[ Date Index ][
Thread Index ]
[ <= Previous by date /
thread ]
[ Next by date /
thread => ]
Adrian Midgley wrote:
akm@xxxxxxx:~ > df Filesystem 1k-blocks Used Available Use% Mounted on /dev/hda3 38053708 20037516 16083172 56% / /dev/hda1 23302 5435 16664 25% /boot /dev/hdb1 19550532 10955972 8594560 57% /2 shmfs 563392 0 563392 0% /dev/shm akm@xxxxxxx:~ > uptime 10:22am up 1 min, 1 user, load average: 0.72, 0.21, 0.07 akm@xxxxxxx:~ > POOH! /dev/hda3 umount failed on shutdown. Reiserfs did an extra synch and hoped it would be all right to remount. It evidently was. So, question to gurus: Could I have fixed (extra synch?) the /hda3 without doing a shutdown?
Urm sync is just to flush data from memory to disk. As in the magic incantation "sync ; sync ; sync ; reboot" (from the days before "reboot -n"). Old habits die hard, I still type "sync" if possible, before hitting the power switch on recalcitrant *nix boxes that refuse to shutdown properly. Reiserfsck requires at least a read only filesystem AFAIK so you would have to stop normal operations to run it on the root filesystem. I've seen one suspected case of ReiserFS corruption ever, I have seen a LOT of deleted files held open by manic processes writing data to them. Are you looking for Zebra? (is that the right medical terminology?). Ensure you have latest Reiser patches for your distro. If it happens again I'd try stopping and restarting services first to eliminate the possibility of it being a file held open. Many big Unix server systems use many small logical partitions to localise the effects of run away disk space usage or other problems. It is not uncommon to see "/var" or even "/var/spool/lp" on it's own partition. With good partitioning tools this isn't as painful as it sounds, but it does make for more admin work :( Simon -- The Mailing List for the Devon & Cornwall LUG Mail majordomo@xxxxxxxxxxxx with "unsubscribe list" in the message body to unsubscribe.