Announcement

**Spork Schivago** · 09-12-2015, 06:16 PM

Re: HDD issues.

Oh, I always want to say I don't think it's anything with the heads, I don't hear any clicking noise or anything like that. Granted, the internals of a hard drive aren't really my strong suite here. Thanks.

**ddscentral** · 09-12-2015, 08:12 PM

Re: HDD issues.

It could be anything: bad heads, bad sectors, dying electronics, etc. But does it really matter ?
Knowing why the drive behaves like that most likely will not help you save the data.

Try to save what you can and replace the drive.

**keeney123** · 09-12-2015, 09:24 PM

Re: HDD issues.

When you say power cycling on and off of the power drive. Are you saying the activation light goes out when you don't see the drive with the software? If this is the case then you could have a bad power connector to the drive. If it only happen in a read mode and not a write mode and it is just in one partition this could be a bad section of the disk. Do you have any test program on your computer that writes and read data to the hard disk. That should tell you the section of the disk you are having problems with. Use to be that one could could read and write to certain places on the disk to be able to tell which ones were bad. The program would not stop when it ran into the first bad section but would store it in memory and continue on. Some programs would remember the bad sections and not try to retrieve them. In this way some of the programs on the disk were recoverable and some programs were corrupt by the bad sectors. You could also retrieve bits of data from all good places on the disk. Then you would have to reverse engineer it to tell which part of the bad program needed to be fixed.

**stj** · 09-12-2015, 10:34 PM

Re: HDD issues.

internal drive, or external?

**Spork Schivago** · 09-13-2015, 11:10 AM

Re: HDD issues.

Originally posted by keeney123 View Post

When you say power cycling on and off of the power drive. Are you saying the activation light goes out when you don't see the drive with the software? If this is the case then you could have a bad power connector to the drive. If it only happen in a read mode and not a write mode and it is just in one partition this could be a bad section of the disk. Do you have any test program on your computer that writes and read data to the hard disk. That should tell you the section of the disk you are having problems with. Use to be that one could could read and write to certain places on the disk to be able to tell which ones were bad. The program would not stop when it ran into the first bad section but would store it in memory and continue on. Some programs would remember the bad sections and not try to retrieve them. In this way some of the programs on the disk were recoverable and some programs were corrupt by the bad sectors. You could also retrieve bits of data from all good places on the disk. Then you would have to reverse engineer it to tell which part of the bad program needed to be fixed.

The hard drive completely looses power and then the power on the hard drive comes back, but this seems to only happen when I read (didn't try writing because the one partition I'm having trouble with was in hibernation mode and my Linux system makes me mount it read only because of this). I left the hard drive connected for over 24 hours, with no reads, and there where no problems. I'm not used to hard drives just turning off like this. It makes using a program like ddrescue a real pain in the ass. Every time the drive shuts down, ddrescue crashes, the drive disappears from /dev for a bit, and then reappears 10 seconds later or so. In X-Windows, it wants me to the type root's password to mount them. I tried setting up a udev rule to turn off USB auto-mount (I'm using a USB external hard drive enclosure) but it didn't work. This is what dmesg says:

Code:

