From: "Peter Arremann" <loony@xxxxxxxxxxxx> > I have 2 machines with such messages and hdd's just running fine for 2+ > years. If ext2/3 fs is not marking such sectors as "bad, i dont use it |month before it really goes. Many people are lucky but in the office (where |we've got about 400 linux boxes, many with more than 1 drive) the stats show |that between the first pending sectors and the final drive failure its an |average of about 5 weeks. Enjoy :-) Overhere are running for 2+ years. Both WD's. one is a 40Gb and the other one is a 120Gb. The bad sector started after 1 week after installation. On the 40gb hdd it is sector 65 and on the 120Gb it is sector 1. |Anyway, reallocation of blocks is something the drive has to handle not the |filesystem. Hmm it would be also better (just my mind) if the fs cant read a (bad) sector for many times then it just skip reading from it :) And why it is reading it, because some data wants read data that still is on that bad sector. The question, how to find out what data is reading data from the bad sector and then you can lets say delete a xxx.conf and replace it with a new one. Then the bad sector wil not used anymore :) But in my life time i never saw a reporting tool: "Hey Mr root, samba tries to read smb.conf but its on a not readable sector reported by smartd. delete the smb.conf, and create a new one so i can skip reading that sector forever" hehe Danny. p.s. i go to bed. sleep well or have a nice day.
