[Bug 1137373] New: Systemd mount and unmount partition during boot
http://bugzilla.suse.com/show_bug.cgi?id=1137373 Bug ID: 1137373 Summary: Systemd mount and unmount partition during boot Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: x86-64 OS: Other Status: NEW Severity: Major Priority: P5 - None Component: Basesystem Assignee: bnc-team-screening@forge.provo.novell.com Reporter: kailed@kailed.net QA Contact: qa-bugs@suse.de Found By: --- Blocker: --- Since snaphot 529 (604 now, skipped few before, sono not sure since when..) one o my partition is mounted and unmounted during boot. # dmesg | grep XFS: [ 5.279229] SGI XFS with ACLs, security attributes, no debug enabled [ 5.280308] XFS (sda2): Mounting V5 Filesystem [ 5.280487] XFS (sdb5): Mounting V5 Filesystem [ 5.327015] XFS (sda2): Ending clean mount [ 5.396838] XFS (sdb5): Ending clean mount [ 6.382957] XFS (sdb5): Unmounting Filesystem # fstab (created at installation time by yast, default subvolumes removed here..): UUID=565e589b-71ff-4d5a-b289-59b29f66c181 / btrfs defaults 0 0 UUID=44dccc07-9527-4ea7-8e15-5dc63486a9cd /local xfs defaults 0 0 UUID=99c78e04-da7c-4291-9944-1e11510d8af8 /home xfs defaults 0 0 UUID=71043c82-a659-4e7b-8a2e-0a5645c424f7 swap swap defaults 0 0 Nothing usefull from journactl: giu 05 18:34:05 fyrfos kernel: XFS (sdb5): Mounting V5 Filesystem ... giu 05 18:34:05 fyrfos kernel: XFS (sdb5): Ending clean mount giu 05 18:34:05 fyrfos systemd[1]: Started udev Kernel Device Manager. giu 05 18:34:05 fyrfos systemd[1]: Mounted /local. giu 05 18:34:05 fyrfos systemd[1]: Reached target Local File Systems. ... giu 05 18:34:06 fyrfos systemd[1]: Stopped target Local File Systems. giu 05 18:34:06 fyrfos systemd[1]: Unmounting /local... ... giu 05 18:34:06 fyrfos kernel: XFS (sdb5): Unmounting Filesystem # mount -a "fix" the problem till next reboot I can also mount the partition with Dolphin. Fs is clean, hdd is ok. I've this setup since more then one year. So.. it's weird. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c1
--- Comment #1 from Daniele Tombolini
http://bugzilla.suse.com/show_bug.cgi?id=1137373
Alynx Zhou
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c2
Franck Bui
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c3
Daniele Tombolini
http://bugzilla.suse.com/show_bug.cgi?id=1137373
Daniele Tombolini
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c4
--- Comment #4 from Franck Bui
From your logs:
[ 27.937471] fyrfos systemd[1]: dev-disk-by\x2duuid-44dccc07\x2d9527\x2d4ea7\x2d8e15\x2d5dc63486a9cd.device: Changed plugged -> dead [ 27.937545] fyrfos systemd[1]: local.mount: Trying to enqueue job local.mount/stop/replace [ 27.937617] fyrfos systemd[1]: Added job local.mount/stop to transaction.
It seems that the device for /local was unplugged during the boot process but it would be great if you could provide the debug logs for udev as well. In order to do so, please resend the journal content but with the following kernel command line options: "debug printk.devkmsg=on". Thanks. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c5
--- Comment #5 from Daniele Tombolini
http://bugzilla.suse.com/show_bug.cgi?id=1137373
Daniele Tombolini
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c6
--- Comment #6 from Franck Bui
http://bugzilla.suse.com/show_bug.cgi?id=1137373
Franck Bui
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c7
--- Comment #7 from Daniele Tombolini
Thanks for the logs.
It seems a known issue that remains unfixed... see: https://github.com/systemd/systemd/pull/12013#issuecomment-479038739
So I'm tempted to close this bug and redirect you to upstream especially since your system is TW which uses the latest version of systemd available.
Why does not happens with /home ? Anyway, is there a suitable workaround ? Maybe calling "mount -a" or reusing generated unit file during boot ? I use this partition for a lot of stuff (daily backup, data..). A small sdd and big hdd is a very common setup nowdays so it's a very big issue, IMHO. Thanks. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c8
--- Comment #8 from Franck Bui
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c9
Franck Bui
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c10
Daniele Tombolini
Daniele, would it be possible to give the following test project a try ?
https://download.opensuse.org/repositories/home:/fbui:/systemd:/next:/ openSUSE-Factory/openSUSE_Tumbleweed/
Thanks.
Done ! Now Systemd and udev are at version 239 and, after a reboot, everything is working as expected ! Let me now if you need more info... Thanks. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c11
--- Comment #11 from Daniele Tombolini
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c12
Franck Bui
Now Systemd and udev are at version 239
That's not the right version as you were supposed to test v242 + some patches... I re-triggered a build of the test project and now the repository contains the correct version of systemd. Can you retry your testing please ? Sorry for the inconvenience. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c13
Daniele Tombolini
(In reply to Daniele Tombolini from comment #10)
Now Systemd and udev are at version 239
That's not the right version as you were supposed to test v242 + some patches...
I re-triggered a build of the test project and now the repository contains the correct version of systemd.
Can you retry your testing please ?
Sorry for the inconvenience.
No problem, version 242-114.2 now and it seems working. "seems" just because I did only one reboot, I don't know if it is enough to say it's OK. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c14
--- Comment #14 from Franck Bui
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c15
--- Comment #15 from Daniele Tombolini
Could you give it a few more tries so we can be confident that your issue is fixed ?
