in fact, nothing was broken (except the free space number)... On Sun, Dec 12, 2004 at 11:35:38AM +0100, Michael Nottebrock wrote: > On Sunday, 12. December 2004 10:06, Divacky Roman wrote: > > Hi, > > > > with recent 6-current I had to reboot my machine (I tried mount my cdrom > > but the process hanged in kernel then everything started to crash). > > > > after panic, fsck was invoken: > > witten ~# fsck / > > ** /dev/ad0s1a (NO WRITE) > > ** Last Mounted on / > > ** Root file system > > ** Phase 1 - Check Blocks and Sizes > > fsck_ufs: cannot alloc 3221471004 bytes for inoinfo > > > > and df shows this: > > witten ~# df > > Filesystem Size Used Avail Capacity Mounted on > > /dev/ad0s1a 143G 4.5T -4.4T 3520% / > > > > > > is this a known problem? what can I do with it? > > Kinda, see this thread: > http://lists.freebsd.org/pipermail/freebsd-current/2004-November/043670.html > > From all of my experience, you can consider this filesystem busted. Try to get > as much data off it as you can (chances are some files are so corrupt that > you'll get a hang when you try to copy them) and take care to preserve file > modes & permissions, then newfs and restore. > > It's kinda sad, but right now soft-updates must probably be considered unsafe > for some configurations. The reliability in the case of crashes is about just > as bad as mounting something async... my guess is that it's not really a new > bug in softupdates, but rather effects from modern big ata drives with big > caches and write caching enabled (and no tagged queueing - although IIRC the > ata tags support is disabled in 5.x and -CURRENT anyway). > > -- > ,_, | Michael Nottebrock | lofi_at_freebsd.org > (/^ ^\) | FreeBSD - The Power to Serve | http://www.freebsd.org > \u/ | K Desktop Environment on FreeBSD | http://freebsd.kde.orgReceived on Mon Dec 13 2004 - 14:15:04 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:24 UTC