Ext4 re-mounting root partition. Is this normal?

holister · 10-29-2010, 12:37 AM

Why does the / partition re-mount? dmesg shows:

Code:

EXT4-fs (sda5): re-mounted. Opts: error=remount-ro,data=writeback,barrier=0

Same partitioning scheme with ext2 does not do this. /sda5 (root) is a "logical" partition. But so are the other 3 which don't re-mount.

1) Booted LiveCD and 1st partitioned the clean drive
2) Turned off journaling on each partition with: tune2fs -O ^has_journal /dev/sdaX
3) Installed Debian "Squeeze" to those partitions.
4) Added following for each partition to /etc/fstab: default,noatime,data=writeback,barrier=0

The systems work but major company info goes on these computers and I need to know what this means and why it does it? Thank you so much for the help.

impert · 10-29-2010, 05:02 PM

I think you'll find this is normal.

Quote:

cam@cwpc:~$ sudo dmesg | grep remount
[ 20.592291] EXT4-fs (sda18): re-mounted. Opts: errors=remount-ro
[ 33.305885] EXT4-fs (sda18): re-mounted. Opts: errors=remount-ro,commit=0

The partition is mounted read-only during boot for safety, and then remounted later so you can use it. Or so I believe.

holister · 10-30-2010, 01:52 AM

I really need to find out for sure. Is this a fact with ext4? Here's why I ask...

After doing a clean install, I made 4 directories in /home. One got files put in it, the rest empty. On next login, the 3 empty directories were gone. I recreated, put files in the 2nd one and 2 left empty. Next login, the 2 empty ones were gone... There is something wrong and I'm trying to put my finger on it.

If it was "re-mounting" read-only, I wouldn't even be able to create the new directories... right?

syg00 · 10-30-2010, 03:14 AM

If you ever suspect "oddities" in a filesystem, fsck it.
No if, no maybe, just do it. Then start looking for answers.

The remount impert mentioned is normal. If you have an error on that f/s, your options some into play ... see that "error=" ?.

holister · 10-30-2010, 03:53 PM

fsck shows all "clean". One final question to put my mind at ease, please... just so I understand...

So, options in /etc/fstab (i.e. Opts: error=remount-ro) refers to errors in the filesystem (in which case, if there were any, it would mount the partition "read-only")... It's not referring to ANY error in the computer in general (i.e. ACPI errors, battery, resource conflicts, etc...)... Correct?

This would help me eliminate one "suspect" and look elsewhere. Thank you.

impert · 10-31-2010, 06:04 PM

Quote:

So, options in /etc/fstab (i.e. Opts: error=remount-ro) refers to errors in the filesystem (in which case, if there were any, it would mount the partition "read-only")

That's how I understand it. Someone else may be able to help you more.

eSelix · 10-31-2010, 07:17 PM

I also have ext4 root partition with option error=remount-ro (which is normal, because you can work with your system even when some errors appear) but without dmesg message. Can you tell which version of kernel you have and check what happens when you boot from liveCD and try to mount sda5?

holister · 10-31-2010, 08:50 PM

Quote:

Originally Posted by eSelix

... Can you tell which version of kernel you have

That's actually what prompted my questions in the first place. In my original post I reported the process I used to install. I cut up the drive using (cfdisk) LiveCD, formatted all for ext4 and turned off journaling. Then installed Debian Squeeze to those partitions. That gave me kernel 2.6.32-5-686.

After reboot, I added mount options to /etc/fstab. The "error=remount-ro" option was there by default and dmesg did NOT have the EXT4-fs: remounted message.

I then compiled kernel 2.6.36 (tried both system-tailored and stock options), rebooted into new kernel and the "ext4-fs remounted" message was there. (the new kernel fixed other issues though, so no complaints). That's what prompted my question (Is this normal?) The "32" kernel didn't show it and "36" kernel did.

Quote:

Originally Posted by eSelix

... what happens when you boot from liveCD and try to mount sda5?

I booted Fedora. No "remounted" message there. Mounted /dev/sda5 to a temp directory. No "remount" message. But I might be misunderstanding something though. I believe the "remount" happens on boot. So if I just mount the partition after the fact, how would I be able to tell?

JZL240I-U · 11-03-2010, 08:29 AM

I seem to remember that this has something to do with the regular fsck-ing during boot time. One can't run fsck on a rw mounted file system, right?

holister · 11-04-2010, 06:55 PM

I've now installed several distros, all of which have this message. I've also tested with all options in /etc/fstab removed and this is the dmesg report:

Code:

main@submain:~$ dmesg | grep mount
[    1.791510] EXT4-fs (sda5): mounted filesystem without journal. Opts: (null)
[    6.758241] EXT4-fs (sda5): re-mounted. Opts: (null)
[    7.023599] EXT4-fs (sda5): re-mounted. Opts: (null)

So, the "remount" message appears to be part to the startup process. Initially, I thought that if the system had errors, the partition would get re-mounted. Apparently that's not the case. The fact that ext4 on 32 kernel didn't have this message and 36 kernel does, remains beyond my understanding. But thank you all for your help. This is a great community.

