[opensuse] ext4 errors
I was looking at some dmesg output when I spotted the following: # dmesg | egrep 'EXT4|JBD' [ 14.262151] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: (null) [ 14.516952] EXT4-fs (sda1): re-mounted. Opts: acl,user_xattr [ 15.911215] EXT4-fs (sda1): warning: mounting fs with errors, running e2fsck is recommended [ 15.917483] EXT4-fs (sda1): re-mounted. Opts: acl,user_xattr [ 23.292887] EXT4-fs error (device sda1): ext4_mb_generate_buddy:741: group 0, 22398 clusters in bitmap, 22417 in gd [ 23.516248] JBD2: Spotted dirty metadata buffer (dev = sda1, blocknr = 0). There's a risk of filesystem corruption in case of system crash. [ 23.517465] EXT4-fs error (device sda1): ext4_mb_generate_buddy:741: group 2, 9625 clusters in bitmap, 9629 in gd [ 23.517822] JBD2: Spotted dirty metadata buffer (dev = sda1, blocknr = 0). There's a risk of filesystem corruption in case of system crash. [ 314.336017] EXT4-fs (sda1): error count: 9 [ 314.343951] EXT4-fs (sda1): initial error at 1362140595: ext4_lookup:1051: inode 2 [ 314.352004] EXT4-fs (sda1): last error at 1362688890: ext4_mb_generate_buddy:741 The disk is an LVM volume attached over iSCSI - any reason to think ext4 should be having a problem with that? -- Per Jessen, Zürich (5.1°C) http://www.dns24.ch/ - free DNS hosting, made in Switzerland. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Fri, 08 Mar 2013 08:04:31 +0100 Per Jessen wrote:
I was looking at some dmesg output when I spotted the following:
# dmesg | egrep 'EXT4|JBD' [ 14.262151] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: (null) [ 14.516952] EXT4-fs (sda1): re-mounted. Opts: acl,user_xattr [ 15.911215] EXT4-fs (sda1): warning: mounting fs with errors, running e2fsck is recommended [ 15.917483] EXT4-fs (sda1): re-mounted. Opts: acl,user_xattr [ 23.292887] EXT4-fs error (device sda1): ext4_mb_generate_buddy:741: group 0, 22398 clusters in bitmap, 22417 in gd [ 23.516248] JBD2: Spotted dirty metadata buffer (dev = sda1, blocknr = 0). There's a risk of filesystem corruption in case of system crash. [ 23.517465] EXT4-fs error (device sda1): ext4_mb_generate_buddy:741: group 2, 9625 clusters in bitmap, 9629 in gd [ 23.517822] JBD2: Spotted dirty metadata buffer (dev = sda1, blocknr = 0). There's a risk of filesystem corruption in case of system crash. [ 314.336017] EXT4-fs (sda1): error count: 9 [ 314.343951] EXT4-fs (sda1): initial error at 1362140595: ext4_lookup:1051: inode 2 [ 314.352004] EXT4-fs (sda1): last error at 1362688890: ext4_mb_generate_buddy:741
The disk is an LVM volume attached over iSCSI - any reason to think ext4 should be having a problem with that?
Hi Per, In re: "last error at 1362688890" # date -u --date="1970-01-01 1362140595 sec GMT" --> Fri Mar 1 12:23:15 UTC 2013 Anything in your recollection, or logs, at this date and time? This clue courtesy of https://bugzilla.kernel.org/show_bug.cgi?id=51431 regards, Carl -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Carl Hartung wrote:
On Fri, 08 Mar 2013 08:04:31 +0100 Per Jessen wrote:
I was looking at some dmesg output when I spotted the following:
# dmesg | egrep 'EXT4|JBD' [ 14.262151] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: (null) [ 14.516952] EXT4-fs (sda1): re-mounted. Opts: acl,user_xattr [ 15.911215] EXT4-fs (sda1): warning: mounting fs with errors, running e2fsck is recommended [ 15.917483] EXT4-fs (sda1): re-mounted. Opts: acl,user_xattr [ 23.292887] EXT4-fs error (device sda1): ext4_mb_generate_buddy:741: group 0, 22398 clusters in bitmap, 22417 in gd [ 23.516248] JBD2: Spotted dirty metadata buffer (dev = sda1, blocknr = 0). There's a risk of filesystem corruption in case of system crash. [ 23.517465] EXT4-fs error (device sda1): ext4_mb_generate_buddy:741: group 2, 9625 clusters in bitmap, 9629 in gd [ 23.517822] JBD2: Spotted dirty metadata buffer (dev = sda1, blocknr = 0). There's a risk of filesystem corruption in case of system crash. [ 314.336017] EXT4-fs (sda1): error count: 9 [ 314.343951] EXT4-fs (sda1): initial error at 1362140595: ext4_lookup:1051: inode 2 [ 314.352004] EXT4-fs (sda1): last error at 1362688890: ext4_mb_generate_buddy:741
The disk is an LVM volume attached over iSCSI - any reason to think ext4 should be having a problem with that?
Hi Per,
In re: "last error at 1362688890"
# date -u --date="1970-01-01 1362140595 sec GMT"
--> Fri Mar 1 12:23:15 UTC 2013
Anything in your recollection, or logs, at this date and time?
