of times, to read the sector, until the CRC value (Fire) gives a correct result. If thousands of attempts never give a good value, then up burbles a CRC error, for your utility to consider. If you wade through a bad patch, your read rate is one sector every 15 seconds. Doing a little math will show the process finished at t=infinity. On a healthy disk, say the sustained read rate is 100MB/sec. Well, the CRC check is nothing out of the ordinary, and the disk can still read at 100MB/sec for the purposes of surface verification. Again, a little math will show that using the sustained rate, your little verification application should have finished *long* ago. What other things could a verification application be doing ? You got me there. Maybe they're mining Ethereum with your processor or something. CHKDSK verifies structure. And there isn't much sense mixing file system structure verification, with physical layer verification. So it's probably not something like that. If you have some strange utility like this, it should be tested on a *healthy* disk, to establish a baseline for overhead. Does it work at the sustained read rate on a healthy disk ? Then, when you try it on a duff disk, and the rate works out to 1MB/sec, you know you've got some bad spots it's beating the **** outta. Paul |
Thread Tools | |
Display Modes | |
|
|