I will update the heading as "solved".

JZL240I-U · 11-05-2010, 04:24 AM

32 is quite outdated and ext4 was under heavy construction for some time since then. That might explain it. For the rest see my previous post.

jbilbo · 11-11-2010, 01:31 PM

This happens because there's a script that change the "commit" value of your ext4 partition. Commit value of 0 is the default and minimum (5 seconds, commit=0 equals commit=5), but growing this number improve performance (600 = 10minutes until journal data commits).

This has nothing to do with errors, or fstab... in Ubuntu you can find the script in /usr/lib/pm-utils/power.d/journal-commit from pm-utils package (power management). This script is hooked to all power events (like plug/unplug the power cable).

You can disable this remount-to-change-commit-value behaviour doing "chmod -x /usr/lib/pm-utils/power.d/journal-commit" in Ubuntu.

mfraser · 01-03-2011, 05:24 AM

I'm seeing this too on a desktop:
[ 2.042565] EXT4-fs (sda2): mounted filesystem with ordered data mode. Opts: (null)
[ 15.593523] EXT4-fs (sda2): re-mounted. Opts: errors=remount-ro
[ 17.108361] EXT4-fs (sda3): mounted filesystem with ordered data mode. Opts: (null)
[ 19.958860] EXT4-fs (sdb1): mounted filesystem with ordered data mode. Opts: (null)
[ 20.047201] EXT4-fs (sdb2): mounted filesystem with ordered data mode. Opts: (null)
[ 22.828642] EXT4-fs (sda2): re-mounted. Opts: errors=remount-ro,commit=0
[ 22.838596] EXT4-fs (sda3): re-mounted. Opts: commit=0
[ 22.856821] EXT4-fs (sdb1): re-mounted. Opts: commit=0
[ 22.873126] EXT4-fs (sdb2): re-mounted. Opts: commit=0
[ 1371.140616] EXT4-fs (sda2): re-mounted. Opts: errors=remount-ro,commit=0
[ 1371.576565] EXT4-fs (sda3): re-mounted. Opts: commit=0
[ 1371.597049] EXT4-fs (sdb1): re-mounted. Opts: commit=0
[ 1371.608489] EXT4-fs (sdb2): re-mounted. Opts: commit=0

DanielEKFA · 02-10-2011, 09:29 AM

I see this all the time at the moment. Ctrl+Alt+F'ing from X to a TTY, if it hasn't been switched to recently, will shortly interrupt audio output, then give this message in /var/log/messages (this is me doing Ctrl+Alt+F2, Alt+F8, Ctrl+Alt+F3, Alt+F8):

Feb 10 15:17:11 lnxsrv kernel: [240123.298463] EXT4-fs (sdj2): re-mounted. Opts: errors=remount-ro,discard,commit=0
Feb 10 15:17:12 lnxsrv pulseaudio[11721]: ratelimit.c: 77 events suppressed
Feb 10 15:17:17 lnxsrv kernel: [240129.531111] EXT4-fs (sdj2): re-mounted. Opts: errors=remount-ro,discard,commit=0
Feb 10 15:17:22 lnxsrv pulseaudio[11721]: ratelimit.c: 83 events suppressed
Feb 10 15:17:27 lnxsrv kernel: [240139.290213] EXT4-fs (sdj2): re-mounted. Opts: errors=remount-ro,discard,commit=0

/dev/sdj2 is my root partition. It's the only ext4 partition on this rig.

The partition isn't re-mounted read-only, though, so it appears to not be because of an error per se. In fact, there appears to be nothing "dangerous" going on, but I really don't like that message!

Kubuntu 10.10 amd64, 2.6.35-24-generic

I'm wondering, since this seems to happen *only* when switching from X to a TTY, if it's related to X. I'm using proprietary nVidia drivers. kdm.log is empty. Hmmm... Actually, I just tailed the syslog, seems there's a lot of things going on here, starting with pulseaudio:

Feb 10 15:26:22 lnxsrv pulseaudio[11721]: ratelimit.c: 86 events suppressed
Feb 10 15:26:26 lnxsrv kernel: [240678.259904] events/1: page allocation failure. order:4, mode:0x40d0
Feb 10 15:26:26 lnxsrv kernel: [240678.259908] Pid: 10, comm: events/1 Tainted: P 2.6.35-24-generic #42-Ubuntu
Feb 10 15:26:26 lnxsrv kernel: [240678.259909] Call Trace:
Feb 10 15:26:26 lnxsrv kernel: [240678.259919] [<ffffffff81108476>] __alloc_pages_slowpath+0x4b6/0x5a0
Feb 10 15:26:26 lnxsrv kernel: [240678.259922] [<ffffffff811086fa>] __alloc_pages_nodemask+0x19a/0x1f0
Feb 10 15:26:26 lnxsrv kernel: [240678.259925] [<ffffffff8113a75a>] alloc_pages_current+0x9a/0x100
Feb 10 15:26:26 lnxsrv kernel: [240678.259927] [<ffffffff811075be>] __get_free_pages+0xe/0x50
Feb 10 15:26:26 lnxsrv kernel: [240678.259931] [<ffffffff813604f4>] complete_change_console+0x214/0x240
Feb 10 15:26:26 lnxsrv kernel: [240678.259934] [<ffffffff8136dae0>] ? console_callback+0x0/0x140
Feb 10 15:26:26 lnxsrv acpid: client 11208[0:0] has disconnected
Feb 10 15:26:26 lnxsrv acpid: last message repeated 2 times
Feb 10 15:26:26 lnxsrv acpid: client connected from 11208[0:0]
Feb 10 15:26:26 lnxsrv acpid: 1 client rule loaded
Feb 10 15:26:26 lnxsrv kernel: [240678.259935] [<ffffffff8136057d>] change_console+0x5d/0xb0
Feb 10 15:26:26 lnxsrv kernel: [240678.259937] [<ffffffff8136dbb5>] console_callback+0xd5/0x140
Feb 10 15:26:26 lnxsrv kernel: [240678.259942] [<ffffffff8107a885>] run_workqueue+0xc5/0x1a0
Feb 10 15:26:26 lnxsrv kernel: [240678.259944] [<ffffffff8107aa03>] worker_thread+0xa3/0x110
Feb 10 15:26:26 lnxsrv kernel: [240678.259946] [<ffffffff8107f730>] ? autoremove_wake_function+0x0/0x40
Feb 10 15:26:26 lnxsrv kernel: [240678.259948] [<ffffffff8107a960>] ? worker_thread+0x0/0x110
Feb 10 15:26:26 lnxsrv kernel: [240678.259950] [<ffffffff8107f1d6>] kthread+0x96/0xa0
Feb 10 15:26:26 lnxsrv kernel: [240678.259953] [<ffffffff8100aee4>] kernel_thread_helper+0x4/0x10
Feb 10 15:26:26 lnxsrv kernel: [240678.259955] [<ffffffff8107f140>] ? kthread+0x0/0xa0
Feb 10 15:26:26 lnxsrv kernel: [240678.259957] [<ffffffff8100aee0>] ? kernel_thread_helper+0x0/0x10
Feb 10 15:26:26 lnxsrv kernel: [240678.259958] Mem-Info:
Feb 10 15:26:26 lnxsrv kernel: [240678.259959] Node 0 DMA per-cpu:
Feb 10 15:26:26 lnxsrv kernel: [240678.259961] CPU 0: hi: 0, btch: 1 usd: 0
Feb 10 15:26:26 lnxsrv kernel: [240678.259963] CPU 1: hi: 0, btch: 1 usd: 0
Feb 10 15:26:26 lnxsrv kernel: [240678.259964] Node 0 DMA32 per-cpu:
Feb 10 15:26:26 lnxsrv kernel: [240678.259965] CPU 0: hi: 186, btch: 31 usd: 0
Feb 10 15:26:26 lnxsrv kernel: [240678.259967] CPU 1: hi: 186, btch: 31 usd: 0
Feb 10 15:26:26 lnxsrv kernel: [240678.259968] Node 0 Normal per-cpu:
Feb 10 15:26:26 lnxsrv kernel: [240678.259969] CPU 0: hi: 186, btch: 31 usd: 60
Feb 10 15:26:26 lnxsrv kernel: [240678.259971] CPU 1: hi: 186, btch: 31 usd: 0
Feb 10 15:26:26 lnxsrv kernel: [240678.259974] active_anon:374567 inactive_anon:154338 isolated_anon:0
Feb 10 15:26:26 lnxsrv kernel: [240678.259975] active_file:92450 inactive_file:283690 isolated_file:0
Feb 10 15:26:26 lnxsrv kernel: [240678.259975] unevictable:3 dirty:58967 writeback:3469 unstable:0
Feb 10 15:26:26 lnxsrv kernel: [240678.259976] free:24177 slab_reclaimable:12458 slab_unreclaimable:11321
Feb 10 15:26:26 lnxsrv kernel: [240678.259977] mapped:45177 shmem:10937 pagetables:15476 bounce:0