Hmm, that is probably around the time it was installed. How about this one: EXT4-fs (sda1): warning: mounting fs with errors, running e2fsck is recommended Normally the fsck is done automatically, do I need to poke something to get it done? This is an experimental box, everything goes :-)
This clue courtesy of https://bugzilla.kernel.org/show_bug.cgi?id=51431
Thanks for googling for me :-) "ext4 stores errors, and reports them again every 24h, which seems to be a source of confusion for most people who encounter it. :(" -- Per Jessen, Zürich (6.4°C) http://www.dns24.ch/ - free DNS hosting, made in Switzerland. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Fri, 8 Mar 2013 02:30:48 -0500 Carl Hartung wrote: ...
In re: "last error at 1362688890"
# date -u --date="1970-01-01 1362140595 sec GMT"
--> Fri Mar 1 12:23:15 UTC 2013
Correction! 'initial' /= 'last'; I was tired! 'Fri Mar 1 12:23:15 UTC 2013' is the *first* error. date -u --date="1970-01-01 1362688890 sec GMT" The *last* was: 'Thu Mar 7 20:41:30 UTC 2013' So this problem is still ongoing? regards, Carl -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Carl Hartung wrote:
On Fri, 8 Mar 2013 02:30:48 -0500 Carl Hartung wrote: ...
In re: "last error at 1362688890"
# date -u --date="1970-01-01 1362140595 sec GMT"
--> Fri Mar 1 12:23:15 UTC 2013
Correction!
'initial' /= 'last'; I was tired!
'Fri Mar 1 12:23:15 UTC 2013' is the *first* error.
date -u --date="1970-01-01 1362688890 sec GMT"
The *last* was: 'Thu Mar 7 20:41:30 UTC 2013'
So this problem is still ongoing?
Well, I only see those few error messages in the dmesg output, but the 'error count' is increasing: # dmesg | egrep -i 'ext4|jbd' [ 17.479460] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: (null) [ 17.587535] EXT4-fs (sda1): re-mounted. Opts: acl,user_xattr [ 20.687199] EXT4-fs (sda1): warning: mounting fs with errors, running e2fsck is recommended [ 20.746813] EXT4-fs (sda1): re-mounted. Opts: acl,user_xattr [ 26.082002] EXT4-fs error (device sda1): ext4_mb_generate_buddy:741: group 2, 11284 clusters in bitmap, 11288 in gd [ 26.148672] EXT4-fs error (device sda1): ext4_mb_generate_buddy:741: group 0, 22349 clusters in bitmap, 22368 in gd [ 318.432008] EXT4-fs (sda1): error count: 19 [ 318.457037] EXT4-fs (sda1): initial error at 1362140595: ext4_lookup:1051: inode 2 [ 318.502440] EXT4-fs (sda1): last error at 1362739992: ext4_mb_generate_buddy:741 -- Per Jessen, Zürich (12.2°C) http://www.dns24.ch/ - free DNS hosting, made in Switzerland. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Fri, 08 Mar 2013 13:53:30 +0100 Per Jessen wrote:
[ 318.432008] EXT4-fs (sda1): error count: 19
More than doubled since your first post. I hope you don't have any important data on that device! ;-) My gut reaction says the underlying physical hardware needs 'local attention' sooner rather than later. I don't know how verbose or accurate iSCSI is compared to original 50-position / 68-position devices and controllers. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Carl Hartung wrote:
On Fri, 08 Mar 2013 13:53:30 +0100 Per Jessen wrote:
[ 318.432008] EXT4-fs (sda1): error count: 19
More than doubled since your first post. I hope you don't have any important data on that device! ;-) My gut reaction says the underlying physical hardware needs 'local attention' sooner rather than later.
The box is experimental, so no problems there. The hardware is fine, like I said it's an LVM volume attached via iSCSI. The LVM volume was carved out from a 10Tb RAID5 array. -- Per Jessen, Zürich (12.8°C) http://www.dns24.ch/ - free DNS hosting, made in Switzerland. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On Fri, 08 Mar 2013 15:50:21 +0100 Per Jessen wrote:
Carl Hartung wrote:
On Fri, 08 Mar 2013 13:53:30 +0100 Per Jessen wrote:
[ 318.432008] EXT4-fs (sda1): error count: 19
More than doubled since your first post. I hope you don't have any important data on that device! ;-) My gut reaction says the underlying physical hardware needs 'local attention' sooner rather than later.
The box is experimental, so no problems there. The hardware is fine, like I said it's an LVM volume attached via iSCSI. The LVM volume was carved out from a 10Tb RAID5 array.
Then it's a bug? ... comparable to, possibly related to, the one I linked to earlier? What's your plan? :-) -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
Carl Hartung wrote:
On Fri, 08 Mar 2013 15:50:21 +0100 Per Jessen wrote:
Carl Hartung wrote:
On Fri, 08 Mar 2013 13:53:30 +0100 Per Jessen wrote:
[ 318.432008] EXT4-fs (sda1): error count: 19
More than doubled since your first post. I hope you don't have any important data on that device! ;-) My gut reaction says the underlying physical hardware needs 'local attention' sooner rather than later.
The box is experimental, so no problems there. The hardware is fine, like I said it's an LVM volume attached via iSCSI. The LVM volume was carved out from a 10Tb RAID5 array.
Then it's a bug? ... comparable to, possibly related to, the one I linked to earlier?
What's your plan? :-)
Try another installation with ext4, if the same problems arise, switch to JFS. (where I have working systems). -- Per Jessen, Zürich (7.1°C) http://www.dns24.ch/ - free DNS hosting, made in Switzerland. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
participants (2)
-
Carl Hartung
-
Per Jessen