[1998407.447046] usb 2-1: new high-speed USB device number 18 using ehci-pci
[1998407.561993] usb 2-1: New USB device found, idVendor=152d, idProduct=2351
[1998407.561999] usb 2-1: New USB device strings: Mfr=10, Product=11, SerialNumber=5
[1998407.562024] usb 2-1: Product: USB/eSATA To SATA Bridge
[1998407.562029] usb 2-1: Manufacturer: JMicron
[1998407.562031] usb 2-1: SerialNumber: D21E428696FF
[1998407.563805] usb-storage 2-1:1.0: USB Mass Storage device detected
[1998407.567876] scsi21 : usb-storage 2-1:1.0
[1998408.574153] scsi 21:0:0:0: Direct-Access   WDC WD75 00BPVT-80HXZT3    PQ: 0 ANSI: 2 CCS
[1998408.574503] sd 21:0:0:0: Attached scsi generic sg3 type 0
[1998408.805627] sd 21:0:0:0: [sdc] 1465149168 512-byte logical blocks: (750 GB/698 GiB)
[1998408.809765] sd 21:0:0:0: [sdc] Write Protect is off
[1998408.809772] sd 21:0:0:0: [sdc] Mode Sense: 34 00 00 00
[1998408.810756] sd 21:0:0:0: [sdc] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[1998408.905788] sdc: sdc1 sdc2 sdc3 sdc4 sdc5 sdc6
[1998408.912506] sd 21:0:0:0: [sdc] Attached SCSI disk
[1998670.350965] sd 21:0:0:0: [sdc] Unhandled sense code
[1998670.350973] sd 21:0:0:0: [sdc]
[1998670.350977] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[1998670.350981] sd 21:0:0:0: [sdc]
[1998670.350984] Sense Key : Medium Error [current]
[1998670.350991] sd 21:0:0:0: [sdc]
[1998670.350995] Add. Sense: Unrecovered read error
[1998670.350999] sd 21:0:0:0: [sdc] CDB:
[1998670.351028] Read(10): 28 00 00 d0 33 80 00 00 80 00
[1998670.351055] end_request: critical medium error, dev sdc, sector 13644672
[1998700.880045] usb 2-1: reset high-speed USB device number 18 using ehci-pci
[1998701.350447] usb 2-1: device firmware changed
[1998701.350490] usb 2-1: USB disconnect, device number 18
[1998701.355400] scsi 21:0:0:0: rejecting I/O to offline device
[1998701.355408] scsi 21:0:0:0: [sdc] killing request
[1998701.355425] scsi 21:0:0:0: [sdc] Unhandled error code
[1998701.355427] scsi 21:0:0:0: [sdc]
[1998701.355429] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK
[1998701.355432] scsi 21:0:0:0: [sdc] CDB:
[1998701.355434] Read(10): 28 00 00 d0 34 01 00 00 7f 00
[1998701.355442] end_request: I/O error, dev sdc, sector 13644801
[1998701.468035] usb 2-1: new high-speed USB device number 19 using ehci-pci
[1998701.583068] usb 2-1: New USB device found, idVendor=152d, idProduct=2351
[1998701.583073] usb 2-1: New USB device strings: Mfr=10, Product=11, SerialNumber=5
[1998701.583076] usb 2-1: Product: USB/eSATA To SATA Bridge
[1998701.583079] usb 2-1: Manufacturer: JMicron
[1998701.583081] usb 2-1: SerialNumber: D21E428696FF
[1998701.584040] usb-storage 2-1:1.0: USB Mass Storage device detected
[1998701.593172] scsi22 : usb-storage 2-1:1.0
[1998702.600231] scsi 22:0:0:0: Direct-Access   WDC WD75 00BPVT-80HXZT3    PQ: 0 ANSI: 2 CCS
[1998702.604286] sd 22:0:0:0: Attached scsi generic sg3 type 0
[1998702.831828] sd 22:0:0:0: [sdc] 1465149168 512-byte logical blocks: (750 GB/698 GiB)
[1998702.834450] sd 22:0:0:0: [sdc] Write Protect is off
[1998702.834456] sd 22:0:0:0: [sdc] Mode Sense: 34 00 00 00
[1998702.835837] sd 22:0:0:0: [sdc] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[1998702.929902] sdc: sdc1 sdc2 sdc3 sdc4 sdc5 sdc6
[1998702.936577] sd 22:0:0:0: [sdc] Attached SCSI disk

This kind of makes me think it is bad sectors on the drive. So essentially, there's no way to repair this, right? Even with a clean room and the proper tools, the platters would have to be replaced, correct? Even though we could use ddrescue to get data off the partition, something is so horribly wrong, when it try reading a bad sector, the drivers actually have to restart the device.

**goontron** · 09-13-2015, 11:12 AM

Re: HDD issues.

get it on an internal PSU bus. Your case may not be providing enough power by the looks of it. Also if you're using the USB bridge you cant send SMART commands beyond the necessary ones for normal operation.

**Spork Schivago** · 09-13-2015, 11:13 AM

Re: HDD issues.

Originally posted by ddscentral View Post

It could be anything: bad heads, bad sectors, dying electronics, etc. But does it really matter ?
Knowing why the drive behaves like that most likely will not help you save the data.

Try to save what you can and replace the drive.

Yes, it does really matter. If it's something like a dying PCB, I might be able to replace the PCB. Depending on the drive, people have had varying degrees of success. Same with changing the heads. People have made portable clean rooms (and granted, this is recommended) and have successfully replaced parts inside the hard drive long enough to recover the data. I know what risks are involved. And it'd be something, if I could find out what was really causing the problem, I'd do just for the fun of it.

**Spork Schivago** · 09-13-2015, 11:18 AM

Re: HDD issues.

