Re: Hardware Imager Challenge

[lcoughey]

This thread is a log of the testing done by Recovery Force from the test drive from HDDOracle mentioned in this thread on their forum here.

Final result summary of unread sectors:

Test (tested by): final number of unread sectors - time to image
HRT DRE (HDD Oracle): 5749
DeepSpar Disk Imager 3 (Recovery Force): 5563 - about 10 hours
PC3000 UDMA DE (Recovery Force): 5595- about 13 hours
PC3000 Express DE (Recovery Force): 5537- about 14 hours
RapidSpar Disk Imager ((Recovery Force): 5585 - about 9 1/2 hours
MRT Ultra DE (Recovery Force): 5567 - about 11 1/4 hours
StarTech SATADUP11 (Recovery Force): 5555 - about 12 1/4 hours
DFL (Sabo Computers):

My thoughts about this test:

• The number of unread sectors are close enough, that I'd consider the differences insignificant
• The patient drive is quite stable, thus not really requiring any of the advance features of hardware imagers
• It is probable that software imagers, such as ddrescue and hddsuperclone would get similar results
• In reality, the only conclusions I can draw from this is that some imagers are significantly faster than others...the DeepSpar products shaved at least a couple hours off the imaging time, which, in theory, would result in less overall wear on the drive failing drive.

[lcoughey]

DDI3 / DDI4 Test

Time to complete the recovery: about 10 hours (the test was initially run with DDI3 with a significantly less aggressive approach, resulting in it taking about a week of many passes. With a re-run with DDI4, we allowed the approach to be more in line with an aggressive long read approach of all the other tests and, ironically enough, got the exact same number of unread sectors as DDI3.)

Unread Sectors: 5563

Spildit-DDI-fullmap - 2.png (3.03 KiB) Viewed 25485 times

Spildit-DRE-File-Report.png (8.04 KiB) Viewed 25485 times

[lcoughey]

PC3000 DE Test

The first time the test was run, it was run with an older PC3000 UDMA card, so just to be fair, we reran the test with the new PC3000 Express card to see if there are any significant differences. The second numbers in () are for the Express card.

Time to complete the recovery: about 13 hours (about 14 hours)

Unread: 5595 (5537)

This was set to run with default timeout settings:

forward timeout (UDMA mode): 5000ms
reverse timeout (PIO mode): 10000ms

PC3000 UDMA DE

Spildit-DE-log.png (15.54 KiB) Viewed 25485 times

PC3000 Express DE

Spildit-PC3K-DE-log.png (5.63 KiB) Viewed 25231 times

PC3000 UDMA DE

Spildit-DE.png (57.69 KiB) Viewed 25485 times

PC3000 Express DE

Spildit-PC3K-DE.png (22.67 KiB) Viewed 25231 times

[lcoughey]

RapidSpar Imager

Time to complete recovery: About 9 1/2 hours

P:SPILDIT-TEST
P:SPILDIT-TEST
Creating Project...
By-Sector Image on Target!
P:SPILDIT-TEST
START Img: 11Apr2017,09:58:56
END Img: 11Apr2017,19:29:02

Unread: 5585

Spildit-RSI.jpg (216.17 KiB) Viewed 25347 times

[lcoughey]

MRT Ultra DE

Time to Complete: 11 1/4 hours

Unread: 5567

Used default imaging settings with 10000ms read timeouts.

Spildit-MRT-DE.png (77.7 KiB) Viewed 25293 times

[lcoughey]

StarTech SATADUP11

Time to Complete: 12 1/4 hours

Unread: 5555

Just changed settings to allow for unlimited read errors.

Spilidit-Startech-SATADUP11.jpg (89.52 KiB) Viewed 25257 times

[LarrySabo]

StarTech SATADUP11
Time to Complete: 12 1/4 hours
Unread: 5555
Just changed settings to allow for unlimited read errors.

Amazing!

[LarrySabo]

My results with the DFL-SRP (DDP) are discussed in Spildit's forum, and were bizarre, to say the least. I ran three tests; the first under Windows 10 was cancelled because it appears that DFL-SRP is not compatible with Win10, even though it seemed to be working normally. (DFL does not claim it's Win10-compatible.) The second was run on a PC running Windows 7 and completed in 11 hrs with zero errors -- not plausible. It was using using a scavenged 250GB drive as a target disk. The third was run on the same PC with output to a file rather than a hard drive, and it completed in just over 11 hrs and had 81 unreadable sectors -- again, not plausible.

Spildit has speculated on the causes for these results in the thread cited above and I returned the patient disk, the clone from the second test, and the image from the third to him for analysis, but the package has still not been received after 4 and a half weeks. Can it get any weirder?

[lcoughey]

Stranger things have happened with Toshiba drives. They usually tend to get worse, not better.