rename() making ext2fs going bonkers? (was: interested in an account)

[Thomas Schwinge]

Hello!

On Fri, Oct 02, 2009 at 10:01:22AM -0600, Eric Blake wrote:
> However, I'm wondering if I just tickled a bug that hung flubber;
> I was running the gnulib test for rename(2), and lost all response from
> the machine.  After killing that make process, even simple commands like
> "ls" failed to do anything for several seconds; when I hit Ctrl-C, I got:
> 
> $ ls
> ^Cls: cannot open directory .: Interrupted system call
> 
> and after I exited my ssh session, I can't seem to get back on.

Yes, such things still happen on Hurd systems :-)...  If this turns out
to be reproducible, we'll obviously have a bigger chance of getting an
understanding of this failure, and fixing it.

Indeed, I also found the machine to be unresponsive, even on the Xen
console, so I restarted the domU, and got this:

    /dev/hd2 was not cleanly unmounted, check forced.
    /dev/hd2: Invalid inode number for '.' in directory inode 804372.
    /dev/hd2: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.

... which also indicates that it is a filesystem-related issue: hd2 is
where /home is on, and where your rename() was acting, I guess.  Usually,
from previous experience, when the system crashes (due to resource
exhaustion in the kernel, for example), /home is fine, but the root
filesystem has some minor inconsistencies (due to open files, etc., as I
understand it).

For reference, here's the complete fsck run; indeed indicating that inode
804372 is related to your gnulib test:

    tschwinge@zenhost:~ $ sudo e2fsck -f /dev/zenhost/flubber-data
    e2fsck 1.41.9 (22-Aug-2009)
    Pass 1: Checking inodes, blocks, and sizes
    Pass 2: Checking directory structure
    Invalid inode number for '.' in directory inode 804372.
    Fix<y>? yes
    Pass 3: Checking directory connectivity
    '..' in /home/ericb/gnulib/testdir1215/build/gltests/test-rename.tdir2 
(804373) is /home/ericb/gnulib/testdir1215/build/gltests/test-rename.tdir 
(804372), should be /home/ericb/gnulib/testdir1215/build/gltests (639644).
    Fix<y>? yes
    Pass 3A: Optimizing directories
    Pass 4: Checking reference counts
    Inode 639644 ref count is 9, should be 8.  Fix<y>? yes
    Inode 639925 ref count is 2, should be 1.  Fix<y>? yes
    Inode 804373 ref count is 3, should be 2.  Fix<y>? yes
    Pass 5: Checking group summary information
    Block bitmap differences:  +(1286463--1286464) +(1617297--1617298) 
+(1617300--1617301)
    Fix<y>? yes
    Free blocks count wrong for group #39 (28393, counted=28391).
    Fix<y>? yes
    Free blocks count wrong for group #49 (28617, counted=28613).
    Fix<y>? yes
    Free blocks count wrong (904952, counted=904946).
    Fix<y>? yes
    Inode bitmap differences:  +(639924--639925) +(804370--804373)
    Fix<y>? yes
    Free inodes count wrong for group #39 (15157, counted=15155).
    Fix<y>? yes
    Free inodes count wrong for group #49 (14830, counted=14826).
    Fix<y>? yes
    Directories count wrong for group #49 (142, counted=146).
    Fix<y>? yes
    Free inodes count wrong (672146, counted=672140).
    Fix<y>? yes
    /dev/zenhost/flubber-data: ***** FILE SYSTEM WAS MODIFIED *****
    /dev/zenhost/flubber-data: 245364/917504 files (5.1% non-contiguous), 
930062/1835008 blocks

flubber is up again.  As that system is primarily used for
non-system-destabilizing things (apart from encountering them for the
first time, obviously), I tend to use a different system for further
exploring such issues: grubber (also listed on
<http://www.gnu.org/software/hurd/public_hurd_boxen.html>).  If you want
to explore this rename() issue further (first try to reproduce it), would
it be fine for you to move this testing to grubber (you already got an
account on there), which is otherwise unused at the moment?  Also, if you
want, I wouldn't hesitate to give you root access, and access to the dom0
so that you can restart hanging / crashed domUs.


Unfortunately, at the moment I don't have time to address the other
topics you and Samuel came up with, but I'll keep the emails around.
Thanks for helping!


Regards,
 Thomas

signature.asc
Description: Digital signature