Originally posted by stj View Post

internal drive, or external?

Internal laptop SATA 3.0Gbps WD 750 GB hooked up to an external enclosure right now. I haven't tried hooking it up directly via the onboard SATA ports on my PC yet. I should probably do that before I tear the thing open. We can successfully read data from other partitions and I was able to successfully clone my wife's 500GB laptop hdd, so I know it's not the enclosure. Before that though, I also tried using E-SATA and tried another USB cable just to rule that out.

Any ideas STJ? dmesg kinda tells me why the drives resetting. It appears it finds a bad sector while trying to read and then for some reason, the driver actually restarts the drive. So I guess damage on the platters is looking more and more like the problem.

**goontron** · 09-13-2015, 11:19 AM

Re: HDD issues.

Originally posted by Spork Schivago View Post

Internal laptop SATA 3.0Gbps WD 750 GB hooked up to an external enclosure right now. I haven't tried hooking it up directly via the onboard SATA ports on my PC yet. I should probably do that before I tear the thing open. We can successfully read data from other partitions and I was able to successfully clone my wife's 500GB laptop hdd, so I know it's not the enclosure. Before that though, I also tried using E-SATA and tried another USB cable just to rule that out.

Any ideas STJ? dmesg kinda tells me why the drives resetting. It appears it finds a bad sector while trying to read and then for some reason, the driver actually restarts the drive. So I guess damage on the platters is looking more and more like the problem.

yes, do that. Those USB laptop cases have sparse power. definitely not 1amp worth! Unless you are on USB3.

**Spork Schivago** · 09-13-2015, 11:22 AM

Re: HDD issues.

Originally posted by goontron View Post

get it on an internal PSU bus. Your case may not be providing enough power by the looks of it. Also if you're using the USB bridge you cant send SMART commands beyond the necessary ones for normal operation.

I cannot retrieve the S.M.A.R.T. data using 3rd party tools when it's running through a USB / eSATA to SATA bridge? I did not know this.

I questioned whether it was something with the enclosure or not at first but I figured because I was able to successfully recover all the data from the other partitions, it wasn't bad. The enclosure is not self powered. It has a transformer to power it.

However, just to rule it out once and for all, I'll hook it up directly to my PC later and give it a shot. I've read of others with the same issue and for them, hooking it up to the SATA controller directly didn't help at all. Good idea though. I'll let you know how it goes.

Thanks

**Spork Schivago** · 09-13-2015, 11:25 AM

Re: HDD issues.

Originally posted by goontron View Post

yes, do that. Those USB laptop cases have sparse power. definitely not 1amp worth! Unless you are on USB3.

Output of the transformer is 12V, 1.5A. It's actually made for desktop hard drives as well as laptop, just not the smaller, laptop hard drives. It feels a bit cheap though. I think maybe you posted this before I posted my post about it being not self-powered and having a separate transformer. I will still try it directly plugged into my Desktop.

**goontron** · 09-13-2015, 11:25 AM

Re: HDD issues.

Originally posted by Spork Schivago View Post

I cannot retrieve the S.M.A.R.T. data using 3rd party tools when it's running through a USB / eSATA to SATA bridge? I did not know this.

I questioned this originally but figured because I was able to successfully recover so much data, it wasn't the case. The enclosure is not self powered. It has a transformer to power it.

However, just to rule it out once and for all, I'll hook it up directly to my PC later and give it a shot. I've read of others with the same issue and for them, hooking it up to the SATA controller directly didn't help at all. Good idea though. I'll let you know how it goes.

Thanks

SMART will attempt to recover the sectors, causing a spike in power draw. I see this with desktop drives, so whether it's as big of a power change on a laptop drive i don't know.

Originally posted by Spork Schivago View Post

Output of the transformer is 12V, 1.5A. It's actually made for desktop hard drives as well as laptop, just not the smaller, laptop hard drives. It feels a bit cheap though. I think maybe you posted this before I posted my post about it being not self-powered and having a separate transformer. I will still try it directly plugged into my Desktop.

let me dig up an old thread about that....

https://www.badcaps.net/forum/showthread.php?t=35254

**Spork Schivago** · 09-13-2015, 12:11 PM

Re: HDD issues.

Wow, great thread there Goontron! I never released how cheap the transformer was!

**SteveNielsen** · 09-13-2015, 12:27 PM

Re: HDD issues.

