Hello! Two days ago when quitting X11, my PC freezed and the HD made a repetitive noise. So I hard-reset it, and on next boot I got the horrible message like "Unexpected inconsistency, run e2fsck by hand". What I did. Usually it solves the problem, but here e2fsck detected a LOT of errors and after a while said "too many errors, restarting e2fsck from the beginnning"... And the same errors got detected/repaired (or supposedly), and the same message, and... I stopped it with ^C. So. I began to look at e2fsck options, and discovered the "-c" switch, which basically tests physically the blocks on my partition /dev/hdb10 (the one mounted as /var). I have exactly 4 bad blocks on my disk. Then the prog followed with inodes checks and other stuff, and like previously, restarted from the start after a short while, and so on... Then I tried with "e2fsck -l mybadblocksfile", and then it worked... Well, I mean that e2fsck reached the end, and that my computer could boot. Though, I'm pretty sure that the bad blocks are NOT marked as "bad", and if I try to access the files in /var, I will get bunches of errors (because I didn't do much and I don't see why the errors could be corrected). My question is simple : what can I do to mark the bad blocks as bad? Another question : if I try to reinstall Mandrake, hence reformating my faulty partition, will the physical errors be detected and corrected automatically? Or will it crashes the installation and put me in big troubles? Thanks for your help (I've already searched the news but didn't find much). Please *copy* your messages to [EMAIL PROTECTED] as I can't subscribe to the list now (wrong operating system if you see what I mean). Gregus
