Re: [gentoo-user] problems getting systemd to work
On Thu, May 15, 2014 at 5:01 PM, Stefan G. Weichinger li...@xunil.at wrote: Am 15.05.2014 22:38, schrieb cov...@ccs.covici.com: image=/boot/vmlinuz-3.6.2-gentoo phew. 3.6.2 is from October 2012 ... Did you recompile it with the suggested options for systemd? Maybe it doesn't matter, but just a thought ... that kernel is quite old. It doesn't matter, at least in theory systemd works with linux-3.0. John, could you please send the output from dracut --print-cmdline? In my case, it lists *ALL* my lvms, and (I think) therefore all of them are mounted. Since your lilo.conf only lists rd.lvm.lv=linux-files/64-root and rd.lvm.lv=linux-files/64-usr, I think that would explain why it doesn't mount the others. If dracut --print-cmdline doesn't print the others, could you try to boot with rd.lvm.lv=linux-files/audio in the kernel command line specified in your lilo.conf? If after booting /audio is mounted, then I think we have found the problem. We'll need just to figure out why dracut --print-cmdline does not print the other lvms. Also, could you try to generate your initramfs again, but this time with add_dracutmodules+=systemd lvm? Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 5:01 PM, Stefan G. Weichinger li...@xunil.at wrote: Am 15.05.2014 22:38, schrieb cov...@ccs.covici.com: image=/boot/vmlinuz-3.6.2-gentoo phew. 3.6.2 is from October 2012 ... Did you recompile it with the suggested options for systemd? Maybe it doesn't matter, but just a thought ... that kernel is quite old. It doesn't matter, at least in theory systemd works with linux-3.0. John, could you please send the output from dracut --print-cmdline? In my case, it lists *ALL* my lvms, and (I think) therefore all of them are mounted. Since your lilo.conf only lists rd.lvm.lv=linux-files/64-root and rd.lvm.lv=linux-files/64-usr, I think that would explain why it doesn't mount the others. If dracut --print-cmdline doesn't print the others, could you try to boot with rd.lvm.lv=linux-files/audio in the kernel command line specified in your lilo.conf? If after booting /audio is mounted, then I think we have found the problem. We'll need just to figure out why dracut --print-cmdline does not print the other lvms. Also, could you try to generate your initramfs again, but this time with add_dracutmodules+=systemd lvm? The secret to activate all the volumes, is to specify the volume groups instead of each volume -- that got them all activated, but systemd still is not working well. The print-cmdline still just prints the volumes necessary to mount the root and user file systems, which makes sense, but why the rest of them do not activate, I have no clue. Now for some systemd problems. The root file system was read only when I logged in, but I could remount it rw -- not sure why this was happening. Some units did start, but most did not. Whenever I tried to start one manually, I got a message like the following: I wrote systemctl start /usr/lib/systemd/system/ntpd.service and got the error that it was unable to start because it could not find ntpd.service.mount:] May 16 01:59:52 ccs kernel: 31systemd[1]: Failed to load configuration for usr-lib-systemd-system-ntpd.service.mount: No such file or directory May 16 01:59:52 ccs kernel: 31systemd[1]: Trying to enqueue job usr-lib-systemd-system-ntpd.service.mount/start/replace May 16 01:59:53 ccs kernel: 31systemd[1]: Sent message type=error sender=n/a destination=n/a object=n/a interface=n/a member=n/a cookie=1 reply_cookie=1 erro\r=Unit usr-lib-systemd-system-ntpd.service.mount failed to load: No such file or directory. May 16 01:59:53 ccs kernel: 31systemd[1]: Failed to process message [type=method_call sender=n/a path=/org/freedesktop/systemd1 interface=org.freedesktop.sys\temd1.Manager member=StartUnit signature=ss]: Unit usr-lib-systemd-system-ntpd.service.mount failed to load: No such file or directory. No matter what unit I tried to start I would get such a message about the service.mount. Also, even though my network names were correct, they did not come up, but I will try to look in the logs to see why not. So we have made some progress, but still a long way to go yet. Note also, that I am not booting into a display manager, just a regular console. What a lot of work just to get the system booted! -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Am 16.05.2014 12:53, schrieb cov...@ccs.covici.com: Now for some systemd problems. The root file system was read only when I logged in, but I could remount it rw -- not sure why this was happening. Some units did start, but most did not. Maybe you only got into emergency mode? Also, even though my network names were correct, they did not come up, but I will try to look in the logs to see why not. So we have made some progress, but still a long way to go yet. Note also, that I am not booting into a display manager, just a regular console. What a lot of work just to get the system booted! Oh yes! very complex stuff all around. If you want send me the log off-list, I am curious to read through it. This really gets a challenge here ;-) Stefan
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 5:01 PM, Stefan G. Weichinger li...@xunil.at wrote: Am 15.05.2014 22:38, schrieb cov...@ccs.covici.com: image=/boot/vmlinuz-3.6.2-gentoo phew. 3.6.2 is from October 2012 ... Did you recompile it with the suggested options for systemd? Maybe it doesn't matter, but just a thought ... that kernel is quite old. It doesn't matter, at least in theory systemd works with linux-3.0. John, could you please send the output from dracut --print-cmdline? In my case, it lists *ALL* my lvms, and (I think) therefore all of them are mounted. Since your lilo.conf only lists rd.lvm.lv=linux-files/64-root and rd.lvm.lv=linux-files/64-usr, I think that would explain why it doesn't mount the others. If dracut --print-cmdline doesn't print the others, could you try to boot with rd.lvm.lv=linux-files/audio in the kernel command line specified in your lilo.conf? If after booting /audio is mounted, then I think we have found the problem. We'll need just to figure out why dracut --print-cmdline does not print the other lvms. Also, could you try to generate your initramfs again, but this time with add_dracutmodules+=systemd lvm? Also, when I booted up, the systemd started a user slice and a new copy of systemd in the slice -- whic I think I don't want -- is this normal? -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
On Fri, May 16, 2014 at 5:53 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 5:01 PM, Stefan G. Weichinger li...@xunil.at wrote: Am 15.05.2014 22:38, schrieb cov...@ccs.covici.com: image=/boot/vmlinuz-3.6.2-gentoo phew. 3.6.2 is from October 2012 ... Did you recompile it with the suggested options for systemd? Maybe it doesn't matter, but just a thought ... that kernel is quite old. It doesn't matter, at least in theory systemd works with linux-3.0. John, could you please send the output from dracut --print-cmdline? In my case, it lists *ALL* my lvms, and (I think) therefore all of them are mounted. Since your lilo.conf only lists rd.lvm.lv=linux-files/64-root and rd.lvm.lv=linux-files/64-usr, I think that would explain why it doesn't mount the others. If dracut --print-cmdline doesn't print the others, could you try to boot with rd.lvm.lv=linux-files/audio in the kernel command line specified in your lilo.conf? If after booting /audio is mounted, then I think we have found the problem. We'll need just to figure out why dracut --print-cmdline does not print the other lvms. Also, could you try to generate your initramfs again, but this time with add_dracutmodules+=systemd lvm? The secret to activate all the volumes, is to specify the volume groups instead of each volume -- that got them all activated, but systemd still is not working well. The print-cmdline still just prints the volumes necessary to mount the root and user file systems, which makes sense, but why the rest of them do not activate, I have no clue. I don't understand the current situation .So now you get ALL your volumes activated, or not? Now for some systemd problems. The root file system was read only when I logged in, but I could remount it rw -- not sure why this was happening. Set systemd.log_level=debug in your command line, and post the exit from journalctl -b. Some units did start, but most did not. Whenever I tried to start one manually, I got a message like the following: I wrote systemctl start /usr/lib/systemd/system/ntpd.service and got the error that it was unable to start because it could not find ntpd.service.mount:] May 16 01:59:52 ccs kernel: 31systemd[1]: Failed to load configuration for usr-lib-systemd-system-ntpd.service.mount: No such file or directory May 16 01:59:52 ccs kernel: 31systemd[1]: Trying to enqueue job usr-lib-systemd-system-ntpd.service.mount/start/replace May 16 01:59:53 ccs kernel: 31systemd[1]: Sent message type=error sender=n/a destination=n/a object=n/a interface=n/a member=n/a cookie=1 reply_cookie=1 erro\r=Unit usr-lib-systemd-system-ntpd.service.mount failed to load: No such file or directory. May 16 01:59:53 ccs kernel: 31systemd[1]: Failed to process message [type=method_call sender=n/a path=/org/freedesktop/systemd1 interface=org.freedesktop.sys\temd1.Manager member=StartUnit signature=ss]: Unit usr-lib-systemd-system-ntpd.service.mount failed to load: No such file or directory. No matter what unit I tried to start I would get such a message about the service.mount. That sounds like a problem with the cgroups hierarchy (which uses a virtual filesystem). I don't remember seeing a problem like that before. Also, even though my network names were correct, they did not come up, but I will try to look in the logs to see why not. systemd will not (AFAIK) start your network, and before the 209 or 210 version it needed helper program (NetwokrManager, connman, ip, ifconfig, etc.) to do it. Now it includes networkd, but you need to set up .network files (like .service files) to configure it. See [1]. So we have made some progress, but still a long way to go yet. Note also, that I am not booting into a display manager, just a regular console. What a lot of work just to get the system booted! Well, you have a setup that is not, by any means, simple. Also, in my experience old LVM configurations seem to cause a lot of troubles to bring to what systemd expects. John, could you also post here your kernel config? Those cgroups errors *may* be related to some missing functionality from the kernel. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
On Fri, May 16, 2014 at 7:46 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 5:01 PM, Stefan G. Weichinger li...@xunil.at wrote: Am 15.05.2014 22:38, schrieb cov...@ccs.covici.com: image=/boot/vmlinuz-3.6.2-gentoo phew. 3.6.2 is from October 2012 ... Did you recompile it with the suggested options for systemd? Maybe it doesn't matter, but just a thought ... that kernel is quite old. It doesn't matter, at least in theory systemd works with linux-3.0. John, could you please send the output from dracut --print-cmdline? In my case, it lists *ALL* my lvms, and (I think) therefore all of them are mounted. Since your lilo.conf only lists rd.lvm.lv=linux-files/64-root and rd.lvm.lv=linux-files/64-usr, I think that would explain why it doesn't mount the others. If dracut --print-cmdline doesn't print the others, could you try to boot with rd.lvm.lv=linux-files/audio in the kernel command line specified in your lilo.conf? If after booting /audio is mounted, then I think we have found the problem. We'll need just to figure out why dracut --print-cmdline does not print the other lvms. Also, could you try to generate your initramfs again, but this time with add_dracutmodules+=systemd lvm? Also, when I booted up, the systemd started a user slice and a new copy of systemd in the slice -- whic I think I don't want -- is this normal? Yes: every user session gets its own cgroup and systemd --user instance. You can track them with loginctl. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
On Fri, May 16, 2014 at 12:17 PM, cov...@ccs.covici.com wrote:
[snip]
I don't understand the current situation .So now you get ALL your
volumes activated, or not?
