Need to mark a few blocks as unusable
#1
Posted 01 October 2012 - 08:14 AM
http://imageshack.us...graphdirect.png
http://imageshack.us...eportdirect.png
http://imageshack.us...hdmapdirect.png
#2
Posted 01 October 2012 - 08:43 AM
doveman, on 01 October 2012 - 08:14 AM, said:
Normally you would do a complete write/erase to have them blocks marked badly (so better make a tested clone of that hard disk).
Otherwise you could try Victoria, finding bad/weak sectors with it and having it forcing the remap.
Obviously YMMV:
http://forum.hddguru...ctor-t6447.html
MHDD and HDAT2 should have a similar option
jaclaz
#3
Posted 01 October 2012 - 08:44 AM
*Edit: same time as Jaclaz, who as usual has a better answer.
GL
This post has been edited by GrofLuigi: 01 October 2012 - 08:46 AM
#4
Posted 01 October 2012 - 09:30 AM
@Jaclaz, It's rather tricky making backups of the Xbox HDD as it uses XFAT, so I have to use special apps to access it, all of which seems to have bugs and keep locking up (e.g. Xplorer360 locks up when trying to delete a folder, which USBXTAFGUI doesn't have any trouble with but did lock up when trying to copy/extract some folders to my PC HDD). I could probably make an IMG of the HDD but I'd rather have the individual files backed up so I can choose what to put back on the HDD. I've probably got most of it backed up already anyway, so it's not that important.
Because it uses XFAT I also have to use special apps to erase/format it and I'm not sure whether those apps would remap the bad sectors, unless that's something that's done at the drive-level irrespective of which app is used. Anyway, better if I can fix it without erasing/formatiing it
This post has been edited by doveman: 01 October 2012 - 09:43 AM
#5
Posted 01 October 2012 - 10:09 AM
#6
Posted 01 October 2012 - 11:03 AM
jaclaz, on 01 October 2012 - 10:09 AM, said:
Yeah, it looks like I might have to, even though that's not really what I want to do. The thing that concerns me is that if I make a clone image and then mark the dodgy blocks as unusable with Victoria, when I go to restore the image won't it want to write to the same dodgy blocks or will it be forced to write somewhere else?
I've never made and restored a Xbox HDD image either, so obviously I'm a bit concerned in case it doesn't work as expected. At least if I had the individual files on my PC HDD I could restore them easily. In fact, it might have been better to copy the files over SMB whilst the HDD was still connected to the Xbox but now I've dismantled the casing to connect it to the PC I can't be bothered to put it all back together, copy the files, then dismantle it again to connect to the PC and fix the dodgy blocks
#7
Posted 01 October 2012 - 12:11 PM
doveman, on 01 October 2012 - 11:03 AM, said:
And AGAIN, I never talked of imaging, I said "tested clone".
This means cloning to another hard disk AND TEST the destination hard disk on the XBOX BEFORE anything else.
If you have NOT another suitable hard disk, start saving money, buy one, clone you current one, test the clone BEFORE doing anything.
doveman, on 01 October 2012 - 11:03 AM, said:
Let's leave alone the Xbox.
You seem like having not a clear idea on how exactly a (modern) hard disk works.
Physical sectors (meaning actual physical space on the surface of a platter) are indexed/addressed in a "misterious" way by the hard disk controller (and it's firmware).
Then there is a re-mapping or if you prefer a "translation" of these addresses in a way that the actual interface/OS can understand.
VERY simplified example, imagine an hard disk with only 10 (ten) setors user addressable, it will have something like 12 real sectors, i.e.:
Physical address (what the controller sees) translated to what the user (interface/OS) sees:
Quote
Pa2->mapped to LBA1
Pa3->mapped to LBA2
Pa4->mapped to LBA3
Pa5 (spare sector) ->NOT mapped
Pa6->mapped to LBA4
Pa7->mapped to LBA5
Pa8->mapped to LBA6
Pa9->mapped to LBA7
Pa10->mapped to LBA8
Pa11->mapped to LBA9
Pa12(spare sector) ->NOT mapped
If (say) Pa3 goes bad, what happens is:
Quote
Pa2->mapped to LBA1
Pa3 (bad sector) ->NOT mapped
Pa4->mapped to LBA3
Pa5 (
Pa6->mapped to LBA4
Pa7->mapped to LBA5
Pa8->mapped to LBA6
Pa9->mapped to LBA7
Pa10->mapped to LBA8
Pa11->mapped to LBA9
Pa12(spare sector) ->NOT mapped
I.e. hard disk bad sectors are "transparent" on the user side.
jaclaz
#8
Posted 01 October 2012 - 12:58 PM
jaclaz, on 01 October 2012 - 12:11 PM, said:
doveman, on 01 October 2012 - 11:03 AM, said:
And AGAIN, I never talked of imaging, I said "tested clone".
