Cameron Hutchison said: > This is also a drawback to XFS - I think. I'm guessing that my harddrive > mapped a new block to cover for a bad block (since I was getting bad > blocks), and while XFS did not cause a panic, processes would get stuck > after entering a particular directory on the filesystem. > > The man page for fsck.xfs says: > > NAME > fsck.xfs - do nothing, successfully > > The authors of XFS seem to think that because it is a journalling > filesystem, a filesystem repair tool is not necessary.
yes this is true, I forgot about it. a few years ago I was replacing a drive on a SGI box which was running XFS. The system refused to run fsck. It told me something like 'you don't need to fsck a xfs volume', even though the disk was failing and data was being lost at the time. perhaps this is why reiserfs does the same, it's nice to be able to mark blocks as bad, but whenever I see bad blocks, I get ready for an RMA, because most likely when bad blocks start showing up on the filesystem itself, the drive has already exausted it's reserve blocks, and there is severe damage on the disk. nate -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