I use a salvaged old 110w ATX PSU to power drives I'm connecting to a USB adapter.

**Spork Schivago** · 09-13-2015, 02:00 PM

Re: HDD issues.

Originally posted by goontron View Post

SMART will attempt to recover the sectors, causing a spike in power draw. I see this with desktop drives, so whether it's as big of a power change on a laptop drive i don't know.

let me dig up an old thread about that....

https://www.badcaps.net/forum/showthread.php?t=35254

I do believe you figured it out Goontron! I just had no idea something like this was happening because of the 1.5A transformer. I knew laptop drives didn't need that much, but wow, boy did I learn something new today.

ddrescue shows:

Code:

[FONT="Courier New"]GNU ddrescue 1.18.1
About to copy 300061 MBytes from /dev/sdc4 to /home/spork/backups/sdc4.img
  Starting positions: infile = 0 B, outfile = 0 B
  Copy block size: 128 sectors    Initial skip size: 128 sectors
Sector size: 512 Bytes

Press Ctrl-C to interrupt
Initial status (read from logfile)
rescued:   5906 MB, errsize:    0 B, errors:    0

Current status
rescued:   6612 MB, errsize:  8650kB, current rate:   524 kB/s
  ipos:   6635 MB,  errors:   143,  average rate:  1613 kB/s
  opos:   6635 MB, run time:  7.88 m, successful read:    0 s ago
Copying non-tried blocks... Pass 1 (forwards)
[/FONT]

Thank you so much!!!! Now I can successfully run ddrescue. Are you familiar with the program? Following the example someone used when they had the same issue, the command is:

Code:

ddrescue -d -v -n -A /dev/sdc4 /home/spork/backups/sdc4.img /home/spork/backups/sdc4.logfile

Supposedly this helped them a bit. I normally use something more like:

Code:

ddrescue -d -r3 /dev/sdc4 /home/spork/backups/sdc4.img /home/spork/backups/sdc4.logfile

And then I just try it over and over again until the drive dies or there's no more errors. The -d means direct disk access (no disk caching, etc), -v means verbose, -n means no-split (skip the splitting phase) and -A means try-again (mark non-split, non-trimmed blocks as non-tried). My -r3 means retry-passed = 3. (exit after <n> retry passes (-1=infinity) [0])

I don't really understand the whole skip the splitting phase and mark non-split, non-trimmed blocks as non-tried. Usually, with a dying drive, first pass, I won't set r at all, just so it doesn't kill the drive faster.

**Spork Schivago** · 09-13-2015, 02:03 PM

Re: HDD issues.

Originally posted by SteveNielsen View Post

I use a salvaged old 110w ATX PSU to power drives I'm connecting to a USB adapter.

I wish I could do that, but this USB / eSATA enclosure won't allow me to. It's more like a caddy that sites on the desk and I push the drives into it, so they're standing vertical. It has a SATA port built into it. It came with our E3 flasher when we bought it for dumping the NAND on the PS3. Replacing the transformer with a higher quality one wouldn't fix this issue, right? I'd need a different USB adapter that allows me to power the drive directly, right?

**fzabkar** · 09-13-2015, 02:31 PM

Re: HDD issues.

Hard drives can report SMART data via USB provided that the USB-SATA bridge firmware supports SCSI-ATA translation (SAT). This allows any ATA command to be tunnelled through the bridge by encapsulating it within a SCSI command packet. Earlier bridges had proprietary methods for doing the same thing, while the earliest bridges had no support for SMART. A tool such as smartctl (smartmontools) should be able to retrieve the SMART data. Otherwise there are Windows tools such as CrystalDiskInfo, HD Sentinel, HDDScan, etc.

WD's drives have problems with oxidisation at the HDA contacts. These contacts can be cleaned with a soft pencil eraser.

Oxidisation on Western Digital PCBs:
http://www.hddoracle.com/viewtopic.p...6&t=649&p=1789

WD's drives also suffer from the "slow" problem. Professional data recovery tools have a single-click solution which is often called the "slow fix" or "dealing with slow responding". AIUI, this turns off reallocation and retries. Essentially it does what ddrescue tries to do, but at the firmware level. You can do the same thing by purchasing a one-month licence for WD Marvel (US$15), or you can apply the free solution described here:

http://forum.hddguru.com/viewtopic.p...29187&start=20

http://www.alexsoft.org/viewtopic.ph...8&p=4345#p4345

There is a free Linux tool, with source code, (by KHONG How Yu) that modifies the same two firmware modules:

http://mod32patch.sourceforge.net/
http://mod2patch.sourceforge.net/

**goontron** · 09-13-2015, 03:10 PM

Re: HDD issues.

