On Mon, 28 Jan 2008 20:25:22 -0600 Kevin Dupuy wrote:

On Sat, 2008-01-26 at 06:02 -0500, Aaron Kulkis wrote:

...

Stephen Berman wrote:

...

Starting four days ago, the kernel sends bogus i_mode messages after automatic beagle indexing, like this:

Jan 25 21:15:02 escher su: (to beagleindex) root on none Jan 25 21:15:02 escher su: (to beagleindex) root on none Jan 25 21:18:09 escher syslog-ng[2198]: SIGHUP received, restarting syslog-ng Jan 25 21:18:10 escher syslog-ng[2198]: new configuration initialized Jan 25 21:18:36 escher kernel: klogd 1.4.1, ---------- state change ---------- Jan 25 21:19:07 escher su: (to nobody) root on none Jan 25 21:19:07 escher su: (to nobody) root on none Jan 25 21:19:11 escher su: (to nobody) root on none Jan 25 21:19:20 escher syslog-ng[2198]: last message repeated 4 times Jan 25 21:19:20 escher kernel: init_special_inode: bogus i_mode (33060) Jan 25 21:19:20 escher kernel: init_special_inode: bogus i_mode (30072) Jan 25 21:19:20 escher kernel: init_special_inode: bogus i_mode (31460) Jan 25 21:19:20 escher kernel: init_special_inode: bogus i_mode (30060) Jan 25 21:19:20 escher kernel: init_special_inode: bogus i_mode (30462) Jan 25 21:19:20 escher kernel: init_special_inode: bogus i_mode (31462) Jan 25 21:19:20 escher kernel: init_special_inode: bogus i_mode (57553) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (30057) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (30170) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (72145) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (71145) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (55440) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (35063)

Can anyone tell me what this means, and is it a cause for concern?

What it means is that Beagle has serious problems, something which several people on this list seem to have a very great difficulty in understanding.

I'm afraid this may be independent of whatever problems beagle may have; I suspect a hardware problem, maybe (and if so, hopefully only) bad RAM, though I ran memtest for seven hours (15 full passes) without getting a single error. Still, it might be helpful to find out why beagle is eliciting these messages (and continues to do so).

...

Delete beagle. It's junk.

You'll probably want to file a bug. I haven't seen that before. Thanks.

I will, perhaps it will generate useful feedback. Steve Berman -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

On Tue, 29 Jan 2008 09:45:30 -0500 JP Rosevear wrote:

On Tue, 2008-01-29 at 14:12 +0100, Stephen Berman wrote:

...

On Mon, 28 Jan 2008 20:25:22 -0600 Kevin Dupuy wrote:

...

...

...

...

Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (30057) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (30170) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (72145) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (71145) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (55440) Jan 25 21:19:21 escher kernel: init_special_inode: bogus i_mode (35063)

Can anyone tell me what this means, and is it a cause for concern?

What it means is that Beagle has serious problems, something which several people on this list seem to have a very great difficulty in understanding.

I'm afraid this may be independent of whatever problems beagle may have; I suspect a hardware problem, maybe (and if so, hopefully only) bad RAM, though I ran memtest for seven hours (15 full passes) without getting a single error. Still, it might be helpful to find out why beagle is eliciting these messages (and continues to do so).

These appear to be file system errors. If they are indeed related to beagle its probably because its trying to read something on the filesystem that is corrupt or has dangling symlinks . Its noticeable that the errors show up 4 minutes after the beagle system wide index process starts, not immediately.

