lkml.org 
[lkml]   [1997]   [Dec]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectProblems with 2.0.33 kernel and IDE.
Howdy.

Forgive my ignorance if this has already been bought up however;

I used to be running 2.0.29 for quite a while without any hitches. Then I
decide to upgrade to 2.0.33 (oh and these are with the 20Cyrix patches)
because of the bug fixes and what have you. Now I'm getting massive
corruption in the form of:

Dec 22 11:44:03 oracle kernel: EXT2-fs error (device 03:03):
ext2_add_entry: bad entry in directory #118361: rec_len % != 0 - offset=0,
inode=623028, rec_len=5133, name_len=0

And I'm not just talking one or two - literally hundreds at a time. This
particular partition is a 1.5gig used for a usenet newspool. Usually when
use decompresses (batched via uucp) and floods into INN (version 1.5.1) the
errors happen.

After unmounting the partition and running e2fsck (version 1.10) twice (to
be sure, and yes it did correct errors on the second run as well) I remount
and it'll eventually start giving errors again.

Now, this does _not_ happen with 2.0.29 (at least, it doesn't spit it out
the the logs anyway). So my question is about the updated IDE drivers I've
seen in the recent stable kernels. The hard drive in question is a Western
Digial Caviar AC3310 (3gig) and it doesn't seem to corrupt any of the other
partitions from what I can see - which makes this even more strange,
perhaps because of the size of the partition? In the mean time I'm going
to back off to 2.0.29 and reformat.

Thanks in advance for any help.

--
This message is Copyright (C) 1997 by Karl Ferguson
Tower Networking Pty Ltd Tel: +61 8 9456 0000 karl@tower.net.au
t/a STAR Online Services Fax: +61 8 9455 2776 ICQ UIN: 2287428


\
 
 \ /
  Last update: 2005-03-22 13:40    [W:1.857 / U:0.036 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site