Yep, they are all activated and they all get mounted.
Cool, one problem less.
Now for some systemd problems. The root file system was read only when
I logged in, but I could remount it rw -- not sure why this was
happening.
Set systemd.log_level=debug in your command line, and post the exit
from journalctl -b.
I had debug in the command line by itself, would that make the correct
log_level? The file is quite large, should I send it to you privately?
I don't think is necessary, I may have found the real problem (see below).
Some units did start, but most did not. Whenever I tried to
start one manually, I got a message like the following:
[snip]
No matter what unit I tried to start I would get such a message about
the service.mount.
That sounds like a problem with the cgroups hierarchy (which uses a
virtual filesystem). I don't remember seeing a problem like that
before.
Also, even though my network names were correct, they did not come up,
but I will try to look in the logs to see why not.
I wrote a service file to start my network adaptors, here it is:
network@.service
[Unit]
Description=Network Connectivity for %i
Wants=network.target
Before=network.target
BindsTo=sys-subsystem-net-devices-%i.device
After=sys-subsystem-net-devices-%i.device
[Service]
Type=oneshot
RemainAfterExit=yes
EnvironmentFile=/etc/conf.d/network@%i
ExecStart=/usr/bin/ip link set dev %i up
ExecStart=/usr/bin/ip addr add ${address}/${netmask} broadcast ${broadcast}
dev %i
ExecStart=/bin/bash -c 'test -n ${gateway} /usr/bin/ip route add default
via
${gateway}'
ExecStart=/bin/bash -c 'test -f /etc/conf.d/postup@%i.sh/bin/bash
/etc/conf.d/postup@%i.sh
ExecStop=/usr/bin/ip addr flush dev %i
ExecStop=/usr/bin/ip link set dev %i down
[Install]
WantedBy=network.target
Did you enabled network@ifaca.service? Also, WantedBy=network.target
doesn't do what you probably think it does (check [1]... and BTW, I
forgot my last footnote, is now on [2]).
I would use WantedBy=multi-user.target.
systemd will not (AFAIK) start your network, and before the 209 or 210
version it needed helper program (NetwokrManager, connman, ip,
ifconfig, etc.) to do it. Now it includes networkd, but you need to
set up .network files (like .service files) to configure it. See [1].
So we have made some progress, but still a long way to go yet. Note
also, that I am not booting into a display manager, just a regular
console.
What a lot of work just to get the system booted!
Well, you have a setup that is not, by any means, simple. Also, in my
experience old LVM configurations seem to cause a lot of troubles to
bring to what systemd expects.
John, could you also post here your kernel config? Those cgroups
errors *may* be related to some missing functionality from the kernel.
[snip kernel config]
John, your kernel is incorrectly configured to be used by systemd.
When you installed systemd, a warning should have appeared about some
missing configure options; you either didn't saw or ignored those
warnings. Install systemd again so you can see them.
From what I can tell, you are missing *AT LEAST* the following options:
CONFIG_AUTOFS4_FS
CONFIG_DMIID
CONFIG_FANOTIFY
CONFIG_INOTIFY_USER
CONFIG_PROC_FS
CONFIG_SYSFS
John, if you don't set them, systemd *CANNOT WORK PROPERLY*. They are
mandatory. I'm surprised you are able to boot to a semi-working state.
Yes, migrating to systemd is a lot of work. But if you don't see (or
ignore) your system messages, that work gets multiplied several times.
Reconfigure, recompile, and reinstall your kernel (don't forget to
reinstall the modules!), regenerate your initramfs, update lilo (if I
remember correctly, you need to run lilo -something-or-another every
time you change kernel and/or initramfs), and try again.
Regards
[1] http://www.freedesktop.org/wiki/Software/systemd/NetworkTarget/
[2] http://www.freedesktop.org/software/systemd/man/systemd.network.html
--
Canek Peláez Valdés
Profesor de asignatura, Facultad de Ciencias
Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote:
On Fri, May 16, 2014 at 12:17 PM, cov...@ccs.covici.com wrote:
[snip]
I don't understand the current situation .So now you get ALL your
volumes activated, or not?
Yep, they are all activated and they all get mounted.
Cool, one problem less.
Now for some systemd problems. The root file system was read only when
I logged in, but I could remount it rw -- not sure why this was
happening.
Set systemd.log_level=debug in your command line, and post the exit
from journalctl -b.
I had debug in the command line by itself, would that make the correct
log_level? The file is quite large, should I send it to you privately?
I don't think is necessary, I may have found the real problem (see below).
Some units did start, but most did not. Whenever I tried to
start one manually, I got a message like the following:
[snip]
No matter what unit I tried to start I would get such a message about
the service.mount.
That sounds like a problem with the cgroups hierarchy (which uses a
virtual filesystem). I don't remember seeing a problem like that
before.
Also, even though my network names were correct, they did not come up,
but I will try to look in the logs to see why not.
I wrote a service file to start my network adaptors, here it is:
network@.service
[Unit]
Description=Network Connectivity for %i
Wants=network.target
Before=network.target
BindsTo=sys-subsystem-net-devices-%i.device
After=sys-subsystem-net-devices-%i.device
[Service]
Type=oneshot
RemainAfterExit=yes
EnvironmentFile=/etc/conf.d/network@%i
ExecStart=/usr/bin/ip link set dev %i up
ExecStart=/usr/bin/ip addr add ${address}/${netmask} broadcast ${broadcast}
dev %i
ExecStart=/bin/bash -c 'test -n ${gateway} /usr/bin/ip route add default
via
${gateway}'
ExecStart=/bin/bash -c 'test -f /etc/conf.d/postup@%i.sh/bin/bash
/etc/conf.d/postup@%i.sh
ExecStop=/usr/bin/ip addr flush dev %i
ExecStop=/usr/bin/ip link set dev %i down
[Install]
WantedBy=network.target
Did you enabled network@ifaca.service? Also, WantedBy=network.target
doesn't do what you probably think it does (check [1]... and BTW, I
forgot my last footnote, is now on [2]).
I would use WantedBy=multi-user.target.
systemd will not (AFAIK) start your network, and before the 209 or 210
version it needed helper program (NetwokrManager, connman, ip,
ifconfig, etc.) to do it. Now it includes networkd, but you need to
set up .network files (like .service files) to configure it. See [1].
So we have made some progress, but still a long way to go yet. Note
also, that I am not booting into a display manager, just a regular
console.
What a lot of work just to get the system booted!
Well, you have a setup that is not, by any means, simple. Also, in my
experience old LVM configurations seem to cause a lot of troubles to
bring to what systemd expects.