Originally posted by fzabkar View Post

Hard drives can report SMART data via USB provided that the USB-SATA bridge firmware supports SCSI-ATA translation (SAT). This allows any ATA command to be tunnelled through the bridge by encapsulating it within a SCSI command packet. Earlier bridges had proprietary methods for doing the same thing, while the earliest bridges had no support for SMART. A tool such as smartctl (smartmontools) should be able to retrieve the SMART data. Otherwise there are Windows tools such as CrystalDiskInfo, HD Sentinel, HDDScan, etc.

WD's drives have problems with oxidisation at the HDA contacts. These contacts can be cleaned with a soft pencil eraser.

Oxidisation on Western Digital PCBs:
http://www.hddoracle.com/viewtopic.p...6&t=649&p=1789

WD's drives also suffer from the "slow" problem. Professional data recovery tools have a single-click solution which is often called the "slow fix" or "dealing with slow responding". AIUI, this turns off reallocation and retries. Essentially it does what ddrescue tries to do, but at the firmware level. You can do the same thing by purchasing a one-month licence for WD Marvel (US$15), or you can apply the free solution described here:

http://forum.hddguru.com/viewtopic.p...29187&start=20

http://www.alexsoft.org/viewtopic.ph...8&p=4345#p4345

There is a free Linux tool, with source code, (by KHONG How Yu) that modifies the same two firmware modules:

http://mod32patch.sourceforge.net/
http://mod2patch.sourceforge.net/

most dont have SAT support. good cases do. but not all.

Originally posted by Spork Schivago View Post

I do believe you figured it out Goontron! I just had no idea something like this was happening because of the 1.5A transformer. I knew laptop drives didn't need that much, but wow, boy did I learn something new today.

ddrescue shows:

Code:

[FONT="Courier New"]GNU ddrescue 1.18.1
About to copy 300061 MBytes from /dev/sdc4 to /home/spork/backups/sdc4.img
  Starting positions: infile = 0 B, outfile = 0 B
  Copy block size: 128 sectors    Initial skip size: 128 sectors
Sector size: 512 Bytes

Press Ctrl-C to interrupt
Initial status (read from logfile)
rescued:   5906 MB, errsize:    0 B, errors:    0

Current status
rescued:   6612 MB, errsize:  8650kB, current rate:   524 kB/s
  ipos:   6635 MB,  errors:   143,  average rate:  1613 kB/s
  opos:   6635 MB, run time:  7.88 m, successful read:    0 s ago
Copying non-tried blocks... Pass 1 (forwards)
[/FONT]

Thank you so much!!!! Now I can successfully run ddrescue. Are you familiar with the program? Following the example someone used when they had the same issue, the command is:

Code:

ddrescue -d -v -n -A /dev/sdc4 /home/spork/backups/sdc4.img /home/spork/backups/sdc4.logfile

Supposedly this helped them a bit. I normally use something more like:

Code:

ddrescue -d -r3 /dev/sdc4 /home/spork/backups/sdc4.img /home/spork/backups/sdc4.logfile

And then I just try it over and over again until the drive dies or there's no more errors. The -d means direct disk access (no disk caching, etc), -v means verbose, -n means no-split (skip the splitting phase) and -A means try-again (mark non-split, non-trimmed blocks as non-tried). My -r3 means retry-passed = 3. (exit after <n> retry passes (-1=infinity) [0])

I don't really understand the whole skip the splitting phase and mark non-split, non-trimmed blocks as non-tried. Usually, with a dying drive, first pass, I won't set r at all, just so it doesn't kill the drive faster.

Assuming that the rating of the brick is fudged is not a bad thing.
I always assume on those bricks that they just tacked the TOTAL rating of the tranny as the rating of both rails. 1.5/2=.75 would be the actual rating per rail. A Samsung laptop drive sitting on my desk has a minimum (they always seem to list the minimal) of .85 amps.

**fzabkar** · 09-13-2015, 03:24 PM

Re: HDD issues.

Originally posted by goontron View Post

most dont have SAT support. High end cases do. but not most.

Not true. SAT support is the norm.

https://www.smartmontools.org/wiki/S...ed_USB-Devices

Announcement

HDD issues.

HDD issues.

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment