Re: [Bug-ddrescue] Slowness during recovery

[colo]

Further on my end. Since I do not have a linux box I can not test the -d flag. However I have learned that linux is so small that there are several boot cd's that run it. This I am trying right now, however the last one I tried did not burn correctly and I need to run to the mall *shudder* on christmas eve *double shudder* and test this again.

But it's really cool to hear that you got somewhere. If I may ask what are your speeds per sec or how long did it take to get the 54 gigs with all of the stop and starting? 

On Wed, Dec 24, 2008 at 1:57 AM, James W. Watts <address@hidden> wrote:

Wanted to follow-up and pass along my progress.

I tried some of the tips that Antonio suggested and I'm happy to report they're working. In my latest message to the distribution list, I have another drive that I'm recovering. It's a 160GB laptop SATA drive. It kept getting stuck around the 220MB mark. So I added the -d (direct) switch and reduced the cluster size as Antonio suggested. But it wasn't until I added the --input-position switch that the magic happened. I tried increasing values (225, 250, 300, etc.) for this parameter, having to unplug the drive manually each time to force ddrescue to stop (CTRL-C was non-responsive). When I set it to 500, it the drive finally got going and kept going until it just got stuck around the 54GB mark. So I unplugged the drive, replugged it, and set the input position to 55GB and now it's continuing onward.

sudo ddrescue -v -d -n --cluster-size=10 --input-position=500MB /dev/sdc /media/sea500gb/tony.img /media/sea500gb/tony.log

Keeping my fingers crossed on this one.

Any thoughts as to why ddrescue would not respond to CTRL-C? The only way I could get it to exit was to unplug the drive and make it think the drive didn't exist any more.

Also, regarding the other switches/parameters that can be set with ddrescue...when is it recommended to use the "-t (truncate)" switch? What does it do exactly? Advantages, disadvantages?

Thanks,

James

---

From: colo <address@hidden>

To: ddrescue <address@hidden>
Sent: Tuesday, December 23, 2008 6:14:12 PM

Subject: Re: [Bug-ddrescue] Slowness during recovery

So I went ahead and tried to patch an older 1.6 

from the post below. I am very head strong and stubborn so I did try.

Using 

sudo ./ddrescue -B -d -v -r0 /dev/disk2s3 /Volumes/Lacie/testthree/savepleasethrid.dmg savepleasetwo.log

I know a tiny tiny bit of C so me f'ng up is easy.

Here is the main.cc file that I tried to hand patch.

http://ldbss.com/ddrescue/main.txt

And here is the console output

Initial status (read from logfile)

rescued:         0 B,  errsize:       0 B,  errors:       0

Current status

rescued:         0 B,  errsize:  8126 PiB,  current rate:        0 B/s

   ipos:    8126 PiB,   errors:       1,    average rate:        0 B/s

   opos:    8126 PiB

ddrescue: internal error: bad size copying a Block

This portion    ipos:    8126 PiB

was constantly updating it's self in the terminal, something that never happened before. It usually takes 30 sec to a minute to get back an update.

I made a modification to main.cc so to support ddrescue's direct mode ("-d") when running on Mac OS X or Darwin systems. It seems to work-- in my test case, without -d a volume copy (with around 300 errors) took more than twice as long.

I don't know enough about configure to make this compile in automatically; just use this to

See previous email...