All times are UTC - 5 hours [ DST ]




Post new topic Reply to topic  [ 104 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next
Author Message
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 25th, 2019, 16:40 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
This document explains the structure of the 12-byte ATA Passthrough CDB:

Using SAT to access SATA drives:
http://www.scsitoolbox.com/Pdfs/UsingSAT.pdf

Attachment:
ATA_Passthrough_CDB_12-byte.gif
ATA_Passthrough_CDB_12-byte.gif [ 33.64 KiB | Viewed 22250 times ]


I'm still looking for the other bytes ...

_________________
A backup a day keeps DR away.


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 25th, 2019, 16:44 
Offline

Joined: January 29th, 2012, 1:43
Posts: 982
Location: United States
Yes, I know about the ATA passthrough, hddsuperclone uses it. I want to know about the encapsulating USB part, which is part of the USB Mass Storage Class.

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 25th, 2019, 17:19 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
USB Mass Storage Class on an Embedded Device:
http://ww1.microchip.com/downloads/en/appnotes/01169a.pdf

Additional documents in attachments.


Attachments:
usbmassbulk_10.pdf [101.18 KiB]
Downloaded 712 times
usb_msc_boot_1.0.pdf [77.36 KiB]
Downloaded 728 times
Command_Data_Status_Flow.gif
Command_Data_Status_Flow.gif [ 15.18 KiB | Viewed 22238 times ]
CBW.gif
CBW.gif [ 17.57 KiB | Viewed 22238 times ]
CSW.gif
CSW.gif [ 14.06 KiB | Viewed 22238 times ]

_________________
A backup a day keeps DR away.
Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 25th, 2019, 17:53 
Offline

Joined: January 29th, 2012, 1:43
Posts: 982
Location: United States
Sweet! @fzabkar, you are the best at searching and finding documentation. That is what I was looking for, and even if I need to know more, I now have the terminology for search criteria.
:D

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 25th, 2019, 18:14 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
There may be some examples of C code in the smartmontools source. Smartctl uses ATA pass-through to access SMART data behind USB-SATA/IDE bridges. There should also be examples of code which implements VSCs for accessing SMART data behind those bridges which do not support pass-through, eg some Cypress, JMicron and Sunplus bridges.

https://www.smartmontools.org/wiki/Supported_USB-Devices

BTW, I have found Christian Franke to be very helpful and personable.

_________________
A backup a day keeps DR away.


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 25th, 2019, 18:42 
Offline

Joined: January 29th, 2012, 1:43
Posts: 982
Location: United States
Using the built-in passthrough in Linux is one thing, which is what I believe smartmontools does (and so does hddsuperclone). I want to take it to a whole new level to see what can be done with USB. I can take over a USB drive in Linux in a way that will make it disappear from the system (even if it is mounted it will "poof" disappear), and give me full raw USB control (Linux is awesome like that). I kind of want to be able to play a little additional competition game with this "USB Stabilizer". That and I really want to see about the possible soft and hard reset capability, even if it doesn't work on all drives (doesn't look like it worked on the one Larry tested with).

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 26th, 2019, 12:55 
Offline
User avatar

Joined: February 24th, 2019, 11:07
Posts: 44
Location: Canada
You can disable auto mount in windows.
I tested it, working not bad.

https://www.tenforums.com/tutorials/117 ... ndows.html

_________________
https://cslpc.ca/


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 26th, 2019, 15:06 
Offline
User avatar

Joined: September 8th, 2009, 18:21
Posts: 15461
Location: Australia
CSL PC wrote:
You can disable auto mount in windows.
I tested it, working not bad.

https://www.tenforums.com/tutorials/117 ... ndows.html

Thanks. That's good to know.

_________________
A backup a day keeps DR away.


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 26th, 2019, 15:10 
Offline

Joined: January 29th, 2012, 1:43
Posts: 982
Location: United States
CSL PC wrote:
You can disable auto mount in windows.
I tested it, working not bad.

https://www.tenforums.com/tutorials/117 ... ndows.html

So I had to do an experiment to prove a point between Windows and Linux with the mounting. I have a drive that has a weak head with bad sectors spread out all over, including in the MFT. It is a 2.5” sata, but I am using a USB adapter for this test (it is a WD adapter for an older passport drive).

So I connect the drive to Windows 7, and nothing happens. This is even with automount turned on. So far so good. Then I fire up a demo of R-Studio (I only have a paid version for Linux). It locks up until I unplug the drive. I plug the drive back in, try a refresh in R-Studio, and all I get is a spinning wheel. So I disable automount with the diskpart method (and the scrub, which I found was important to get good results), reboot, and try again. This time it did show up with R-Studio quickly with no hang. So if you do disable automount, make sure to also do the scrub, otherwise if the drive had ever been recognized before, it could still automount even if you don’t think it should.

Then I switch to Linux (automount is disabled, and I did not see any reason to enable it for this test). I opened R-Studio, plugged the drive in, hit refresh in R-Studio, and almost instantly the drive and partitions showed up, and I could work with it. To be fair, I tried to click on the drive in the file manager so it would try to mount, and it did hang with a spinning wheel for about a minute, but then it gave a popup message that it could not mount the drive, and no more hang. So even when trying to mount, Linux gave up after about a minute.

So I tried to see what would happen when trying to mount the drive in Windows. I am not sure how long I would need to wait to see if it would ever respond, but after a few minutes I pulled the plug on the drive, no need to punish it any more than needed. It would probably have hung for a long time, or indefinitely (from past experience). Windows just sucks when it comes to working with a failing drive, but it does seem much better as long as it does not try to mount. If it tries to mount, either pack a big lunch for the eternal wait, or unplug/reboot. At least Linux knows when to give up.

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 26th, 2019, 20:38 
Offline
User avatar

Joined: February 24th, 2019, 11:07
Posts: 44
Location: Canada
fzabkar wrote:
CSL PC wrote:
You can disable auto mount in windows.
I tested it, working not bad.

https://www.tenforums.com/tutorials/117 ... ndows.html

Thanks. That's good to know.

Welcome :wink:

_________________
https://cslpc.ca/


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 26th, 2019, 20:47 
Offline
User avatar

Joined: February 24th, 2019, 11:07
Posts: 44
Location: Canada
maximus wrote:
CSL PC wrote:
You can disable auto mount in windows.
I tested it, working not bad.

https://www.tenforums.com/tutorials/117 ... ndows.html

So I had to do an experiment to prove a point between Windows and Linux with the mounting. I have a drive that has a weak head with bad sectors spread out all over, including in the MFT. It is a 2.5” sata, but I am using a USB adapter for this test (it is a WD adapter for an older passport drive).

So I connect the drive to Windows 7, and nothing happens. This is even with automount turned on. So far so good. Then I fire up a demo of R-Studio (I only have a paid version for Linux). It locks up until I unplug the drive. I plug the drive back in, try a refresh in R-Studio, and all I get is a spinning wheel. So I disable automount with the diskpart method (and the scrub, which I found was important to get good results), reboot, and try again. This time it did show up with R-Studio quickly with no hang. So if you do disable automount, make sure to also do the scrub, otherwise if the drive had ever been recognized before, it could still automount even if you don’t think it should.

Then I switch to Linux (automount is disabled, and I did not see any reason to enable it for this test). I opened R-Studio, plugged the drive in, hit refresh in R-Studio, and almost instantly the drive and partitions showed up, and I could work with it. To be fair, I tried to click on the drive in the file manager so it would try to mount, and it did hang with a spinning wheel for about a minute, but then it gave a popup message that it could not mount the drive, and no more hang. So even when trying to mount, Linux gave up after about a minute.

So I tried to see what would happen when trying to mount the drive in Windows. I am not sure how long I would need to wait to see if it would ever respond, but after a few minutes I pulled the plug on the drive, no need to punish it any more than needed. It would probably have hung for a long time, or indefinitely (from past experience). Windows just sucks when it comes to working with a failing drive, but it does seem much better as long as it does not try to mount. If it tries to mount, either pack a big lunch for the eternal wait, or unplug/reboot. At least Linux knows when to give up.


Scott,
This is a good option for someone that can't afford to buy hardware tool, using WDmarvel, Sediv, etc, If automount is on you'll not be able to repair the drive. New WDmarvel work on AHCI, you can connect the drive after the bios and work on it.
I'm not sure if Sediv will work (I think no).

As for Hddsuperclone, I've tested it with many drives, once you have it you no longer need to use data extractor. Great tool.

_________________
https://cslpc.ca/


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 26th, 2019, 20:55 
Offline
User avatar

Joined: February 24th, 2019, 11:07
Posts: 44
Location: Canada
fzabkar wrote:
Thanks for the confirmation and explanation. It would seem that your only real competitor is ddrescue. ;-)

Not really.
This tool eliminate the need of data extractor. Believe me, I've tested and compare it, It can do the same as DE that cost 3K.
I don't think that even Scott know how powerful his tool is.

_________________
https://cslpc.ca/


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 26th, 2019, 21:28 
Offline

Joined: January 29th, 2012, 1:43
Posts: 982
Location: United States
CSL PC wrote:
fzabkar wrote:
Thanks for the confirmation and explanation. It would seem that your only real competitor is ddrescue. ;-)

Not really.
This tool eliminate the need of data extractor. Believe me, I've tested and compare it, It can do the same as DE that cost 3K.
I don't think that even Scott know how powerful his tool is.

I made the free version too good :wink: (Or are you using the pro version? I don't think so...)
Not enough reason to buy the pro for many. That is why I will experiment with raw USB to see if it is something that is worth it.
The one thing I do have going is that since it is the best free tool, it gets enough attention that some do look at and want the pro features :)

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 27th, 2019, 9:36 
Offline
User avatar

Joined: February 24th, 2019, 11:07
Posts: 44
Location: Canada
maximus wrote:
CSL PC wrote:
fzabkar wrote:
Thanks for the confirmation and explanation. It would seem that your only real competitor is ddrescue. ;-)

Not really.
This tool eliminate the need of data extractor. Believe me, I've tested and compare it, It can do the same as DE that cost 3K.
I don't think that even Scott know how powerful his tool is.

I made the free version too good :wink: (Or are you using the pro version? I don't think so...)
Not enough reason to buy the pro for many. That is why I will experiment with raw USB to see if it is something that is worth it.
The one thing I do have going is that since it is the best free tool, it gets enough attention that some do look at and want the pro features :)

Hi Scott,
I'm nissimezra from data medic. You gave me the pro key to test, I was very sick at the time and I couldn't test it.
I'm using the free one and it is doing great job. I even cloned a drive that was not detected by linux but I was surprised to see that it was detected by HSC and even the drive was ready.

It is exceptional tool, not in ddrescue category but more on the paid DE.

_________________
https://cslpc.ca/


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 30th, 2019, 17:18 
Offline

Joined: January 29th, 2012, 1:43
Posts: 982
Location: United States
I have been doing some experimenting with USB, and I can say it is a bit complicated, and taking me some time to figure out even at a basic level. But I do have a result when using my Sabrent USB to SATA/IDE adapter. The SATA drive attached has simulated bad sectors (created by write uncorrectable), and a read of one of those bad sectors takes just over 4 seconds with direct AHCI.

So the test reads a good sector, then a bad sector, then another good sector. The sectors are far enough apart that there is no chance of buffering. A normal test takes just over 4 seconds to process and read both good sectors, which is expected.

When using a USB timeout and a Bulk-Only Mass Storage Reset (along with clearing the endpoint-in halt), I had result of 1.5 seconds. I couldn’t get it much lower without having other issues, but that is a result. The only explanation is that my Sabrent adapter is performing either a soft or hard reset on the SATA drive. Interesting.

I have not yet tried using the ATA passthrough to do any resets, nor have I tested on a true USB only drive. I am having enough trouble making things stable, considering I am still in the learning stages with this, and any how-to documentation is very lacking.

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 30th, 2019, 17:36 
Offline

Joined: January 29th, 2012, 1:43
Posts: 982
Location: United States
A setback that I will have for further testing is that I only have two USB only drives, and neither one has any bad sectors, nor do they support write uncorrectable.

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 30th, 2019, 20:32 
Offline

Joined: January 29th, 2012, 1:43
Posts: 982
Location: United States
I do have an older WD Passport 250GB, which is just a SATA drive with a USB bridge board. The drive does have bad sectors, but the scary thing is that with the UBS bridge, it does not report any errors! In testing as SATA, the drive takes about 4 seconds for a bad sector. So I ran the same USB timeout test on it for reading a good sector, a bad sector, and then another good sector. And it took just over 1 second. So this USB bridge also did a reset, this time I can say it was a soft reset because of the quick response (the USB timeout is 1000ms). So I do have to admit that it is possible to perform some forms of resets to USB drives, although what is possible will depend on the USB bridge.

