Dmitry Kostin wrote:
Have you tried --reopen-on-error ? It also reopens the file after slow reads.
Yes, I use -O - but it doesn't help. I think, some pause is still needed.
I'll implement --pause-on-error in the next prerelease then.
I don't understand the meaning of ' max_max_skipbs = 1<< 30 ' - HOW SHOULD I
CHANGE IT either to get no limitation or to limit to about 1Tb?
Setting a max_max_skipbs larger than 2 GiB requires major changes to
ddrescue. By the moment you can try 'max_max_skipbs = 2000000000' (2GB).
Now I want to change the time to start skipping on slow, too (set it to about
1-2 s). Where can it be done?
In the same 'slow_read' function (line 150 of rescuebook.h). You will
need to remove the comparison with the average rate (a_rate), but this
may perhaps cause false positives.
Or, can you please add one more option such as "sensitivity" or something to
set the interval (1-2-5-10-20-30) instead of built-in 30 seconds. (?)
Maybe, add to "-T / --timeout" option two more parameters of sensitivity: time
to start skipping and time to exit on slow.
This sounds too complicated. I'll probably inplement just one option to
set the initial time to disable 'slow_read'.
Maybe, there is a quick way to insert/call the <pause> procedure into/from the
skipping module code? Just to wait for --pause=<seconds> after each skip.
Yes. Insert 'sleep( pause );' before lines 323 and 379 of rescuebook.cc.
Best regards,
Antonio.
_______________________________________________
Bug-ddrescue mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/bug-ddrescue
