On 06/06/2019 12:00 PM, Carlos E. R. wrote:
Hi,
My only 3 TB Seagate disk died suddenly, at 19447 hours of use. Smartctl was perfect one day, thousand of bad sectors the next day. Report attached.
I bought a new disk and tried dd_rhelp to try recover some. It coredumped after a while. So I switched to ddrescue instead, which did finished - but it recovered less than 1%.
Ouch! I feel your pain. I just had a 42.3 box scatter in the most bizarre way I've ever seen. Was rsync'ing files to it. (no doubt a capacitor was getting flaky) The machine froze during rsync, would boot just fine, but then would not allow a user login (boots fine to kdm or to console). It's like it no longer has the links/whatever needed to authenticate. I booted via the install, chrooted, ran zypper up, and there were a number of shared-object files reported as zero-length, e.g.: Additional rpm output: /sbin/ldconfig: File /usr/lib64/libwiretap.so.7 is empty, not checked. /sbin/ldconfig: File /usr/lib64/libwscodecs.so.1.1.0 is empty, not checked. /sbin/ldconfig: File /usr/lib64/libwiretap.so.7.0.14 is empty, not checked. /sbin/ldconfig: File /usr/lib64/libwiretap.so.7 is empty, not checked. /sbin/ldconfig: File /usr/lib64/libwscodecs.so.1.1.0 is empty, not checked. /sbin/ldconfig: File /usr/lib64/libwiretap.so.7.0.14 is empty, not checked. So something is FUBAR. The chroot went fine and running zypper-up went fine, but clearly there were libraries that got hosed somehow. (? bad/corrupt address caused files to be written wiping out libs? -- never thought that was possible running rsync as a non-privileged user) So a board issue can radically affect how/what the disk does. This was on an old laptop (probably 12 years old or so), so board replacement is a nogo. Just sucks to lose a beautiful 17" display (even if the little ticks on the 'F' and 'J' keys are worn off, just improved your hand position :) -- David C. Rankin, J.D.,P.E.