[Bug 809226] New: unable to shutdown the system
https://bugzilla.novell.com/show_bug.cgi?id=809226 https://bugzilla.novell.com/show_bug.cgi?id=809226#c0 Summary: unable to shutdown the system Classification: openSUSE Product: openSUSE 12.3 Version: Final Platform: x86-64 OS/Version: Other Status: NEW Severity: Critical Priority: P5 - None Component: Basesystem AssignedTo: bnc-team-screening@forge.provo.novell.com ReportedBy: chris@infosun.fim.uni-passau.de QAContact: qa-bugs@suse.de Found By: --- Blocker: --- User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:19.0) Gecko/20100101 Firefox/19.0 I made an update installation of openSUSE 12.3 final over openSUSE 12.2. In 12.2 shutdown issues were NOT present. After a shutdown command the system hangs: no output, no log, no HW power off. Using log off or restart in kde or graphical login manager, using reboot, shutdown -h -P now, shutdown -h now, shutdown -r now, reboot, halt, halt -p on a (root) console is all the same. Using acpi=off or fail safe boot mode does not help. My mainboard is an Intel 5 Series DP55WG if this should be of interest. What type of log file should I provide? /var/log/messages, /var/log/boot.log and dmesg shows nothing significant, at least for me. Reproducible: Always Steps to Reproduce: 1. start system 2. shutdown via different possibilities, eg, shutdown -h now from console 3. Actual Results: The systems hangs and does not react. Gettys on ALT-STRG-F2 to F6 are gone (cursor blinks) -F7 is empty, F10 -shows the start log. Changing these consoles is the only thing I can do in this situation, despite of a HW-reset. Expected Results: List of services stop and finally a message System stopped and HW power off. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c1
--- Comment #1 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c
FeiXiang Zhang
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c2
Dr. Werner Fink
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c3
Frederic Crozat
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c4
--- Comment #4 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c5
--- Comment #5 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c6
Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c
Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c7
--- Comment #7 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c8
--- Comment #8 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c9
--- Comment #9 from Frederic Crozat
Btw. I am using kernel-default, and not kernel-desktop. Hope that is not the problem...
clearly not, don't worry :) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c10
--- Comment #10 from Frederic Crozat
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c11
--- Comment #11 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c12
--- Comment #12 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c15
--- Comment #15 from Frederic Crozat
I think I have found the trigger of the problem. Executing 'systemctl stop nfsserver.service' empties the directory /proc ! I never have seen something likewise before. This is only true if /etc/exports is not empty. However, before shutting down the exports work like a charm and can be mounted.
My current content of /etc/exports is
/srv/export *(bind=/,rw,fsid=0,insecure,no_root_squash,sync,no_subtree_check) /srv/export/bak *(bind=/bak,rw,nohide,no_root_squash,sync,no_subtree_check)
Reproduced. It think it is related to bind=/ I'll investigate. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c16
--- Comment #16 from Frederic Crozat
https://bugzilla.novell.com/show_bug.cgi?id=809226 https://bugzilla.novell.com/show_bug.cgi?id=809226#c17 --- Comment #17 from Dr. Werner Fink2013-03-15 13:07:28 UTC --- Why is writting to stdout cause such an error? Shouldn't stdout not be connected to the same device as stderr? Is this realted to bug #728774 ? @@ -103,10 +103,10 @@ sort | while read export dir; do if [ $warned = no ] ; then - echo - echo "NFS: The usage of 'bind=' in /etc/exports is deprecated." - echo "NFS: Is it no longer needed and may cease to work." - echo "NFS: Please remove these settings." + echo >&2 + echo "NFS: The usage of 'bind=' in /etc/exports is deprecated." >&2 + echo "NFS: Is it no longer needed and may cease to work." >&2 + echo "NFS: Please remove these settings." >&2 warned=yes fi test -d "$export" || mkdir -p "$export" -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c18
--- Comment #18 from Frederic Crozat
Why is writting to stdout cause such an error? Shouldn't stdout not be connected to the same device as stderr? Is this realted to bug #728774 ?
It is not, it is writing to /run/nfs/binds.mount instead of stdout / stderr :) I just used stderr because it looked better (and also because stdout is redirected to /run/nfs/bind.mounts in this part of the shell, see line 115 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c19
--- Comment #19 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c20
--- Comment #20 from Neil Brown
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c21
--- Comment #21 from Christian Bachmaier
NFSv3 and NFSv4 exports can coexist on a server. After enabling the support for NFSv4 in the initial configuration dialog, those exports for which fsid=0 and bind=/target/path are not included in the option list are considered v3 exports.