John, could you also post here your kernel config? Those cgroups
errors *may* be related to some missing functionality from the kernel.
[snip kernel config]
John, your kernel is incorrectly configured to be used by systemd.
When you installed systemd, a warning should have appeared about some
missing configure options; you either didn't saw or ignored those
warnings. Install systemd again so you can see them.
From what I can tell, you are missing *AT LEAST* the following options:
CONFIG_AUTOFS4_FS
configured as a module.
CONFIG_DMIID
set to Y
CONFIG_FANOTIFY
set to y
CONFIG_INOTIFY_USER
set to y
CONFIG_PROC_FS
set to y
CONFIG_SYSFS
set to y
John, if you don't set them, systemd *CANNOT WORK PROPERLY*. They are
mandatory. I'm surprised you are able to boot to a semi-working state.
Yes, migrating to systemd is a lot of work. But if you don't see (or
ignore) your system messages, that work gets multiplied several times.
Reconfigure, recompile, and reinstall your kernel (don't forget to
reinstall the modules!), regenerate your initramfs, update lilo (if I
remember correctly, you need to run lilo -something-or-another every
time you change kernel and/or initramfs), and try again.
Regards
[1] http://www.freedesktop.org/wiki/Software/systemd/NetworkTarget/
[2] http://www.freedesktop.org/software/systemd/man/systemd.network.html
--
Canek Peláez Valdés
Profesor de asignatura, Facultad de Ciencias
Universidad Nacional Autónoma de México
--
Your life is like a penny. You're going to lose it. The question is:
How do
you spend it?
John Covici
cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
On Fri, May 16, 2014 at 1:11 PM, cov...@ccs.covici.com wrote:
Canek Peláez Valdés can...@gmail.com wrote:
On Fri, May 16, 2014 at 12:17 PM, cov...@ccs.covici.com wrote:
[snip]
I don't understand the current situation .So now you get ALL your
volumes activated, or not?
Yep, they are all activated and they all get mounted.
Cool, one problem less.
Now for some systemd problems. The root file system was read only when
I logged in, but I could remount it rw -- not sure why this was
happening.
Set systemd.log_level=debug in your command line, and post the exit
from journalctl -b.
I had debug in the command line by itself, would that make the correct
log_level? The file is quite large, should I send it to you privately?
I don't think is necessary, I may have found the real problem (see below).
Some units did start, but most did not. Whenever I tried to
start one manually, I got a message like the following:
[snip]
No matter what unit I tried to start I would get such a message about
the service.mount.
That sounds like a problem with the cgroups hierarchy (which uses a
virtual filesystem). I don't remember seeing a problem like that
before.
Also, even though my network names were correct, they did not come up,
but I will try to look in the logs to see why not.
I wrote a service file to start my network adaptors, here it is:
network@.service
[Unit]
Description=Network Connectivity for %i
Wants=network.target
Before=network.target
BindsTo=sys-subsystem-net-devices-%i.device
After=sys-subsystem-net-devices-%i.device
[Service]
Type=oneshot
RemainAfterExit=yes
EnvironmentFile=/etc/conf.d/network@%i
ExecStart=/usr/bin/ip link set dev %i up
ExecStart=/usr/bin/ip addr add ${address}/${netmask} broadcast
${broadcast} dev %i
ExecStart=/bin/bash -c 'test -n ${gateway} /usr/bin/ip route add
default via
${gateway}'
ExecStart=/bin/bash -c 'test -f /etc/conf.d/postup@%i.sh/bin/bash
/etc/conf.d/postup@%i.sh
ExecStop=/usr/bin/ip addr flush dev %i
ExecStop=/usr/bin/ip link set dev %i down
[Install]
WantedBy=network.target
Did you enabled network@ifaca.service? Also, WantedBy=network.target
doesn't do what you probably think it does (check [1]... and BTW, I
forgot my last footnote, is now on [2]).
I would use WantedBy=multi-user.target.
systemd will not (AFAIK) start your network, and before the 209 or 210
version it needed helper program (NetwokrManager, connman, ip,
ifconfig, etc.) to do it. Now it includes networkd, but you need to
set up .network files (like .service files) to configure it. See [1].
So we have made some progress, but still a long way to go yet. Note
also, that I am not booting into a display manager, just a regular
console.
What a lot of work just to get the system booted!
Well, you have a setup that is not, by any means, simple. Also, in my
experience old LVM configurations seem to cause a lot of troubles to
bring to what systemd expects.
John, could you also post here your kernel config? Those cgroups
errors *may* be related to some missing functionality from the kernel.
[snip kernel config]
John, your kernel is incorrectly configured to be used by systemd.
When you installed systemd, a warning should have appeared about some
missing configure options; you either didn't saw or ignored those
warnings. Install systemd again so you can see them.
From what I can tell, you are missing *AT LEAST* the following options:
CONFIG_AUTOFS4_FS
configured as a module.
CONFIG_DMIID
set to Y
CONFIG_FANOTIFY
set to y
CONFIG_INOTIFY_USER
set to y
CONFIG_PROC_FS
set to y
CONFIG_SYSFS
set to y
I beg your pardon; GMail cut the config file and I didn't notice.
Well then, please send me privately the output from journalctl -b.
Regards.
--
Canek Peláez Valdés
Profesor de asignatura, Facultad de Ciencias
Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote:
On Fri, May 16, 2014 at 1:11 PM, cov...@ccs.covici.com wrote:
Canek Peláez Valdés can...@gmail.com wrote:
On Fri, May 16, 2014 at 12:17 PM, cov...@ccs.covici.com wrote:
[snip]
I don't understand the current situation .So now you get ALL your
volumes activated, or not?
Yep, they are all activated and they all get mounted.
Cool, one problem less.
Now for some systemd problems. The root file system was read only
when
I logged in, but I could remount it rw -- not sure why this was
happening.
Set systemd.log_level=debug in your command line, and post the exit
from journalctl -b.
I had debug in the command line by itself, would that make the correct
log_level? The file is quite large, should I send it to you privately?
I don't think is necessary, I may have found the real problem (see below).
Some units did start, but most did not. Whenever I tried to
start one manually, I got a message like the following:
[snip]
No matter what unit I tried to start I would get such a message about
the service.mount.
That sounds like a problem with the cgroups hierarchy (which uses a
virtual filesystem). I don't remember seeing a problem like that
before.
Also, even though my network names were correct, they did not come up,
but I will try to look in the logs to see why not.
I wrote a service file to start my network adaptors, here it is:
network@.service
[Unit]
Description=Network Connectivity for %i
Wants=network.target
Before=network.target
BindsTo=sys-subsystem-net-devices-%i.device
After=sys-subsystem-net-devices-%i.device
[Service]
Type=oneshot
RemainAfterExit=yes
EnvironmentFile=/etc/conf.d/network@%i
ExecStart=/usr/bin/ip link set dev %i up
ExecStart=/usr/bin/ip addr add ${address}/${netmask} broadcast
${broadcast} dev %i
ExecStart=/bin/bash -c 'test -n ${gateway} /usr/bin/ip route add
default via
${gateway}'
ExecStart=/bin/bash -c 'test -f /etc/conf.d/postup@%i.sh/bin/bash
/etc/conf.d/postup@%i.sh
ExecStop=/usr/bin/ip addr flush dev %i
ExecStop=/usr/bin/ip link set dev %i down
[Install]
WantedBy=network.target
Did you enabled network@ifaca.service? Also, WantedBy=network.target
doesn't do what you probably think it does (check [1]... and BTW, I
forgot my last footnote, is now on [2]).
I would use WantedBy=multi-user.target.
systemd will not (AFAIK) start your network, and before the 209 or 210
version it needed helper program (NetwokrManager, connman, ip,
ifconfig, etc.) to do it. Now it includes networkd, but you need to
set up .network files (like .service files) to configure it. See [1].
So we have made some progress, but still a long way to go yet. Note
also, that I am not booting into a display manager, just a regular
console.
What a lot of work just to get the system booted!
Well, you have a setup that is not, by any means, simple. Also, in my
experience old LVM configurations seem to cause a lot of troubles to
bring to what systemd expects.
John, could you also post here your kernel config? Those cgroups
errors *may* be related to some missing functionality from the kernel.
[snip kernel config]
John, your kernel is incorrectly configured to be used by systemd.
When you installed systemd, a warning should have appeared about some
missing configure options; you either didn't saw or ignored those
warnings. Install systemd again so you can see them.
From what I can tell, you are missing *AT LEAST* the following options:
CONFIG_AUTOFS4_FS
configured as a module.
CONFIG_DMIID
set to Y
CONFIG_FANOTIFY
set to y
CONFIG_INOTIFY_USER
set to y
CONFIG_PROC_FS
set to y
CONFIG_SYSFS
set to y
I beg your pardon; GMail cut the config file and I didn't notice.
