On Wednesday 18 February 2015, Michael J Dur wrote:

Hello Moby,

There is a bug open that is the explanation for this behaviour. You can read the details here:

https://bugzilla.opensuse.org/show_bug.cgi?id=918226

For the moment you can go back to the previous version of systemd (and add a lock so it doesn't upgrade until the fix is out) with this command:

zypper -n in --oldpackage systemd-32bit-208-23.3.x86_64 systemd-208-23.3.x86_64 systemd-sysvinit-208-23.3.x86_64 && zypper al systemd-32bit systemd systemd-sysvinit

Regarding this workaround. How could I safely reboot such affected system? Or at least stop services to remount ro etc. $ shutdown -r now Failed to open /dev/initctl: No such device or address Failed to talk to init daemon. $ systemctl stop named Failed to get D-Bus connection: Failed to connect to socket /run/systemd/private: Connection refused Even killing a deamon does not work anymore: $ killall named $ ps aux | grep named named 2027 0.0 0.0 0 0 ? Zs Feb18 0:20 [named] <defunct> cu, Rudi -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

On Thursday 19 February 2015, Michael J Dur wrote:

I'm able to reboot with 'reboot -f', or 'reboot -f & exit'

Hm, this is what I did with about 15 machines yesterday. But this is really bad way.

via ssh.

If login still worked it was luck because it seems that any login produces zombie processes. Some of our machines had already "two many open files" "to many processes" ... couldn't do nothing anymore.

There is ongoing discussion and the current patch offered for testing has not fixed the problem,

I downgraded all our systems. So far it seems only those that have had systemd segfault require a reboot after the downgrade.

On 2015-02-19 05:20, Ruediger Meier wrote:

...

On Wednesday 18 February 2015, Michael J Dur wrote:

...

Hello Moby,

There is a bug open that is the explanation for this behaviour. You can read the details here:

https://bugzilla.opensuse.org/show_bug.cgi?id=918226

For the moment you can go back to the previous version of systemd (and add a lock so it doesn't upgrade until the fix is out) with this command:

zypper -n in --oldpackage systemd-32bit-208-23.3.x86_64 systemd-208-23.3.x86_64 systemd-sysvinit-208-23.3.x86_64 && zypper al systemd-32bit systemd systemd-sysvinit

Regarding this workaround. How could I safely reboot such affected system? Or at least stop services to remount ro etc.

$ shutdown -r now Failed to open /dev/initctl: No such device or address Failed to talk to init daemon.

$ systemctl stop named Failed to get D-Bus connection: Failed to connect to socket /run/systemd/private: Connection refused

Even killing a deamon does not work anymore: $ killall named $ ps aux | grep named named 2027 0.0 0.0 0 0 ? Zs Feb18 0:20 [named] <defunct>

cu, Rudi -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

Il 19/02/2015 15:16, Moby ha scritto:

On 02/19/2015 06:03 AM, Ruediger Meier wrote:

...

On Thursday 19 February 2015, Michael J Dur wrote:

...

I'm able to reboot with 'reboot -f', or 'reboot -f & exit' Hm, this is what I did with about 15 machines yesterday. But this is really bad way.

...

via ssh. If login still worked it was luck because it seems that any login produces zombie processes. Some of our machines had already "two many open files" "to many processes" ... couldn't do nothing anymore.

...

There is ongoing discussion and the current patch offered for testing has not fixed the problem,

I downgraded all our systems. So far it seems only those that have had systemd segfault require a reboot after the downgrade.

On 2015-02-19 05:20, Ruediger Meier wrote:

...

On Wednesday 18 February 2015, Michael J Dur wrote:

...

Hello Moby,

There is a bug open that is the explanation for this behaviour. You can read the details here:

https://bugzilla.opensuse.org/show_bug.cgi?id=918226

For the moment you can go back to the previous version of systemd (and add a lock so it doesn't upgrade until the fix is out) with this command:

zypper -n in --oldpackage systemd-32bit-208-23.3.x86_64 systemd-208-23.3.x86_64 systemd-sysvinit-208-23.3.x86_64 && zypper al systemd-32bit systemd systemd-sysvinit Regarding this workaround. How could I safely reboot such affected system? Or at least stop services to remount ro etc.

$ shutdown -r now Failed to open /dev/initctl: No such device or address Failed to talk to init daemon.

$ systemctl stop named Failed to get D-Bus connection: Failed to connect to socket /run/systemd/private: Connection refused

Even killing a deamon does not work anymore: $ killall named $ ps aux | grep named named 2027 0.0 0.0 0 0 ? Zs Feb18 0:20 [named] <defunct>

cu, Rudi Downgrading systemd, as suggested by Michael, fixed the issue. Now I have systemd downgraded and locked in zypper do it does not get messed with. A very serious issue, IMHO, since pid 1 process is jacked up, and in a stable release as opposed to tumbleweed/factory. I also saw growing numbers of zombies with the broken version, but the machine did not complain about running out of resources.

Regards,

Any news about a patch ? That is a big problem, on about 30 of my remote systems.... :( Testing the backroll on a virtual machine.... -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

On 02/23/2015 11:57 AM, Michael J Dur wrote:

On 2015-02-23 12:49, Claudio ML wrote:

...

Il 19/02/2015 15:16, Moby ha scritto:

...

On 02/19/2015 06:03 AM, Ruediger Meier wrote:

...

On Thursday 19 February 2015, Michael J Dur wrote:

...