Thus the suggested config at the top only exports via nfs3, ie, 'mount -o vers=4 server:/bak /mnt' or 'mount server:/bak /mnt' fail, ie, they hang for at least 5 minutes! Although both should automatically fall back to nfs3, as man nfs tells:
nfsvers=n The NFS protocol version number used to contact the server's NFS service. If the server does not support the requested version, the mount request fails. If this option is not specified, the client negotiates a suit- able version with the server, trying version 4 first, version 3 second, and version 2 last.
It is very likely that Neil's test was over nfs3. In /etc/sysconfig/nfs I have set USE_KERNEL_NFSD_NUMBER="4" NFS3_SERVER_Support="yes" which should be correct, but maybe that is of interest. For the suggested version: Executing 'mount -o nfsvers=3 server:/ /mnt' logs
2013-03-20T18:50:29.576808+01:00 server rpc.mountd[8598]: Cannot export /, possibly unsupported filesystem or fsid= required in the clients /var/log/messages
A further reason pro bind=... . Under /bak I have a different file system mounted as under /, ie, a partition on another hard disk. Using my configuration makes it possible on the client with a simple 'mount server:/ /mnt' to have full access to /mnt/bak = /bak on the server. Please note that the provided examples are a simplifies version of my server config, eg, I use Kerberos. If you are asking me, bind= is useful and necessary in the current situation. It should not be deprecated. The only option I see is to use bind-mounts within /etc/fstab, which will always be present even if the nfsserver is not running, however. @Neil: What is your advice? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c22
--- Comment #22 from Christian Bachmaier
Thus the suggested config at the top only exports via nfs3, ie, 'mount -o vers=4 server:/bak /mnt' or 'mount server:/bak /mnt' fail, ie, they hang for at least 5 minutes!
Maybe these hangs are gone if executing the mount commands on a different client machine, not on the server itself. !!!!!! However, the exports seem to be nfs3. !!!!!! With my config 'mount server:/ /mnt' is no problem.
For the suggested version: Executing 'mount -o nfsvers=3 server:/ /mnt' logs
2013-03-20T18:50:29.576808+01:00 server rpc.mountd[8598]: Cannot export /, possibly unsupported filesystem or fsid= required in the clients /var/log/messages
These messages are gone if removing the insecure option. This seems also to trigger the entry as a nfs4 export. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c23
--- Comment #23 from Neil Brown
su
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c24
--- Comment #24 from Christian Bachmaier
cat /etc/exports / *(rw,no_root_squash,sync,no_subtree_check,sec=sys:krb5:krb5i:krb5p) /bak *(rw,no_root_squash,sync,no_subtree_check) /srv/tftpboot/systemrescuecd *(ro,no_subtree_check,all_squash,insecure) service nfsserver restart exportfs -av mount localhost:/bak /mnt cat /proc/mounts | grep /mnt server:/bak /mnt nfs rw,relatime,vers=3,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp6,timeo=600,retrans=2,sec=sys,mountaddr=fd7c:878c:ae10:5cd4::2,mountvers=3,mountport=49193,mountproto=udp6,local_lock=none,addr=fd7c:878c:ae10:5cd4::2 0 0 umount /mnt mount -o vers=4 localhost:/bak /mnt
mcedit /etc/exports
cat /etc/exports / *(rw,no_root_squash,sync,no_subtree_check) /bak *(rw,no_root_squash,sync,no_subtree_check) /srv/tftpboot/systemrescuecd *(ro,no_subtree_check,all_squash,insecure) service nfsserver restart exportfs -av mount localhost:/bak /mnt cat /proc/mounts | grep /mnt server:/bak /mnt nfs4 rw,relatime,vers=4.0,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp6,port=0,timeo=600,retrans=2,sec=sys,clientaddr=fd7c:878c:ae10:5cd4::2,local_lock=none,addr=fd7c:878c:ae10:5cd4::2 0 0
The only change was to eliminate the sec= parameter in the first line... I executed no other cmds in the meanwhile and now this is the full content in my /etc/exportfs. Probably this is the reason why I could not reproduce Neil's advice. Without the sec= param, he seems to be right. My orig. bind= config (above) does not suffer from that. A further problem:
