hfsprescue time expectations

[timblaktu]

Hi Elmar, is it unusual for hfsprescue to take several days to run step 1 on a ~250GB hfs+ filesystem on a damaged ssd?

I am trying to recover files on a damaged ssd in an 11,1 imac that won't boot. I put your latest precompiled hfsprescue tar on a bootable osx el capitan install image I had sitting around on an sd card. I have booted into the osx version on the sd card, and have been executing hfsprescue from terminal. It has been running for about a day and has only made about 1% progress. I am tailing the step 1 log file in another window so I can watch which sector, block, offset is currently being processed, and it's been on the same 4k block for several hours now.

I did notice that, at 0.04% progress in step 1, it took hours to process the block/offset it was at. I am assuming that this is normal for certain damaged disks that encounter i/o errors (and associated latencies), but I just wanted to confirm with you that this behavior is indeed expected.

Also I wanted to ask you whether there are other tools that can help me narrow the scope of my recovery process.
* I did see that hfsprescue has the option to recover individual files. Is it required for step 1 to complete before this can be done?
* Is there a way to time-limit each read operation that occurs in step 1?

Thank you for sharing this software and service! Much appreciated!

[Elmar]

Hello,

Hi Elmar, is it unusual for hfsprescue to take several days to run step 1 on a ~250GB hfs+ filesystem on a damaged ssd?

Several days is unusual. I can't remember exactly, but a 1 TB USB 3.0 needs about 6 hours to complete step 1. I think, step 1 should be completed on a 250 GB internal SSD within 10 minutes.

I suggest to create a disk image from the SSD and work with that. Maybe its faster in your case.

* I did see that hfsprescue has the option to recover individual files. Is it required for step 1 to complete before this can be done?

Completing step 1 is required to restore files.

* Is there a way to time-limit each read operation that occurs in step 1?

No.


Please keep me informed.

Best regards
Elmar

[timblaktu]

Thanks. I think im going to wait it out a bit longer, bc i can see the i/o errors racking up in /var/log/system.log. At least i know hfsprescue is not hung.

If after another day it is still hitting i/o errors, i will try to run dd to create an image, zeroing out the bad sections. (I think dd has the option for zero, and also customizeable i/o timeouts). Then will try to run hfsprescue on the image.

Thanks for your help.

[mitlisp]

I just scanned 5 TB. It took me about 24 hours. Thats about the same amount of time it took for commercial software to do a deep scan. A couple of things to note:
It hung a few times. You can tell when it has hung when the amount of bytes scanned stayed static for ~10 minutes.

A few reasons why it hung was mainly due to the system's environment:
-I had multiple programs running in the background.
-Hard drive heat up (due to CPU Load).
-Hard drive went into sleep mode.

The fix was to:
-Reduce the CPU load by dedicating the unit to only run the the scanning software. This also reduced the heat
-Turn off Hard drive sleep mode:
1-I went to: System Preferences>Energy Saver> Unchecked>Prevent Hardware from sleeping automatically when...
2-Turned off sleep settings using terminal commands: PMSET and CAFFEINATE. I downloaded anti-sleep software but that only crashed the system after about the 13th hour. (probably overkill)