On Tue, 25 Apr 2006 19:18:32 +0100 Dave Russell <d.russell@xxxxxxxxxxxxxxxx> opined: > I have a system that was running FC4 up until very recently. > > I first tried to install FC5 in the same way I do all my home > installs, using a boot iso and a local http source (loopback > mounted FC5 DVD iso, sha1sum'd to ensure integrity). > > The install fails each time (on average) at between 5 and 60% > through the rpm install phase. > > The machine locks hard, and it is not possible to switch to another > VT or do anything except power off the machine. > > A DVD install performs in exactly the same way. > > On the handful of times that I am able to get a very minimal > install installed (by removing all the packages from the install > selection), the machine soon locks up when adding more packages or > running yum update. > > The lockup's appear to be ext3 throwing wobblers and re-mounting > filesystems as read only. > > EXT3-fs error (device dm-0): ext3_delete_entry: bad entry in > directory #1376454: > inode out of bounds - offset=0, inode=914432000, rec_len=12, > name_len=1 Aborting journal on device dm-0. > ext3_abort called. > EXT3-fs error (device dm-0): ext3_journal_start_sb: Detected > aborted journal Remounting filesystem read-only > EXT3-fs error (device dm-0) in ext3_unlink: IO failure > > My initial suspicions were a co-incidental imminent hardware > failure... however I have run memtest86 overnight to confirm no > issues with memory, and have also done a read/write destructive > badblocks test on the array which also reported no problems. > > I've raised a bugzilla case, but so far no joy.... can anyone help, > or suggest anything else I should try? > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=189570 > > Here's hoping. > Have you tried fsck from a rescue disk? -- Our DNSRBL - Eliminate Spam: http://www.TQMcube.com Multi-RBL Check: http://www.TQMcube.com/rblcheck.php The Dirty Dozen Spammiest Ranges: http://tqmcube.com/dirty12.php