Feb 10 15:26:26 lnxsrv kernel: [240678.259978] Node 0 DMA free:15484kB min:28kB low:32kB high:40kB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:15656kB mlocked:0kB dirty:0kB writeback:0kB mapped:0kB shmem:0kB slab_reclaimable:0kB slab_unreclaimable:0kB kernel_stack:0kB pagetables:0kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no
Feb 10 15:26:26 lnxsrv kernel: [240678.259984] lowmem_reserve[]: 0 2999 4009 4009
Feb 10 15:26:26 lnxsrv kernel: [240678.259987] Node 0 DMA32 free:77332kB min:6048kB low:7560kB high:9072kB active_anon:1233848kB inactive_anon:340652kB active_file:290740kB inactive_file:991900kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:3071904kB mlocked:0kB dirty:193688kB writeback:11200kB mapped:129436kB shmem:30260kB slab_reclaimable:31352kB slab_unreclaimable:19584kB kernel_stack:1992kB pagetables:36964kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no
Feb 10 15:26:26 lnxsrv kernel: [240678.259993] lowmem_reserve[]: 0 0 1010 1010
Feb 10 15:26:26 lnxsrv kernel: [240678.259995] Node 0 Normal free:3892kB min:2036kB low:2544kB high:3052kB active_anon:264420kB inactive_anon:276700kB active_file:79060kB inactive_file:142860kB unevictable:12kB isolated(anon):0kB isolated(file):0kB present:1034240kB mlocked:0kB dirty:42180kB writeback:2676kB mapped:51272kB shmem:13488kB slab_reclaimable:18480kB slab_unreclaimable:25700kB kernel_stack:3560kB pagetables:24940kB unstable:0kB bounce:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no
Feb 10 15:26:26 lnxsrv kernel: [240678.260002] lowmem_reserve[]: 0 0 0 0
Feb 10 15:26:26 lnxsrv kernel: [240678.260003] Node 0 DMA: 1*4kB 1*8kB 1*16kB 1*32kB 1*64kB 0*128kB 0*256kB 0*512kB 1*1024kB 1*2048kB 3*4096kB = 15484kB
Feb 10 15:26:26 lnxsrv kernel: [240678.260009] Node 0 DMA32: 5553*4kB 6716*8kB 9*16kB 11*32kB 6*64kB 0*128kB 0*256kB 1*512kB 0*1024kB 0*2048kB 0*4096kB = 77332kB
Feb 10 15:26:26 lnxsrv kernel: [240678.260033] Node 0 Normal: 829*4kB 20*8kB 4*16kB 1*32kB 1*64kB 2*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 3892kB
Feb 10 15:26:26 lnxsrv kernel: [240678.260033] 455227 total pagecache pages
Feb 10 15:26:26 lnxsrv kernel: [240678.260033] 68090 pages in swap cache
Feb 10 15:26:26 lnxsrv kernel: [240678.260033] Swap cache stats: add 6184254, delete 6116164, find 3304538/4017869
Feb 10 15:26:26 lnxsrv kernel: [240678.260033] Free swap = 2081936kB
Feb 10 15:26:26 lnxsrv kernel: [240678.260033] Total swap = 3040148kB
Feb 10 15:26:26 lnxsrv kernel: [240678.275426] 1048560 pages RAM
Feb 10 15:26:26 lnxsrv kernel: [240678.275428] 33804 pages reserved
Feb 10 15:26:26 lnxsrv kernel: [240678.275429] 383054 pages shared
Feb 10 15:26:26 lnxsrv kernel: [240678.275430] 747392 pages non-shared
Feb 10 15:26:26 lnxsrv acpid: client connected from 11208[0:0]
Feb 10 15:26:26 lnxsrv acpid: 1 client rule loaded
Feb 10 15:26:26 lnxsrv acpid: client connected from 11208[0:0]
Feb 10 15:26:26 lnxsrv acpid: 1 client rule loaded
Feb 10 15:26:27 lnxsrv pulseaudio[11721]: ratelimit.c: 80 events suppressed
Feb 10 15:26:29 lnxsrv anacron[19975]: Anacron 2.3 started on 2011-02-10
Feb 10 15:26:29 lnxsrv anacron[19975]: Normal exit (0 jobs run)
Feb 10 15:26:29 lnxsrv kernel: [240681.963748] EXT4-fs (sdj2): re-mounted. Opts: errors=remount-ro,discard,commit=0

Does this look like a bug to you, too? OP, assuming that you're not referring to the initial remount from dmesg that happens at boot, could you try to do the same? I.e. in a terminal window,
$ tail -f /var/log/syslog
then switch to a TTY and back?

JZL240I-U · 02-15-2011, 04:36 AM

Here is some related information:

http://searchenterpriselinux.techtar...ystem-barriers