Well then, please send me privately the output from journalctl -b.
I did find one error so far, I had network@.eth0 and eth2 when I should
have had network@eth0 and eth2 -- thanks to Stefan.
I will send you what logs I have.
Thanks.
Regards.
--
Canek Peláez Valdés
Profesor de asignatura, Facultad de Ciencias
Universidad Nacional Autónoma de México
--
Your life is like a penny. You're going to lose it. The question is:
How do
you spend it?
John Covici
cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. The thing is, I *ALWAYS* use the -H (host only) option, so it was my understanding that the command line should be automatically generated, and it was on 036, but now it doesn't in 037. I think that's a bug. John, with respect to your case: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
On Thu, May 15, 2014 at 1:18 AM, Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. I'm an idiot; I didn't saw the documentation about hostonly_cmdline; BTW Jc, you used host_cmdline, I think the former is the correct one. So, to resume: there is no bug, is just that before hostonly_cmdline was yes by default, and now is no by default. This change was documented, but I failed to notice it (and I think the ebuild in Gentoo should print an einfo message). Anyway, I think that explains all my problems; John, I don't know if it will solve yours. Again: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? And finally, have you tried with --hostonly-cmdline? Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. The thing is, I *ALWAYS* use the -H (host only) option, so it was my understanding that the command line should be automatically generated, and it was on 036, but now it doesn't in 037. I think that's a bug. John, with respect to your case: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? I did not try the -H, I may test with that later. I did look at the --print-cmdline and copied the volumes they mentioned, but I have other lvm volumes in my fstab and none of them were activated, only the ones I specified in the command line! This is where I have run into problems. I have quite a few lvms, I want them all activated! Also, since I wrote the last message, I have been looking at the journalctl output and discovered a couple of things which I would like some help on, but getting the lvms to work is more important. First, whatever happened to DefaultControllers -- I want to disable those cpu hierarchies, but that option seems to have disappeared without a trace, although you can google and see it in some documentation. The keyword also was not accepted in an install section I have, what is the matter with that? I want to use my sysklogd for my syslog, how can I use that with systemd? Thanks so much for all your help on these things. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 1:18 AM, Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. I'm an idiot; I didn't saw the documentation about hostonly_cmdline; BTW Jc, you used host_cmdline, I think the former is the correct one. So, to resume: there is no bug, is just that before hostonly_cmdline was yes by default, and now is no by default. This change was documented, but I failed to notice it (and I think the ebuild in Gentoo should print an einfo message). Anyway, I think that explains all my problems; John, I don't know if it will solve yours. Again: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? And finally, have you tried with --hostonly-cmdline? I did not use --hostonly-cmdline because I have other parameters and wanted to keep everyting together -- should that make a difference? If I did use that then would I have append= in lilo.conf or the grub equivalemt? -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Am 15.05.2014 11:39, schrieb cov...@ccs.covici.com: I did not try the -H, I may test with that later. I did look at the --print-cmdline and copied the volumes they mentioned, but I have other lvm volumes in my fstab and none of them were activated, only the ones I specified in the command line! This is where I have run into problems. I have quite a few lvms, I want them all activated! Sure. I remember having an extra lvm.service for systemd to have all the LVs activated ... with that unit-file it worked more reliably for me (maybe not needed since some time). For sure that service file is only run *after* the initrd has found/activated/mounted your LVM-based root ... might be a workaround to specify the root-LV in the kernel command line (plus maybe rd.auto rd.lvm=1 ?) and then let the service file activate the rest of the LVs. Just to get you started at last ;-) Also, since I wrote the last message, I have been looking at the journalctl output and discovered a couple of things which I would like some help on, but getting the lvms to work is more important. First, whatever happened to DefaultControllers -- I want to disable those cpu hierarchies, but that option seems to have disappeared without a trace, although you can google and see it in some documentation. The keyword also was not accepted in an install section I have, what is the matter with that? What keyword? I don't understand right now. I want to use my sysklogd for my syslog, how can I use that with systemd? systemd's journal will be written to a socket if you configure it in /etc/systemd/journald.conf I would check man journald.conf and the option: ForwardToSyslog= and then let your chosen log-daemon listen there. IMO you should take a look at journalctl then anyway ... new concepts, but powerful features. Stefan
Re: [gentoo-user] problems getting systemd to work
Stefan G. Weichinger li...@xunil.at wrote: Am 15.05.2014 11:39, schrieb cov...@ccs.covici.com: I did not try the -H, I may test with that later. I did look at the --print-cmdline and copied the volumes they mentioned, but I have other lvm volumes in my fstab and none of them were activated, only the ones I specified in the command line! This is where I have run into problems. I have quite a few lvms, I want them all activated! Sure. I remember having an extra lvm.service for systemd to have all the LVs activated ... with that unit-file it worked more reliably for me (maybe not needed since some time). For sure that service file is only run *after* the initrd has found/activated/mounted your LVM-based root ... might be a workaround to specify the root-LV in the kernel command line (plus maybe rd.auto rd.lvm=1 ?) and then let the service file activate the rest of the LVs. Just to get you started at last ;-) Also, since I wrote the last message, I have been looking at the journalctl output and discovered a couple of things which I would like some help on, but getting the lvms to work is more important. First, whatever happened to DefaultControllers -- I want to disable those cpu hierarchies, but that option seems to have disappeared without a trace, although you can google and see it in some documentation. The keyword also was not accepted in an install section I have, what is the matter with that? What keyword? I don't understand right now. I want to use my sysklogd for my syslog, how can I use that with systemd? systemd's journal will be written to a socket if you configure it in /etc/systemd/journald.conf I would check man journald.conf and the option: ForwardToSyslog= and then let your chosen log-daemon listen there. IMO you should take a look at journalctl then anyway ... new concepts, but powerful features. Sure, but what I was looking for was a way to start syslogd and klogd using systemd -- I do have a socket option so they can listen on the socket so that should be OK. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 1:18 AM, Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. I'm an idiot; I didn't saw the documentation about hostonly_cmdline; BTW Jc, you used host_cmdline, I think the former is the correct one. So, to resume: there is no bug, is just that before hostonly_cmdline was yes by default, and now is no by default. This change was documented, but I failed to notice it (and I think the ebuild in Gentoo should print an einfo message). Anyway, I think that explains all my problems; John, I don't know if it will solve yours. Again: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? And finally, have you tried with --hostonly-cmdline? OK, I was looking through the journal output and I think the key to the lvm's not activating is the following: 4 12:54:57 ccs systemd[1]: Got notification message for unit systemd-journald.service 4 12:54:57 ccs systemd[1]: systemd-journald.service: Got notification message from PID 1750 (WATCHDOG=1...) 4 12:54:57 ccs systemd[1]: systemd-journald.service: got WATCHDOG=1 4 12:54:57 ccs systemd[1]: Received SIGCHLD from PID 2603 (lvm). 4 12:54:57 ccs systemd[1]: Child 2602 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Child 2603 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Child 2610 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Job dev-mapper-linux\x2d\x2dfiles\x2dportage.device/start timed out. So what is not installed? Also, for the first two lines, I get hundreds of thatpair of lines, how can I prevent such. So, between the lvm problem and the udev renaming my eth0 devices these are the key as to why things are going wrong -- with openrc udev is not renaming eth0 at all. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Am 15.05.2014 12:19, schrieb cov...@ccs.covici.com: Sure, but what I was looking for was a way to start syslogd and klogd using systemd -- I do have a socket option so they can listen on the socket so that should be OK. So you look for service files? A quick google finds examples for these 2 daemons here: http://www.linuxfromscratch.org/hints/downloads/files/systemd.txt If they work it would be great to file a bug for adding systemd unit files to app-admin/sysklogd at bugs.gentoo.org ( I didn't check if the ebuild brings unit-files but at least I see it doesn't have a systemd USE flag). Stefan