This means cloning to another hard disk AND TEST the destination hard disk on the XBOX BEFORE anything else.
If you have NOT another suitable hard disk, start saving money, buy one, clone you current one, test the clone BEFORE doing anything.
Yeah, that isn't going to happen
Quote
That's cool then. Thanks for explaining that.
#9
Posted 01 October 2012 - 01:50 PM
map /iso/Victoria34.IMG (fd0)
map --hook
chainloader (fd0)
it says:
Fat12 BPB found with 0xEB (jmp) leading the boot sector
probed C/H/S = 80/2/18, probed total sectors = 2000
chainloader (fd0)
Error 13: Invalid or unsupported executable format
I can boot the MHDD ISO OK but then that doesn't seem to recognise any of my drives. It presents three interfaces (not the correct ones) and whichever one I select, when I press F2 to scan it says drive not ready.
#10
Posted 01 October 2012 - 09:12 PM
doveman, on 01 October 2012 - 08:14 AM, said:
doveman, on 01 October 2012 - 09:30 AM, said:
doveman, on 01 October 2012 - 11:03 AM, said:
[...]
At least if I had the individual files on my PC HDD I could restore them easily. In fact, it might have been better to copy the files over SMB whilst the HDD was still connected to the Xbox but now I've dismantled the casing to connect it to the PC I can't be bothered to put it all back together, copy the files, then dismantle it again to connect to the PC and fix the dodgy blocks
doveman, on 01 October 2012 - 12:58 PM, said:
It looks like you know what you don't want to do, therefor implying you know what you do want to do. And you keep saying you already have most of it backed up. So I guess at this point you want to be on your own. OK, have fun and let us know if you are ever able to recover any of your data so someone else might eventually benefit from the knowledge. Have a nice day.
Cheers and Regards
#11
Posted 01 October 2012 - 09:20 PM
Yeah, I do know what I want to do. Mark the dodgy blocks as unusable as I said in my first post. I'm not trying to recover any data as I haven't lost any.
#12
Posted 02 October 2012 - 05:19 AM
I would use the newer 3.52 version....
or maybe the Windows version
http://www.majorgeek...dows_d5688.html
or possibly the 4.46 b:
http://www.benchmark...ml?/be_hdd.html
one. (same place holds the 3.52 for DOS)
Small correction
doveman, on 01 October 2012 - 09:20 PM, said:
Quote
jaclaz
This post has been edited by jaclaz: 02 October 2012 - 05:20 AM
#13
Posted 02 October 2012 - 06:58 AM
jaclaz, on 02 October 2012 - 05:19 AM, said:
I would use the newer 3.52 version....
or maybe the Windows version
http://www.majorgeek...dows_d5688.html
or possibly the 4.46 b:
http://www.benchmark...ml?/be_hdd.html
one. (same place holds the 3.52 for DOS)
Ah, thanks. I thought I'd found an official site and 34 was the last version
Quote
Quote
jaclaz
Well I'd be rather unlucky if I did, as I've got an image of the HDD now as well as backups of much of the data on another HDD
By the way, I'm not sure how dependable HDDScan is now, as I have another 2.5" 500GB drive which I use for PC stuff, which I've had in an e-SATA enclosure which it turns out has been causing problems. So I scanned that with HDDScan and it showed there were 6 less than perfect blocks, so I reformatted it to get those blocks marked and re-scanned it and it showed an orange block within seconds. So I stopped it and re-started the scan and repeated this a few times and it didn't show that block as being problematic after the first scan.
#14
Posted 02 October 2012 - 07:29 AM
doveman, on 02 October 2012 - 06:58 AM, said:
The news? being that HDDScan is made AFAICR by the same nice Russian guys that make Victoria.
http://hdd-911.com/i...ocman&Itemid=31
You have to understand that there are several things connected with testing a hard disk at that level.
A "bad block" is normally "bad", once it became "bad", it will remain "bad".
The case with "slow" or "weak" sectors is different, the "slowness" may be a "glitch in the matrix" and never occur again.
As well a "weak" sectors can be (and it happens more often than not) "regenerated" by simply writing to it again a few times different values.
That is what basically tools like HDDregenerator do (whatever HDDregenerator does is seemingly additionally capable to "regenerate" also "bad" sectors, most probably only some "kinds" of "bad" sectors).
So yes, in my experience it is perfectly normal that a few sectors appear as "slow" and in a subsequent scan they do not.
jaclaz
#15
Posted 02 October 2012 - 09:28 AM
jaclaz, on 02 October 2012 - 07:29 AM, said:
A "bad block" is normally "bad", once it became "bad", it will remain "bad".