cat /etc/exports / *(rw,no_root_squash,sync,no_subtree_check) /bak *(rw,no_root_squash,sync,no_subtree_check) /srv/tftpboot/systemrescuecd *(ro,no_subtree_check,all_squash,insecure) service nfsserver restart exportfs -av mount localhost:/ /mnt cat /proc/mounts | grep /mnt betty:/ /mnt nfs4 rw,relatime,vers=4.0,rsize=1048576,wslsof: WARNING: can't stat() fuse.gvfsd-fuse file system /run/user/1000/gvfs Output information may be incomplete. lsof: WARNING: can't stat() fuse.gvfsd-fuse file system /var/run/user/1000/gvfs Output information may be incomplete. ize=1048576,namlen=255,hard,proto=tcp6,port=0,timeo=600,retrans=2,sec=sys,clientaddr=fd7c:878c:ae10:5cd4::2,local_lock=none,addr=fd7c:878c:ae10:5cd4::2 0 0 cd /mnt ls -la cd / umount /mnt umount.nfs4: /mnt: device is busy umount -f /mnt umount.nfs4: /mnt: device is busy lsof | grep /mnt lsof: WARNING: can't stat() fuse.gvfsd-fuse file system /run/user/1000/gvfs Output information may be incomplete. lsof: WARNING: can't stat() fuse.gvfsd-fuse file system /var/run/user/1000/gvfs Output information may be incomplete.
I found no suitable way to unmount /mnt then. This behavior only shows when mount / to /mnt. Mounting /bak to /mnt and then unmouting it makes no busy problem. Again, my orig. config with bind= (above) does not have this problem. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c25
--- Comment #25 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c26
--- Comment #26 from Dr. Werner Fink
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c27
--- Comment #27 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c
Frederic Crozat
su
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c28
--- Comment #28 from Christian Bachmaier
cat /etc/exports / *(rw,no_root_squash,sync,no_subtree_check) /bak *(rw,no_root_squash,sync,no_subtree_check,sec=sys:krb5:krb5i:krb5p) /srv/tftpboot/systemrescuecd *(ro,no_subtree_check,all_squash,insecure) service nfsserver restart exportfs -av mount localhost:/bak /mnt cat /proc/mounts | grep /mnt server:/bak /mnt nfs4 rw,relatime,vers=4.0,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp6,port=0,timeo=600,retrans=2,sec=sys,clientaddr=xxxx:xxxx:xxxx:xxxx::2,local_lock=none,addr=xxxx:xxxx:xxxx:xxxx::2 0 0
Please, what is the right thing to do, using bind=... further on? Imho comment 24 shows another bug then. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c29
Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c30
Neil Brown
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c31
--- Comment #31 from Neil Brown
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c32
--- Comment #32 from Neil Brown
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c33
--- Comment #33 from Neil Brown
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c34
Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c35
--- Comment #35 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c36
--- Comment #36 from Frederic Crozat
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c37
--- Comment #37 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c38
--- Comment #38 from Frederic Crozat
Frederic, can you at least reproduce it? Seems that Neil gives no answer about reproducability :-/
sorry but I'm only here to fix systemd bugs (and there is none here, in fact). I was able to fix the initial bug which was basicaly making /proc and /sys unmounted. I didn't try to reproduce nor fix other bugs, since they don't seem related to systemd at all (and my TODO list is huge already). Sorry. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c39
Neil Brown
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c40
Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c41
--- Comment #41 from Neil Brown
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c42
--- Comment #42 from Neil Brown
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c43
--- Comment #43 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c44
--- Comment #44 from Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c
Christian Bachmaier
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c45
--- Comment #45 from Bernhard Wiedemann
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c46
Neil Brown
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c47
--- Comment #47 from Bernhard Wiedemann
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c
Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c48
--- Comment #48 from Swamp Workflow Management
https://bugzilla.novell.com/show_bug.cgi?id=809226
https://bugzilla.novell.com/show_bug.cgi?id=809226#c49
--- Comment #49 from Swamp Workflow Management
http://bugzilla.novell.com/show_bug.cgi?id=809226
Swamp Workflow Management
http://bugzilla.novell.com/show_bug.cgi?id=809226
Swamp Workflow Management
participants (1)
-
bugzilla_noreply@novell.com