I'm able to reboot with 'reboot -f', or 'reboot -f & exit' Hm, this is what I did with about 15 machines yesterday. But this is really bad way.

...

via ssh. If login still worked it was luck because it seems that any login produces zombie processes. Some of our machines had already "two many open files" "to many processes" ... couldn't do nothing anymore.

...

There is ongoing discussion and the current patch offered for testing has not fixed the problem,

I downgraded all our systems. So far it seems only those that have had systemd segfault require a reboot after the downgrade.

On 2015-02-19 05:20, Ruediger Meier wrote:

...

On Wednesday 18 February 2015, Michael J Dur wrote: > Hello Moby, > > There is a bug open that is the explanation for this behaviour. > You can read the details here: > > https://bugzilla.opensuse.org/show_bug.cgi?id=918226 > > For the moment you can go back to the previous version of systemd > (and add a lock so it doesn't upgrade until the fix is out) with > this command: > > zypper -n in --oldpackage systemd-32bit-208-23.3.x86_64 > systemd-208-23.3.x86_64 systemd-sysvinit-208-23.3.x86_64 && zypper > al systemd-32bit systemd systemd-sysvinit Regarding this workaround. How could I safely reboot such affected system? Or at least stop services to remount ro etc.

$ shutdown -r now Failed to open /dev/initctl: No such device or address Failed to talk to init daemon.

$ systemctl stop named Failed to get D-Bus connection: Failed to connect to socket /run/systemd/private: Connection refused

Even killing a deamon does not work anymore: $ killall named $ ps aux | grep named named 2027 0.0 0.0 0 0 ? Zs Feb18 0:20 [named] <defunct>

cu, Rudi Downgrading systemd, as suggested by Michael, fixed the issue. Now I have systemd downgraded and locked in zypper do it does not get messed with. A very serious issue, IMHO, since pid 1 process is jacked up, and in a stable release as opposed to tumbleweed/factory. I also saw growing numbers of zombies with the broken version, but the machine did not complain about running out of resources.

Regards,

Any news about a patch ? That is a big problem, on about 30 of my remote systems.... :( Testing the backroll on a virtual machine....

I believe the patch was put into the repos this morning. I saw 208-32.1 come down to our mirror.

-Michael zypper ref and then zypper up gave me the patch and it works fine.

-- --Moby They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety. -- Benjamin Franklin -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org

Il 24/02/2015 09:44, Claudio ML ha scritto:

Il 23/02/2015 20:31, Moby ha scritto:

...

On 02/23/2015 11:57 AM, Michael J Dur wrote:

...

On 2015-02-23 12:49, Claudio ML wrote:

...

Il 19/02/2015 15:16, Moby ha scritto:

...

On 02/19/2015 06:03 AM, Ruediger Meier wrote:

...

On Thursday 19 February 2015, Michael J Dur wrote: > I'm able to reboot with 'reboot -f', or 'reboot -f & exit' Hm, this is what I did with about 15 machines yesterday. But this is really bad way.

> via ssh. If login still worked it was luck because it seems that any login produces zombie processes. Some of our machines had already "two many open files" "to many processes" ... couldn't do nothing anymore.

> There is ongoing discussion and the current patch offered for > testing > has not fixed the problem, > > I downgraded all our systems. So far it seems only those that have > had systemd segfault require a reboot after the downgrade. > > On 2015-02-19 05:20, Ruediger Meier wrote: >> On Wednesday 18 February 2015, Michael J Dur wrote: >>> Hello Moby, >>> >>> There is a bug open that is the explanation for this behaviour. >>> You can read the details here: >>> >>> https://bugzilla.opensuse.org/show_bug.cgi?id=918226 >>> >>> For the moment you can go back to the previous version of systemd >>> (and add a lock so it doesn't upgrade until the fix is out) with >>> this command: >>> >>> zypper -n in --oldpackage systemd-32bit-208-23.3.x86_64 >>> systemd-208-23.3.x86_64 systemd-sysvinit-208-23.3.x86_64 && zypper >>> al systemd-32bit systemd systemd-sysvinit >> Regarding this workaround. How could I safely reboot such affected >> system? Or at least stop services to remount ro etc. >> >> $ shutdown -r now >> Failed to open /dev/initctl: No such device or address >> Failed to talk to init daemon. >> >> $ systemctl stop named >> Failed to get D-Bus connection: Failed to connect to >> socket /run/systemd/private: Connection refused >> >> Even killing a deamon does not work anymore: >> $ killall named >> $ ps aux | grep named >> named 2027 0.0 0.0 0 0 ? Zs Feb18 0:20 [named] >> <defunct> >> >> >> cu, >> Rudi Downgrading systemd, as suggested by Michael, fixed the issue. Now I have systemd downgraded and locked in zypper do it does not get messed with. A very serious issue, IMHO, since pid 1 process is jacked up, and in a stable release as opposed to tumbleweed/factory. I also saw growing numbers of zombies with the broken version, but the machine did not complain about running out of resources.

Regards,

Any news about a patch ? That is a big problem, on about 30 of my remote systems.... :( Testing the backroll on a virtual machine.... I believe the patch was put into the repos this morning. I saw 208-32.1 come down to our mirror.

-Michael zypper ref and then zypper up gave me the patch and it works fine.

Here no patch avaiable for now...

Any way to force the update ?

Claudio.

Sorry, my fault, the patch is present. systemd-sysvinit-208-32.1.x86_64 Claudio. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org