Thanks.
After few more reboot (7-8) I can confirm that for me the bug is fixed. Let me know if it's enough.. Thanks. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1137373
Daniele Tombolini
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c16
Franck Bui
http://bugzilla.suse.com/show_bug.cgi?id=1137373
Fabian Vogt
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c17
--- Comment #17 from Franck Bui
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c31
Franck Bui
http://bugzilla.suse.com/show_bug.cgi?id=1137373
Arvin Schnell
http://bugzilla.suse.com/show_bug.cgi?id=1137373
http://bugzilla.suse.com/show_bug.cgi?id=1137373#c32
--- Comment #32 from Franck Bui
https://bugzilla.suse.com/show_bug.cgi?id=1137373
Arvin Schnell
https://bugzilla.suse.com/show_bug.cgi?id=1137373 https://bugzilla.suse.com/show_bug.cgi?id=1137373#c34 --- Comment #34 from владимир путин <1000Hz.radiowave@gmail.com> --- Can this btrfsmaintenance just be removed by default? It seem to cause more troubles than good. Besides, balancing of btrfs nowadays is not recommended, according to btrfs kernel wiki. And balancing with -susage=... is plain dangerous. Modern btrfs tools wont even allow to do it anymore unless you force it. And scrap won't help if you don't have redundant copy in some sort of raid. So this package basically does a lot of troubles all around the system without any real benefit. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c35
--- Comment #35 from Franck Bui
Can this btrfsmaintenance just be removed by default? It seem to cause more
For btrfsmaintenance complaints can you please use bug #1165780 instead ? This bug deals specifically with the race that might happen when PID1 is reloading. Thanks -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c36
--- Comment #36 from Michael Hirmke
(In reply to владимир путин from comment #34)
Can this btrfsmaintenance just be removed by default? It seem to cause more
For btrfsmaintenance complaints can you please use bug #1165780 instead ?
This bug deals specifically with the race that might happen when PID1 is reloading.
Thanks
In the meantime I found a workaround. Whenever an update triggered a reload of PID 1, run "init 3" before running "init 6". Since then, I never had this problem any longer. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c37
--- Comment #37 from Michael Hirmke
https://bugzilla.suse.com/show_bug.cgi?id=1137373
Hans-Peter Jansen
https://bugzilla.suse.com/show_bug.cgi?id=1137373
Bj�rn Voigt
https://bugzilla.suse.com/show_bug.cgi?id=1137373
Franck Bui
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c41
--- Comment #41 from Franck Bui
Hello guys, I don't know how but for me the problem is fixed since the end of september (more or less). Is it really fixed ? Am I lucky ?
Unfortunately it's not fixed yet. This issue is pretty complex to solve. Please note that it's not specific to SUSE in anyway so any feedback should be reported directly to upstream: https://github.com/systemd/systemd/issues/12953 Now this race was likely to happen on SUSE distros due to the btrfsmaintenance thing that triggered a lot of PID1 reloading during early boot for no good reason. This problem was fixed on TW (still need to be fixed on Leap :-/) and that's the reason why the race is unlikely to happen on this distro since now. While waiting for the workaround, Leap users can follow the procedure described in comment #25. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c42
--- Comment #42 from Michael Hirmke
This problem was fixed on TW (still need to be fixed on Leap :-/) and that's the reason why the race is unlikely to happen on this distro since now.
While waiting for the workaround, Leap users can follow the procedure described in comment #25.
But I still habe the problem on *Tumbleweed* (latest snapshot)! -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c43
--- Comment #43 from Franck Bui
But I still habe the problem on *Tumbleweed* (latest snapshot)!
Then you need to figure out what's triggering PID1 reloads during early boot. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c44
--- Comment #44 from Michael Hirmke
(In reply to Michael Hirmke from comment #42)
But I still habe the problem on *Tumbleweed* (latest snapshot)!
Then you need to figure out what's triggering PID1 reloads during early boot.
So how can I do that? As mentioned, it only happens after having run a zypper dup before reboot. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c45
--- Comment #45 from Franck Bui
So how can I do that? As mentioned, it only happens after having run a zypper dup before reboot.
Then it's probably a different issue. Can you please open a new bug report with a description of the steps you did to trigger the issue ? Also please attach the debug logs (the output of 'journactl -b -oshort-monotonic') when the problem occurs. You can run 'systemctl log-level debug' just before running zypper, no need for the complete debug logs since it doesn't happen during the boot process. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c46
--- Comment #46 from Michael Hirmke
(In reply to Michael Hirmke from comment #44)
So how can I do that? As mentioned, it only happens after having run a zypper dup before reboot.
Then it's probably a different issue.
Can you please open a new bug report with a description of the steps you did to trigger the issue ?
Also please attach the debug logs (the output of 'journactl -b -oshort-monotonic') when the problem occurs. You can run 'systemctl log-level debug' just before running zypper, no need for the complete debug logs since it doesn't happen during the boot process.
There already was a but report last year: https://bugzilla.suse.com/show_bug.cgi?id=1155170 Nothing has changed since then. But of course I can provide the logs, after it has happened again. -- You are receiving this mail because: You are on the CC list for the bug.
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c49
Stefan Hundhammer
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c53
Franck Bui
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c57
--- Comment #57 from Swamp Workflow Management
https://bugzilla.suse.com/show_bug.cgi?id=1137373
https://bugzilla.suse.com/show_bug.cgi?id=1137373#c58
--- Comment #58 from Swamp Workflow Management
participants (2)
-
bugzilla_noreply@novell.com
-
bugzilla_noreply@suse.com