I have Western Digital Caviar Black drive (model WD2002FAEX-007BA0).
System is running Linux Mint 14. Filesystem on drive ext4.
Recently I've got some i/o errors on one of the files. Then I've checked smart and saw Current_Pending_Sector count go up. (few sectors, don't remember exact number)
I've copied all information to another disk and tried to do several badblocks check (in read-only and non destructive mode) using following commands
fsck -vck /dev/sdb1 and fsck -vcck /dev/sdb1
After each scan it reported more and more badblocks. Current_Pending_Sector went as high as 90. Since drive is 2TB, each full scan takes up to day to complete. So I decided to selectively scan only block ranges with badblocks on that partition.
badblocks -b 4096 -sv -p 2 /dev/sdb1 48696066 48674506
after few scans badblock reported more that 1K errors. Then I've decided to wipe whole partition and do destructive write test with badblocks on whole drive.
badblocks -wsv ~/bb-sdb.txt /dev/sdb
but badblocks didn't reported any errors, and magically Current_Pending_Sector went to zero.
Full SMART status available here(pastebin)
Question is why Current_Pending_Sector went to zero? (I understand meaning of pending sector, but question is rather why it went down after full destructive write scan) And can I still trust this drive?