Re: [gentoo-user] problems getting systemd to work
Am 15.05.2014 13:50, schrieb cov...@ccs.covici.com: Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 1:18 AM, Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. I'm an idiot; I didn't saw the documentation about hostonly_cmdline; BTW Jc, you used host_cmdline, I think the former is the correct one. So, to resume: there is no bug, is just that before hostonly_cmdline was yes by default, and now is no by default. This change was documented, but I failed to notice it (and I think the ebuild in Gentoo should print an einfo message). Anyway, I think that explains all my problems; John, I don't know if it will solve yours. Again: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? And finally, have you tried with --hostonly-cmdline? OK, I was looking through the journal output and I think the key to the lvm's not activating is the following: 4 12:54:57 ccs systemd[1]: Got notification message for unit systemd-journald.service 4 12:54:57 ccs systemd[1]: systemd-journald.service: Got notification message from PID 1750 (WATCHDOG=1...) 4 12:54:57 ccs systemd[1]: systemd-journald.service: got WATCHDOG=1 4 12:54:57 ccs systemd[1]: Received SIGCHLD from PID 2603 (lvm). 4 12:54:57 ccs systemd[1]: Child 2602 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Child 2603 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Child 2610 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Job dev-mapper-linux\x2d\x2dfiles\x2dportage.device/start timed out. So what is not installed? My search tells me that this might be a misinterpreted return code. I might repeat myself but the thread gets quite large now: Did you enable lvm2-lvmetad.service or socket (and set use_lvmetad=1 in lvm.conf)? I think you don't have to, I just ask to check. What release of lvm2, btw? Also, for the first two lines, I get hundreds of thatpair of lines, how can I prevent such. The PID1 stuff ? So, between the lvm problem and the udev renaming my eth0 devices these are the key as to why things are going wrong -- with openrc udev is not renaming eth0 at all. We'll take care of eth0 as well as soon your box boots correctly ;-) Stefan
Re: [gentoo-user] problems getting systemd to work
Stefan G. Weichinger li...@xunil.at wrote: Am 15.05.2014 12:19, schrieb cov...@ccs.covici.com: Sure, but what I was looking for was a way to start syslogd and klogd using systemd -- I do have a socket option so they can listen on the socket so that should be OK. So you look for service files? A quick google finds examples for these 2 daemons here: http://www.linuxfromscratch.org/hints/downloads/files/systemd.txt If they work it would be great to file a bug for adding systemd unit files to app-admin/sysklogd at bugs.gentoo.org ( I didn't check if the ebuild brings unit-files but at least I see it doesn't have a systemd USE flag). OK, I will check, I did not see that in my search last time. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Stefan G. Weichinger li...@xunil.at wrote: Am 15.05.2014 13:50, schrieb cov...@ccs.covici.com: Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 1:18 AM, Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. I'm an idiot; I didn't saw the documentation about hostonly_cmdline; BTW Jc, you used host_cmdline, I think the former is the correct one. So, to resume: there is no bug, is just that before hostonly_cmdline was yes by default, and now is no by default. This change was documented, but I failed to notice it (and I think the ebuild in Gentoo should print an einfo message). Anyway, I think that explains all my problems; John, I don't know if it will solve yours. Again: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? And finally, have you tried with --hostonly-cmdline? OK, I was looking through the journal output and I think the key to the lvm's not activating is the following: 4 12:54:57 ccs systemd[1]: Got notification message for unit systemd-journald.service 4 12:54:57 ccs systemd[1]: systemd-journald.service: Got notification message from PID 1750 (WATCHDOG=1...) 4 12:54:57 ccs systemd[1]: systemd-journald.service: got WATCHDOG=1 4 12:54:57 ccs systemd[1]: Received SIGCHLD from PID 2603 (lvm). 4 12:54:57 ccs systemd[1]: Child 2602 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Child 2603 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Child 2610 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Job dev-mapper-linux\x2d\x2dfiles\x2dportage.device/start timed out. So what is not installed? My search tells me that this might be a misinterpreted return code. I might repeat myself but the thread gets quite large now: Did you enable lvm2-lvmetad.service or socket (and set use_lvmetad=1 in lvm.conf)? Yep, did not see that starting. I think you don't have to, I just ask to check. What release of lvm2, btw? 105-r2 Also, for the first two lines, I get hundreds of thatpair of lines, how can I prevent such. The PID1 stuff ? Nope, the notifications about journald and watchdog=1 So, between the lvm problem and the udev renaming my eth0 devices these are the key as to why things are going wrong -- with openrc udev is not renaming eth0 at all. We'll take care of eth0 as well as soon your box boots correctly ;-) Stefan -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Am 15.05.2014 14:38, schrieb cov...@ccs.covici.com: Stefan G. Weichinger li...@xunil.at wrote: Am 15.05.2014 13:50, schrieb cov...@ccs.covici.com: Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 1:18 AM, Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. I'm an idiot; I didn't saw the documentation about hostonly_cmdline; BTW Jc, you used host_cmdline, I think the former is the correct one. So, to resume: there is no bug, is just that before hostonly_cmdline was yes by default, and now is no by default. This change was documented, but I failed to notice it (and I think the ebuild in Gentoo should print an einfo message). Anyway, I think that explains all my problems; John, I don't know if it will solve yours. Again: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? And finally, have you tried with --hostonly-cmdline? OK, I was looking through the journal output and I think the key to the lvm's not activating is the following: 4 12:54:57 ccs systemd[1]: Got notification message for unit systemd-journald.service 4 12:54:57 ccs systemd[1]: systemd-journald.service: Got notification message from PID 1750 (WATCHDOG=1...) 4 12:54:57 ccs systemd[1]: systemd-journald.service: got WATCHDOG=1 4 12:54:57 ccs systemd[1]: Received SIGCHLD from PID 2603 (lvm). 4 12:54:57 ccs systemd[1]: Child 2602 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Child 2603 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Child 2610 (lvm) died (code=exited, status=5/NOTINSSTALLED) 4 12:54:57 ccs systemd[1]: Job dev-mapper-linux\x2d\x2dfiles\x2dportage.device/start timed out. So what is not installed? My search tells me that this might be a misinterpreted return code. I might repeat myself but the thread gets quite large now: Did you enable lvm2-lvmetad.service or socket (and set use_lvmetad=1 in lvm.conf)? Yep, did not see that starting. I think you don't have to, I just ask to check. What release of lvm2, btw? 105-r2 Would you test downgrading to 2.02.104 for checking if that changes something? Or 2.02.106 ... I find various bugs on b.g.o. around lvm2 Stefan
Re: [gentoo-user] problems getting systemd to work
2014-05-15 0:47 GMT-06:00 Canek Peláez Valdés can...@gmail.com: I'm an idiot; I didn't saw the documentation about hostonly_cmdline; BTW Jc, you used host_cmdline, I think the former is the correct one. I don't know right now how host_cmdline got in my config(that was a crazy day between dracut, systemd, and lvm documentations), who knows where I saw it or if I mistyped it, anyway thanks for the correction. So, to resume: there is no bug, is just that before hostonly_cmdline was yes by default, and now is no by default. This change was documented, but I failed to notice it (and I think the ebuild in Gentoo should print an einfo message). I agree, that notification in the ebulid would have saved trouble, and I think it should be included since for sure many will run into these problems in future. So, would this qualify for a gentoo bug, and make the developer include that notification? Again, thanks for clarifying more the working of dracut. Anyway, I think that explains all my problems; John, I don't know if it will solve yours. Again: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? And finally, have you tried with --hostonly-cmdline? Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
2014-05-15 6:38 GMT-06:00 cov...@ccs.covici.com: Did you enable lvm2-lvmetad.service or socket (and set use_lvmetad=1 in lvm.conf)? Yep, did not see that starting. I have seen an odd behaviour regarding this sometimes, particulary when I upgrade kernel, for some reason in my first reboot, my other LVs (I have another 2 VGs) aren't activated, but on the second reboot(Ctrl+Alt+Supr when it has got stuck), it works fine. this has happened to me in the last 4 kernel upgrades.
Re: [gentoo-user] problems getting systemd to work
On Thu, May 15, 2014 at 2:18 AM, Canek Peláez Valdés can...@gmail.com wrote: OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. The thing is, I *ALWAYS* use the -H (host only) option, so it was my understanding that the command line should be automatically generated, and it was on 036, but now it doesn't in 037. I think that's a bug. John, with respect to your case: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? There are two new options, --hostonly-cmdline (store kernel command line arguments needed in the initramfs) and --no-hostonly-cmdline (Do not store kernel command line arguments needed in the initramfs).
