Ryan Babchishin said: > Is there any way for me to repair this on the fly? Even better - without > having to scan the whole array (since I know where the error is)?
there sure is! provided you don't mind even more curroption to your filesystem tell fsck to force check. it will cause massive damage I bet :) if you do care about curroption, remount it read-only and do the fix, people can still read the data but won't be able to change anything. first, terminate all processes that are using the filesystem, then mount /filesystem -o remount,ro and run fsck on it. it is unusual to have a filesystem curropt itself while mounted though it's not totally unheard of, a few months ago I had my company's main NFS server running solaris and UFS curropt part of it's filesystem, maybe it's coincidence but the machine itself(ultra10) started beeping as well, spent some time looking at the logs(syslog server) and discovered this, took about 45 minutes to fsck the arrays(30GB & 100GB). I just told the users, tough shit. If I let it continue you may lose yet more data, they said fine and took a break. once the filesystems were repaired the beeping stopped. never did find out if it was related, the ultra10 user manual did not mention anything, and our on-site solaris certified admins didn't know either. nate -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED]?subject=unsubscribe https://listman.redhat.com/mailman/listinfo/redhat-list