The case with "slow" or "weak" sectors is different, the "slowness" may be a "glitch in the matrix" and never occur again.
As well a "weak" sectors can be (and it happens more often than not) "regenerated" by simply writing to it again a few times different values.
That is what basically tools like HDDregenerator do (whatever HDDregenerator does is seemingly additionally capable to "regenerate" also "bad" sectors, most probably only some "kinds" of "bad" sectors).
So yes, in my experience it is perfectly normal that a few sectors appear as "slow" and in a subsequent scan they do not
That makes sense as I was doing the Verify test in HDDScan, which I presume reads, writes then reads again from each block, so it could be the write that's fixing it.
I'm not sure that's what's happening though (with all the blocks anyway) as I just took the Xbox 250GB HDD which I did a Verify scan on yesterday and when first starting it, it showed block 0 as dodgy, as it did yesterday, so if writing to it had fixed it I wouldn't have expected it to show as dodgy again today. I then restarted the test a few times and each time it identified block 0 as OK (<5ms a few times but on the last few runs <50ms), so it seems likely that block 0 identifying as <500ms is related to spin-up time (which seems to be borne out by the fact that if I leave it for a while (I mean only several minutes) before starting the test again, it's more likely to show as orange). I'll have to see if I've got an app to check and change the spindown time. As you can see from the third image in my OP, many blocks identify as <20ms so that seems fairly normal, whilst under <50ms is considerably more rare but I presume the fact it's coloured grey rather than orange or red indicates that it's still within acceptable limits.
The first run also identified block 1024 as dodgy, which it also failed to do on subsequent runs until about the fifth and sixth time when it flagged it up again but it was OK again on the seventh, eighth, ninth.
Anyway, I'll run it through Victoria 4.46b, which has loaded albeit with "unknown error or driver not exist" followed by "PORTTALK error. Invalid Driver Handle" followed by "Porttalk driver not installed. Available only API Access" (v43 does the same) but I tested the Get SMART function and that worked so hopefully the other functions will. I'm not sure which settings I need to use though. On the Test page it has four options, Ignore, Remap, Erase or Restore. Clearly I don't want to use Erase but which should I use? It has an option to do a Buttrefly scan as well but I don't like the sound of that
#16
Posted 02 October 2012 - 09:54 AM
doveman, on 02 October 2012 - 09:28 AM, said:
How is the hard disk connected?
Which OS are you running?
The porttalk driver used is:
http://retired.beyon...lk/porttalk.htm
Quote
exclusive access to IO Ports on a Windows NT/2000/XP system.
doveman, on 02 October 2012 - 09:28 AM, said:
Hmmm.
API mode represents AFAIK a much "less powerful" kind of access.
jaclaz
#17
Posted 02 October 2012 - 10:23 AM
jaclaz, on 02 October 2012 - 09:54 AM, said:
Which OS are you running?
The porttalk driver used is:
http://retired.beyon...lk/porttalk.htm
Quote
exclusive access to IO Ports on a Windows NT/2000/XP system.
It's just a normal SATA connection and I'm running Win7, although it's running in Test Mode so driver signing shouldn't be an issue. The Porttalk.sys driver is supplied with Victoria and in the same folder.
I think I've got a working XP partition on this system so I'll try booting to that. I still need to know which option to select on the Test page though.
#18
Posted 02 October 2012 - 10:51 AM
doveman, on 02 October 2012 - 10:23 AM, said:
I guess you are setting some new peaks for the meaning of "assuming".
I would say that if you get a message like:
Quote
it could
Quote
A recently installed program tried to install an unsigned driver. This version of Windows requires all drivers to have a valid digital signature. the driver is unavailable and the program that uses this driver might not work correctly.
Uninstall the program or device that uses this driver and check the publisher''s support website to get a digitally signed driver."
doveman, on 02 October 2012 - 10:23 AM, said:
Better.
doveman, on 02 October 2012 - 10:23 AM, said:
Maybe this helps.
http://docs7.chomiku...-Windows-v4.pdf
jaclaz
#19
Posted 02 October 2012 - 01:57 PM
I think I'll just run it in API mode and see if that does any good. To be honest, I'm somewhat doubting that the problems I've been having with my Xbox can be accounted for by the HDD having 5-6 blocks that are a bit slower than they should be!
#20
Posted 28 February 2013 - 10:57 PM
- ← SMART error, wrong size in BIOS
- Hard Drive and Removable Media issues
- Recovering from "diskpart clean" + possible MFT errors →



Help

Back to top