Re: [gentoo-user] problems getting systemd to work
On Thu, May 15, 2014 at 4:48 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 1:18 AM, Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. I'm an idiot; I didn't saw the documentation about hostonly_cmdline; BTW Jc, you used host_cmdline, I think the former is the correct one. So, to resume: there is no bug, is just that before hostonly_cmdline was yes by default, and now is no by default. This change was documented, but I failed to notice it (and I think the ebuild in Gentoo should print an einfo message). Anyway, I think that explains all my problems; John, I don't know if it will solve yours. Again: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? And finally, have you tried with --hostonly-cmdline? I did not use --hostonly-cmdline because I have other parameters and wanted to keep everyting together -- should that make a difference? I think so. If I did use that then would I have append= in lilo.conf or the grub equivalemt? Nothing; dracut would take care of everything (in theory). Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
On Thu, May 15, 2014 at 4:39 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. The thing is, I *ALWAYS* use the -H (host only) option, so it was my understanding that the command line should be automatically generated, and it was on 036, but now it doesn't in 037. I think that's a bug. John, with respect to your case: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? I did not try the -H, I may test with that later. I did look at the --print-cmdline and copied the volumes they mentioned, but I have other lvm volumes in my fstab and none of them were activated, only the ones I specified in the command line! This is where I have run into problems. I have quite a few lvms, I want them all activated! I just have /, /boot, /usr and /home (encrypted) as lvms, but all are activated. Also, since I wrote the last message, I have been looking at the journalctl output and discovered a couple of things which I would like some help on, but getting the lvms to work is more important. First, whatever happened to DefaultControllers -- I want to disable those cpu hierarchies, but that option seems to have disappeared without a trace, although you can google and see it in some documentation. That went away with the new cgroup handling that is being coordinated between systemd and the kernel: http://cgit.freedesktop.org/systemd/systemd/commit/?id=7ac807320a7416463d7ff3ef6ede574863a601c5 The keyword also was not accepted in an install section I have, what is the matter with that? I want to use my sysklogd for my syslog, how can I use that with systemd? I think you can do that with systemd.log_target=kmsg in the kernel command line. Thanks so much for all your help on these things. John, could you please include here the output of lsblk, your fstab, your dracut.conf, and your lilo.conf? Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Am 15.05.2014 20:23, schrieb Canek Peláez Valdés: John, could you please include here the output of lsblk, your fstab, your dracut.conf, and your lilo.conf? .. I agree! it's hard to keep track and overview in here :-)
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Thu, May 15, 2014 at 4:39 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 5:26 PM, cov...@ccs.covici.com wrote: [snip] Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. OK, I was a little mystified about why dracut-036 worked on my system and 037 didn't. Before I tried any workaround, I wanted to know what changed from the previous version to the current one. So I generated an initramfs with dracut-036-r4 and another one with dracut-037-r1, and I tried to see what changed from one to the other. The answer is surprisingly easy: in /etc/cmdline.d/, the following files where in the 036-r4 version, but not in the 037-r4: 90crypt.conf 90lvm.conf 90mdraid.conf base.conf Te contents of those files are (90crypt.conf is empty): 90lvm.conf rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 90mdraid.conf rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 base.conf ro So I just changed my /etc/default/grub file: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash rd.lvm.lv=vg/vol1 rd.lvm.lv=vg/vol4 rd.lvm.lv=vg/vol3 rd.md.uuid=f4a59e68:fbe4039f:a39fc86d:e9e91e12 I regenerated my GRUB2 config, and now again my LVM test system works perfectly with the latest dracut version. The thing is, I *ALWAYS* use the -H (host only) option, so it was my understanding that the command line should be automatically generated, and it was on 036, but now it doesn't in 037. I think that's a bug. John, with respect to your case: did you used dracut --print-cmdline to get the command line? Also, have you tried to use -H to generate your initramfs? I did not try the -H, I may test with that later. I did look at the --print-cmdline and copied the volumes they mentioned, but I have other lvm volumes in my fstab and none of them were activated, only the ones I specified in the command line! This is where I have run into problems. I have quite a few lvms, I want them all activated! I just have /, /boot, /usr and /home (encrypted) as lvms, but all are activated. Also, since I wrote the last message, I have been looking at the journalctl output and discovered a couple of things which I would like some help on, but getting the lvms to work is more important. First, whatever happened to DefaultControllers -- I want to disable those cpu hierarchies, but that option seems to have disappeared without a trace, although you can google and see it in some documentation. That went away with the new cgroup handling that is being coordinated between systemd and the kernel: http://cgit.freedesktop.org/systemd/systemd/commit/?id=7ac807320a7416463d7ff3ef6ede574863a601c5 The keyword also was not accepted in an install section I have, what is the matter with that? I want to use my sysklogd for my syslog, how can I use that with systemd? I think you can do that with systemd.log_target=kmsg in the kernel command line. Thanks so much for all your help on these things. John, could you please include here the output of lsblk, your fstab, your dracut.conf, and your lilo.conf? output of lsblk: NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT sda 8:00 931.5G 0 disk ├─sda1 8:10 3G 0 part /boot ├─sda2 8:20 1000M 0 part /mnt/oldgentoo └─sda3 8:30 927.6G 0 part ├─linux--files-64--root 253:00 3G 0 lvm / ├─linux--files-swap 253:10 2G 0 lvm [SWAP] ├─linux--files-64--usr 253:2065G 0 lvm /usr ├─linux--files-usr--src 253:3035G 0 lvm /usr/src ├─linux--files-tmp 253:4010G 0 lvm /tmp ├─linux--files-64--var 253:5020G 0 lvm /var ├─linux--files-home 253:6030G 0 lvm /home ├─linux--files-audio 253:70 990G 0 lvm /audio ├─linux--files-usr--bbs 253:80 256M 0 lvm /usr/bbs ├─linux--files-hard2 253:90 704M 0 lvm /hard2 ├─linux--files-scratch 253:10
Re: [gentoo-user] problems getting systemd to work
Am 15.05.2014 22:38, schrieb cov...@ccs.covici.com: image=/boot/vmlinuz-3.6.2-gentoo phew. 3.6.2 is from October 2012 ... Did you recompile it with the suggested options for systemd? Maybe it doesn't matter, but just a thought ... that kernel is quite old. Stefan
Re: [gentoo-user] problems getting systemd to work
Stefan G. Weichinger li...@xunil.at wrote: Am 15.05.2014 22:38, schrieb cov...@ccs.covici.com: image=/boot/vmlinuz-3.6.2-gentoo phew. 3.6.2 is from October 2012 ... Did you recompile it with the suggested options for systemd? Maybe it doesn't matter, but just a thought ... that kernel is quite old. I did recompile a number of times, but for now I have to stick with this one. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:52 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:31 AM, cov...@ccs.covici.com wrote: [snip] OK, I will try dracut, I hope it works with dracut. This is my kernel command line and RAID/LVM related stuff from GRUB2: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash GRUB_PRELOAD_MODULES=lvm mdraid1x And this is my dracut.conf (minus comments): add_dracutmodules+=crypt lvm mdraid systemd add_drivers+=autofs4 ipv6 dm-crypt aes sha256 fscks=umount mount /sbin/fsck* e2fsck That's it. I didn't touched anything else to make dracut+systemd work with LVM and RAID (and LUKS, but that doesn't matter). Also, dracut comes with extensive and very clear documentation; check the man pages included. but I still want to know what systemd is doing, what processes its spawning, etc. -- how can I find this out -- I thought to use the confirm_spawn, but it times out and keeps going, what can I do instead? You can use bootchart: man 1 systemd-bootchart It will produce a chart with all the processes, and how long it takes for every one of them. But remember, the order depends on which one finishes before, and that can change from boot to boot. Thanks people for all your responses, this is a great list. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Well, since I am unable to see, the graph would not do me any good, any way to get it in text form? What I want to see (and I know the order may change) is which starts first and so on, to make sure targets, etc. do what I want them to do. Try adding this to your kernel command line: systemd.log_target=console systemd.log_level=debug. It will add a lot of output, including what is being executed. Everything is documented in the man page: man 1 systemd. Well, I rebooted under dracut, but it did not do the lvmscan and so the job trying to find the root file system timed out after 90 seconds. It took me to the emergency shell which I had specified, and I was able to do the lvm_scan and them magically root got mounted under sysroot, but I had no idea what to do next to maybe get things going. So how can I get dracut to do its lvm_scan -- I even added the line add_dracutmodules+=lvm but no joy. I saved the rdsosreport.txt and if that is of any use, I can post it. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:52 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:31 AM, cov...@ccs.covici.com wrote: [snip] OK, I will try dracut, I hope it works with dracut. This is my kernel command line and RAID/LVM related stuff from GRUB2: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash GRUB_PRELOAD_MODULES=lvm mdraid1x And this is my dracut.conf (minus comments): add_dracutmodules+=crypt lvm mdraid systemd add_drivers+=autofs4 ipv6 dm-crypt aes sha256 fscks=umount mount /sbin/fsck* e2fsck That's it. I didn't touched anything else to make dracut+systemd work with LVM and RAID (and LUKS, but that doesn't matter). Also, dracut comes with extensive and very clear documentation; check the man pages included. but I still want to know what systemd is doing, what processes its spawning, etc. -- how can I find this out -- I thought to use the confirm_spawn, but it times out and keeps going, what can I do instead? You can use bootchart: man 1 systemd-bootchart It will produce a chart with all the processes, and how long it takes for every one of them. But remember, the order depends on which one finishes before, and that can change from boot to boot. Thanks people for all your responses, this is a great list. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Well, since I am unable to see, the graph would not do me any good, any way to get it in text form? What I want to see (and I know the order may change) is which starts first and so on, to make sure targets, etc. do what I want them to do. Try adding this to your kernel command line: systemd.log_target=console systemd.log_level=debug. It will add a lot of output, including what is being executed. I also in my kernel command line said rd.lvm=1 just to make sure. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
2014-05-14 3:43 GMT-06:00 cov...@ccs.covici.com: I also in my kernel command line said rd.lvm=1 just to make sure. Have tried this too(only rd.lvm), until I didn't tell what to recognize nothing was done, this might be a bug, but I forgot about it .