Indeed I (did) have some file system corruption under /var, which I first noticed on the same day the bogus i_mode messages first appeared (the corruption appeared on booting in the morning, the i_mode messages after automatic beagle indexing in the evening). Here is a list of all the corrupt files I have found: /var/log: ?--xr--r-t 14904 808936814 962670624 1109408364 Sep 21 2021 acpid ?r-Sr----T 14200 1852990268 1818570799 825372976 Feb 7 1993 boot.msg ?--xr-Sr-x 12132 1668509791 1701327154 1600481121 May 11 2031 cups-autoconfig.log /var/log/YaST2: ?--xrws--- 30821 544104243 825383712 1601463655 Mar 14 1995 config_diff_2007_10_25.log ?---r----- 15451 1987000888 728047648 809120562 Oct 26 2030 disk_sda ?--Srw---- 8306 1918841394 1835361640 1851869706 Mar 19 1997 macro_inst_initial.ycp b---r-S--- 12346 842670177 976433970 101, 110 Feb 26 1996 perl-BL-standalone-log ?--sr-xr-T 25974 174202912 1632268141 811809883 Dec 17 2024 volume_info-2 br-srwS-w- 14899 875720801 959459898 110, 114 Aug 25 1996 y2log-3 ?--xr---wx 12599 540684849 1528854576 959723574 Aug 11 1995 y2log_bootloader ?---r-S--- 8245 825237560 891303968 1663055160 Sep 18 1997 y2logmkinitrd br-xrwS-w- 30768 1701537132 1852976950 58, 56 Feb 1 1993 y2start.log /var/run: br-srwS-wT 28257 1852989293 1818577418 101, 32 Aug 14 1995 acpid.socket ?---rw---- 12346 1936011312 1751348334 808464432 Feb 23 1996 syslog-ng.pid /var/run/dbus: total used in directory 1414116333 available 7247964 ?r--rw--w- 8224 1969303584 1953249900 858929720 Dec 14 2026 pid br--rwxr-T 25957 1751331376 1919251043 74, 10 Oct 17 2021 system_bus_socket /var/lock: ?---rwS--T 30768 1919246946 1797271344 542981680 Mar 18 2029 dmraid /var/lock/subsys: ?---rwx-w- 25908 8243 12337 1751348069 Jan 21 12:54 smbfs ?r--rw---T 14641 791962912 2016425269 1701340019 Jun 12 2018 syslogd /var/lib/samba: ?---r-xrwx 25960 1601451384 1634755443 1242183265 Dec 10 2000 sessionid.tdb /var/adm/fillup-templates: ?--Srw--wt 25959 1634546480 1886413168 1668506912 Feb 19 1987 sysconfig.auditd /var/adm/perl-modules: b--sr---wx 15677 1717989989 941636925 32, 110 Jul 23 2002 perl-XML-Twig /var/cache/preload: br-xrw--wt 8285 825386344 807419442 48, 49 Feb 25 1996 later.preload /var/lib/YaST2: ?--xrwx--- 14899 538980400 1012609594 808464432 Aug 25 1996 backup_boot_sectors br-xrw--wt 26995 2038064488 1886417525 114, 32 Dec 27 2024 bootloader.ycp br-x--x-w- 25966 1701257326 1601466981 52, 97 Jul 1 2025 failed_packages /var/lib/dbus: br--rwxr-T 25208 842674480 976433195 49, 50 Nov 6 2030 machine-id This corruption resulted in several severe problems -- no system logging, no samba. I solved, or at least worked around, these by renaming the directories containing the corrupt files and copying the remaining valid files to new directories with the original names (I could delete some but not all of the corrupt files). After rebooting, some of the files that had been corrupted were now back and ok, others I restored from backups (two files, y2log-3 and perl-XML-Twig, must have been added after my last backup, so I cannot restore them). So far I haven't observed any further corruption or other problems obviously related to the corrupted files. I dealt with the last group of corrupted files last night after beagle indexing, so I have to wait till tonight to see if it has any effect on the bogus i_mode messages.

Possibly you want to try fsck (and maybe back up your data).

I haven't tried fsck since these problems appeared (I have backed up my data). I'm inexperienced with manually running fsck, having only done it a couple of times. Is there anything I need to be aware of before or while running it? My /var is on the same partition as /sbin, so I cannot run fsck on it unmounted; it that a problem? Is there a risk that it could cause more corruption?

I added a comment to your bug as well.

Is it preferable to follow this up on bugzilla instead of this list? Steve Berman -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org