You are not logged in.
Pages: 1
First to say my system is three years old and is runnig fine so I'm not a newbie
but recently I spotted that during boot my root partition is not fscked:
Lis 13 21:58:56 localhost kernel: sd 2:0:0:0: [sda] Attached SCSI disk
Lis 13 21:58:57 localhost systemd[1]: Found device ST31000528AS.
Lis 13 21:58:57 localhost systemd[1]: Mounting /sysroot...
Lis 13 21:58:57 localhost systemd[1]: Mounted /sysroot.
Lis 13 21:58:57 localhost systemd[1]: Starting Initrd Root File System.
Lis 13 21:58:57 localhost systemd[1]: Reached target Initrd Root File System.
Lis 13 21:58:57 localhost systemd[1]: Starting Reload Configuration from the Real Root...
Lis 13 21:58:57 localhost kernel: EXT4-fs (sda3): warning: maximal mount count reached, running e2fsck is recommended
Lis 13 21:58:57 localhost kernel: EXT4-fs (sda3): mounted filesystem with ordered data mode. Opts: (null)
Lis 13 21:58:57 localhost systemd[1]: Reloading.
This is output from:
# systemctl status systemd-fsck-root
systemd-fsck-root.service - File System Check on Root Device
Loaded: loaded (/usr/lib/systemd/system/systemd-fsck-root.service; static)
Active: inactive (dead)
start condition failed at Ned 2013-10-13 21:58:59 CEST; 22min ago
ConditionPathIsReadWrite=!/ was not met
Docs: man:systemd-fsck-root.service(8)
Says failed. This is my fstab:
#
# /etc/fstab: static file system information
#
# <file system> <dir> <type> <options> <dump> <pass>
shm /dev/shm tmpfs nodev,nosuid 0 0
/dev/sda3 / ext4 noatime,auto_da_alloc,errors=remount-ro 0 1
/dev/sda5 /home ext4 noatime,auto_da_alloc,errors=remount-ro 0 2
/dev/sda6 swap swap defaults 0 0
/dev/sdb1 /media/Store ext4 noatime,auto_da_alloc,errors=remount-ro 0 2
#TMP to RAM
tmpfs /tmp tmpfs noatime,nodev,nosuid,mode=1777,size=8G 0 0
I always read the output during pacman -Syyu and follow instructions so I don't think I've missed something during recent updates...
Also this is my HOOKS section in mkinitcpio.conf:
HOOKS="base systemd autodetect modconf block resume filesystems keymap keyboard fsck shutdown"
I tried moving fsck hook after systemd but it doesnt help. If I put fsck.mode=force to kernel parameter, it fscks all other partition but not root. Does anyone have a clue what could be the cause of it?
More important stuff:
#grep rw /etc/grub.d/10_linux
linux ${rel_dirname}/${basename} root=${linux_root_device_thisversion} rw ${args}
And also the grub.cfg says:
grep rw /boot/grub/grub.cfg
linux /boot/vmlinuz-linux root=UUID=d91af415-069d-49a8-b0fe-b7933ea0726c rw
linux /boot/vmlinuz-linux root=UUID=d91af415-069d-49a8-b0fe-b7933ea0726c rw
Also tried with tune2fs -C 40 /dev/sda3 but nothing happens...
Thanks ![]()
Offline
But when the systemd was upgraded it stated I should add systemd hook to mkinitcpio.conf instead of timestamp hook. And also remove udev which I used. https://wiki.archlinux.org/index.php/mk … mmon_hooks
If other partitions do get checked with fsck, then obviously there is a bug in systemd? Here nothing states about systemd hook breaking fsck on boot https://wiki.archlinux.org/index.php/Fsck and in previous link also nothing states...
Offline
Thanks -- very familiar with the post_upgrade messaging since I wrote it myself. The systemd hook overrides the base hook, which is what allows fsck to work in early userspace (assuming the needed binaries exist). fsck support in early userspace with systemd doesn't work when root is mounted read-write. The alternative is to simply mount root as RO, which then means that the filesystem gets fsck'd twice.
Really, if systemd isn't solving any problems for you in early userspace (and in this case it isn't), there's no point in using it. The hook is still a WIP.
Offline
Tnx for the link and explanation. I don't need systemd and I can live without timestamp. With no systemd in hooks, evertyhing works so for now I'm not going to use it until te upgrade tutorials says I must replace it ![]()
Btw. was last year on LinuxCon, listen to systemd by Leonard, your link reminded me.. ![]()
Offline
Pages: 1