Re: [gentoo-user] problems getting systemd to work
2014-05-14 3:40 GMT-06:00 cov...@ccs.covici.com: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:52 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:31 AM, cov...@ccs.covici.com wrote: [snip] OK, I will try dracut, I hope it works with dracut. This is my kernel command line and RAID/LVM related stuff from GRUB2: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash GRUB_PRELOAD_MODULES=lvm mdraid1x And this is my dracut.conf (minus comments): add_dracutmodules+=crypt lvm mdraid systemd add_drivers+=autofs4 ipv6 dm-crypt aes sha256 fscks=umount mount /sbin/fsck* e2fsck That's it. I didn't touched anything else to make dracut+systemd work with LVM and RAID (and LUKS, but that doesn't matter). Also, dracut comes with extensive and very clear documentation; check the man pages included. but I still want to know what systemd is doing, what processes its spawning, etc. -- how can I find this out -- I thought to use the confirm_spawn, but it times out and keeps going, what can I do instead? You can use bootchart: man 1 systemd-bootchart It will produce a chart with all the processes, and how long it takes for every one of them. But remember, the order depends on which one finishes before, and that can change from boot to boot. Thanks people for all your responses, this is a great list. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Well, since I am unable to see, the graph would not do me any good, any way to get it in text form? What I want to see (and I know the order may change) is which starts first and so on, to make sure targets, etc. do what I want them to do. Try adding this to your kernel command line: systemd.log_target=console systemd.log_level=debug. It will add a lot of output, including what is being executed. Everything is documented in the man page: man 1 systemd. Well, I rebooted under dracut, but it did not do the lvmscan and so the job trying to find the root file system timed out after 90 seconds. It took me to the emergency shell which I had specified, and I was able to do the lvm_scan and them magically root got mounted under sysroot, but I had no idea what to do next to maybe get things going. So how can I get dracut to do its lvm_scan -- I even added the line add_dracutmodules+=lvm This is a must if you are not running 'dracut --lvmconf' but no joy. I saved the rdsosreport.txt and if that is of any use, I can post it. In the dracut manpage, I found the '--print-cmdline' argument , it generates a suggested kernel command line , I hadn't used it before, but this what it printed: rd.lvm.lv=gentoovg/usr rd.lvm.lv=gentoovg/root root=/dev/mapper/gentoovg-root rootflags=defaults rootfstype=ext4 So I guess, its neccesary to have this in the command line. At the time I struggled for an entire day with the same timeout when my LVs where not recognized, did research trial/error with dracut and genkernel, got them both working, stayed with dracut because I liked it, I tried to avoid this struggle for you on my first response, but I see is going to take you longer... -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com PD: RTFM.
Re: [gentoo-user] problems getting systemd to work
On Wed, May 14, 2014 at 4:43 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:52 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:31 AM, cov...@ccs.covici.com wrote: [snip] OK, I will try dracut, I hope it works with dracut. This is my kernel command line and RAID/LVM related stuff from GRUB2: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash GRUB_PRELOAD_MODULES=lvm mdraid1x And this is my dracut.conf (minus comments): add_dracutmodules+=crypt lvm mdraid systemd add_drivers+=autofs4 ipv6 dm-crypt aes sha256 fscks=umount mount /sbin/fsck* e2fsck That's it. I didn't touched anything else to make dracut+systemd work with LVM and RAID (and LUKS, but that doesn't matter). Also, dracut comes with extensive and very clear documentation; check the man pages included. but I still want to know what systemd is doing, what processes its spawning, etc. -- how can I find this out -- I thought to use the confirm_spawn, but it times out and keeps going, what can I do instead? You can use bootchart: man 1 systemd-bootchart It will produce a chart with all the processes, and how long it takes for every one of them. But remember, the order depends on which one finishes before, and that can change from boot to boot. Thanks people for all your responses, this is a great list. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Well, since I am unable to see, the graph would not do me any good, any way to get it in text form? What I want to see (and I know the order may change) is which starts first and so on, to make sure targets, etc. do what I want them to do. Try adding this to your kernel command line: systemd.log_target=console systemd.log_level=debug. It will add a lot of output, including what is being executed. I also in my kernel command line said rd.lvm=1 just to make sure. John, which version of dracut are you using? Also, what was the command line you used to generate it? Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 4:43 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:52 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:31 AM, cov...@ccs.covici.com wrote: [snip] OK, I will try dracut, I hope it works with dracut. This is my kernel command line and RAID/LVM related stuff from GRUB2: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash GRUB_PRELOAD_MODULES=lvm mdraid1x And this is my dracut.conf (minus comments): add_dracutmodules+=crypt lvm mdraid systemd add_drivers+=autofs4 ipv6 dm-crypt aes sha256 fscks=umount mount /sbin/fsck* e2fsck That's it. I didn't touched anything else to make dracut+systemd work with LVM and RAID (and LUKS, but that doesn't matter). Also, dracut comes with extensive and very clear documentation; check the man pages included. but I still want to know what systemd is doing, what processes its spawning, etc. -- how can I find this out -- I thought to use the confirm_spawn, but it times out and keeps going, what can I do instead? You can use bootchart: man 1 systemd-bootchart It will produce a chart with all the processes, and how long it takes for every one of them. But remember, the order depends on which one finishes before, and that can change from boot to boot. Thanks people for all your responses, this is a great list. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Well, since I am unable to see, the graph would not do me any good, any way to get it in text form? What I want to see (and I know the order may change) is which starts first and so on, to make sure targets, etc. do what I want them to do. Try adding this to your kernel command line: systemd.log_target=console systemd.log_level=debug. It will add a lot of output, including what is being executed. I also in my kernel command line said rd.lvm=1 just to make sure. John, which version of dracut are you using? Also, what was the command line you used to generate it? I am using version 037 and the command line was very simple dracut --force -M and here is its output: Script started on Wed 14 May 2014 04:52:19 AM EDT Executing: /usr/bin/dracut --force -M 00bootchart: Could not find command '/sbin/bootchartd'! 00dash: Could not find command '/bin/dash'! 50plymouth: Could not find command 'plymouthd'! 50plymouth: Could not find command 'plymouth'! 90btrfs: Could not find command 'btrfs'! 90dmraid: Could not find command 'dmraid'! 90mdraid: Could not find command 'mdadm'! 90multipath: Could not find command 'multipath'! 95fcoe-uefi: Could not find command 'dcbtool'! 95fcoe-uefi: Could not find command 'fipvlan'! 95fcoe-uefi: Could not find command 'lldpad'! 95iscsi: Could not find command 'iscsistart'! 95iscsi: Could not find command 'iscsi-iname'! 95nbd: Could not find command 'nbd-client'! 97biosdevname: Could not find command 'biosdevname'! bash caps caps: does not work with systemd in the initramfs modsign dracut-install: ERROR: installing '/usr/bin/keyctl' /usr/libexec/dracut-install -D /var/tmp/initramfs.dpEwDV /usr/bin/keyctl i18n network crypt dm Skipping udev rule: 64-device-mapper.rules Skipping udev rule: 60-persistent-storage-dm.rules Skipping udev rule: 55-dm.rules kernel-modules Omitting driver i2o_scsi lvm Skipping udev rule: 64-device-mapper.rules Skipping udev rule: 56-lvm.rules Skipping udev rule: 60-persistent-storage-lvm.rules cifs nfs resume rootfs-block terminfo udev-rules Skipping udev rule: 91-permissions.rules Skipping udev rule: 80-drivers-modprobe.rules systemd usrmount base fs-lib shutdown *** Including modules done *** *** Installing kernel module dependencies and firmware *** *** Installing kernel module dependencies and firmware done *** *** Resolving executable dependencies *** *** Resolving executable dependencies done*** *** Stripping files *** *** Stripping files done *** *** Store current command line parameters *** *** Creating image file *** *** Creating image file done *** Script done on Wed 14 May 2014 04:52:50 AM EDT -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