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: March 31st, 2019, 18:01 
Offline

Joined: January 29th, 2012, 1:43
Posts: 982
Location: United States
So I did some more USB testing, and in short this is what I found. It is possible to send a soft or hard reset via the ATA pass-through command. But whether it works or not is up to the USB bridge. And also, when using a USB timeout, a command can’t be sent until after performing a Bulk-Only Mass Storage Reset, which is the reset that worked to automatically do a soft/hard reset in my previous test. So whether or not being able to send the soft or hard reset pass-through commands is worthwhile is undetermined in my testing. There was no need with the Bulk-Only Mass Storage Reset.

When attempting to perform the soft and hard resets, I used my ear to tell if it actually performed the reset, as there should be some form of head movement, and in my testing with adapters I know the exact sound of a soft and hard reset for a particular drive. One USB adapter (the Sabrent) definitely performed the resets on command. It was the only one. The other adapters were either not consistent, did not seem to do anything, or actually failed to perform the commands. The two WD USB only drives that I have did not seem to do anything (no sound at all from the resets).

So what does this all mean? It shows that being able to soft/hard reset a USB drive is entirely up to the USB bridge. I did find my cheap adapter was not able to take advantage of a timeout, and I also have an older 120GB WD passport that also did not have results with a timeout. With the test of the USB Stabilizer that Larry did, the comparison results from HDDSuperClone indicates to me that the drive that he tested with did not respond to a timeout reset.

So while the DeepSpar USB Stabilizer obviously works, their claim that it is (mostly) because of the resets may not be as true as they indicate. I think they should have pointed out how it also bypasses the OS retries. That would have made me feel better anyway :)

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: April 1st, 2019, 20:47 
Offline

Joined: January 29th, 2012, 1:43
Posts: 982
Location: United States
Here is a crazy scenario where direct USB could be very productive. This potential involves the standard WD slow issue.

First, IF (and a big IF) you could perform soft resets on the drive using a timeout, you could potentially very much increase the response time for the next read. The downside is that any reads skipped by the timeout are likely not actually bad, just timed out because of the slow issue. A setting for a second read attempt could be used for this. Faster, but not great.

Now for the possible big kicker. The OS does not normally perform very large reads at once, at least from my experience in Linux. And in fact, in Linux the SCSI passthrough for a USB drive is limited to 240 sectors of 512 bytes in size. This is likely due to an internal buffer setting. But I just did a simple test with direct USB SCSI command (not a pass-through). Note that there was no data integrity test with this, but I would think the data should be good, as no error was reported. Through my Sabrent USB adapter, connected to a WD drive that has the “slow issue”, I did a test with a cluster size of 128 (64KB) and 32,768 (16MB). The read times were the same 2.5 seconds between reads. It is possible to get up to 32MB per read. Doing the math, it could be possible to read between 100-200 times faster with this method, assuming there are not many actual bad sectors (no bad head). This is only a proof of concept test, but there is definite potential. Can the USB Stabilizer do this with some setting? That could be a big winner if it worked. I just did a basic test using the two USB only drives that I have (which are both WD), and it would seem that it would work on both.

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone


Top
 Profile  
 
 Post subject: Re: New DeepSpar USB Stabilizer
PostPosted: April 2nd, 2019, 8:24 
Offline

Joined: December 8th, 2010, 11:37
Posts: 738
Location: Ottawa, Canada
maximus wrote:
Can the USB Stabilizer do this with some setting? That could be a big winner if it worked. I just did a basic test using the two USB only drives that I have (which are both WD), and it would seem that it would work on both.

The block size was determined by the imaging program used. I don't believe there's a setting in the Stabilizer program to change that. It is displayed on the sector map screen and is constant throughout the imaging task, 128 sectors for all but the DMDE task. I can change it for successive DFL passes when I do multi-pass imaging with the DFL, going from 2048 to 128 on passes 1 and 3. The Stabilizer could be using a different (varying) block size and reporting what size the imaging program is using, I suppose.

_________________
Sabo Computer Repairs & Data Recovery


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 104 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next

All times are UTC - 5 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 22 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group