On Wed, May 14, 2014 at 10:05 AM, cov...@ccs.covici.com wrote: [snip] I am using version 037 and the command line was very simple As I told you in [1], the 037 version has problems with LVM; please also check the thread in [2], specially the post in [3]. Please try again with dracut-036-r4. Regards. [1] http://article.gmane.org/gmane.linux.gentoo.user/274340 [2] http://thread.gmane.org/gmane.linux.gentoo.user/274152 [3] http://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg145647.html (for some reason it doesn't appear in gmane). -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:05 AM, cov...@ccs.covici.com wrote: [snip] I am using version 037 and the command line was very simple As I told you in [1], the 037 version has problems with LVM; please also check the thread in [2], specially the post in [3]. Please try again with dracut-036-r4. Regards. [1] http://article.gmane.org/gmane.linux.gentoo.user/274340 [2] http://thread.gmane.org/gmane.linux.gentoo.user/274152 [3] http://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg145647.html (for some reason it doesn't appear in gmane). portage won't let me downgrade unless I downgrade systemd to 2.10 -- should I do that? -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
2014-05-14 9:42 GMT-06:00 cov...@ccs.covici.com: Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:05 AM, cov...@ccs.covici.com wrote: [snip] I am using version 037 and the command line was very simple As I told you in [1], the 037 version has problems with LVM; please also check the thread in [2], specially the post in [3]. Please try again with dracut-036-r4. Regards. [1] http://article.gmane.org/gmane.linux.gentoo.user/274340 [2] http://thread.gmane.org/gmane.linux.gentoo.user/274152 [3] http://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg145647.html (for some reason it doesn't appear in gmane). portage won't let me downgrade unless I downgrade systemd to 2.10 -- should I do that? I'm using dracut 037-r1 and got it to work. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
On Wed, May 14, 2014 at 10:42 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:05 AM, cov...@ccs.covici.com wrote: [snip] I am using version 037 and the command line was very simple As I told you in [1], the 037 version has problems with LVM; please also check the thread in [2], specially the post in [3]. Please try again with dracut-036-r4. Regards. [1] http://article.gmane.org/gmane.linux.gentoo.user/274340 [2] http://thread.gmane.org/gmane.linux.gentoo.user/274152 [3] http://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg145647.html (for some reason it doesn't appear in gmane). portage won't let me downgrade unless I downgrade systemd to 2.10 -- should I do that? You mean 210. I'm on 208; my LVM test system is basically stable. However Jc says he's got it to work with dracut 037-r1; perhaps try that version first? What I do know is that with dracut 037, me and Stefan ran intro troubles. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
2014-05-14 9:51 GMT-06:00 Canek Peláez Valdés can...@gmail.com: On Wed, May 14, 2014 at 10:42 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:05 AM, cov...@ccs.covici.com wrote: [snip] I am using version 037 and the command line was very simple As I told you in [1], the 037 version has problems with LVM; please also check the thread in [2], specially the post in [3]. Please try again with dracut-036-r4. Regards. [1] http://article.gmane.org/gmane.linux.gentoo.user/274340 [2] http://thread.gmane.org/gmane.linux.gentoo.user/274152 [3] http://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg145647.html (for some reason it doesn't appear in gmane). portage won't let me downgrade unless I downgrade systemd to 2.10 -- should I do that? You mean 210. I'm on 208; my LVM test system is basically stable. However Jc says he's got it to work with dracut 037-r1; perhaps try that version first? What I do know is that with dracut 037, me and Stefan ran intro troubles. I described it as a 'workaround' in my first post, aware of the fact that the behavior is not normal, but dracut acctually gives help finding this workaround, wich is telling the kernel what LVs to use at boot. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
On Wed, May 14, 2014 at 10:55 AM, Jc García jyo.gar...@gmail.com wrote: 2014-05-14 9:51 GMT-06:00 Canek Peláez Valdés can...@gmail.com: On Wed, May 14, 2014 at 10:42 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:05 AM, cov...@ccs.covici.com wrote: [snip] I am using version 037 and the command line was very simple As I told you in [1], the 037 version has problems with LVM; please also check the thread in [2], specially the post in [3]. Please try again with dracut-036-r4. Regards. [1] http://article.gmane.org/gmane.linux.gentoo.user/274340 [2] http://thread.gmane.org/gmane.linux.gentoo.user/274152 [3] http://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg145647.html (for some reason it doesn't appear in gmane). portage won't let me downgrade unless I downgrade systemd to 2.10 -- should I do that? You mean 210. I'm on 208; my LVM test system is basically stable. However Jc says he's got it to work with dracut 037-r1; perhaps try that version first? What I do know is that with dracut 037, me and Stefan ran intro troubles. I described it as a 'workaround' in my first post, aware of the fact that the behavior is not normal, but dracut acctually gives help finding this workaround, wich is telling the kernel what LVs to use at boot. Sorry, I forgot about that. I just tried my test LVM system with dracur-037-r1 and it didn't work. I have not tried yet your workaround, but there is obviously something fishy going on with the latest version of dracut; I would stay with 036-r4. I will try your workaround, although my system is a little more complicated, since it uses RAID and LUKS besides LVM. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:55 AM, Jc García jyo.gar...@gmail.com wrote: 2014-05-14 9:51 GMT-06:00 Canek Peláez Valdés can...@gmail.com: On Wed, May 14, 2014 at 10:42 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:05 AM, cov...@ccs.covici.com wrote: [snip] I am using version 037 and the command line was very simple As I told you in [1], the 037 version has problems with LVM; please also check the thread in [2], specially the post in [3]. Please try again with dracut-036-r4. Regards. [1] http://article.gmane.org/gmane.linux.gentoo.user/274340 [2] http://thread.gmane.org/gmane.linux.gentoo.user/274152 [3] http://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg145647.html (for some reason it doesn't appear in gmane). portage won't let me downgrade unless I downgrade systemd to 2.10 -- should I do that? You mean 210. I'm on 208; my LVM test system is basically stable. However Jc says he's got it to work with dracut 037-r1; perhaps try that version first? What I do know is that with dracut 037, me and Stefan ran intro troubles. I described it as a 'workaround' in my first post, aware of the fact that the behavior is not normal, but dracut acctually gives help finding this workaround, wich is telling the kernel what LVs to use at boot. Sorry, I forgot about that. I just tried my test LVM system with dracur-037-r1 and it didn't work. I have not tried yet your workaround, but there is obviously something fishy going on with the latest version of dracut; I would stay with 036-r4. I will try your workaround, although my system is a little more complicated, since it uses RAID and LUKS besides LVM. Here is what I got when I tried to emerge 036-r4 : Script started on Wed 14 May 2014 11:45:56 AM EDT These are the packages that would be merged, in order: Calculating dependencies . done! [ebuild UD ] sys-kernel/dracut-036-r4 [037] USE=systemd -debug (-selinux) 250 kB [blocks B ] =sys-apps/systemd-210 (=sys-apps/systemd-210 is blocking sys-kernel/dracut-036-r4) Total: 1 package (1 downgrade), Size of downloads: 250 kB Conflict: 1 block (1 unsatisfied) * Error: The above package list contains packages which cannot be * installed at the same time on the same system. (sys-apps/systemd-212-r4:0/2::gentoo, installed) pulled in by sys-apps/systemd:0/2= required by (gnome-base/gvfs-1.20.1:0/0::gnome, installed) =sys-apps/systemd-183:0/2= required by (net-misc/networkmanager-0.9.8.10:0/0::gnome, installed) sys-apps/systemd:0/2=[abi_x86_64(-)] required by (media-sound/pulseaudio-5.0-r1:0/0::gentoo, installed) =sys-apps/systemd-186:0= required by (sys-apps/accountsservice-0.6.37:0/0::gentoo, installed) =sys-apps/systemd-206 required by (sys-process/procps-3.3.9-r1:0/0::gentoo, installed) =sys-apps/systemd-183:0/2= required by (gnome-base/gnome-session-3.12.1:0/0::gnome, installed) =sys-apps/systemd-186:0/2= required by (sys-apps/accountsservice-0.6.37:0/0::gentoo, installed) =sys-apps/systemd-208 required by (sys-fs/udisks-2.1.3:2/2::gentoo, installed) sys-apps/systemd required by (net-wireless/bluez-5.18:0/3::gentoo, installed) =sys-apps/systemd-207 required by (sys-apps/gentoo-systemd-integration-4:0/0::gentoo, installed) =sys-apps/systemd-183:0= required by (net-misc/networkmanager-0.9.8.10:0/0::gnome, installed) =sys-apps/systemd-44:0= required by (x11-misc/colord-1.2.0:0/2::gentoo, installed) sys-apps/systemd:0= required by (sys-apps/dbus-1.8.2:0/0::gentoo, installed) =sys-apps/systemd-186:0=[pam] required by (gnome-base/gdm-3.12.1:0/0::gnome, installed) =sys-apps/systemd-31 required by (gnome-base/gnome-shell-3.12.1:0/0::gnome, installed) =sys-apps/systemd-44:0/2= required by (x11-misc/colord-1.2.0:0/2::gentoo, installed) sys-apps/systemd required by @selected sys-apps/systemd required by (gnome-base/gnome-settings-daemon-3.12.1:0/0::gnome, installed) =sys-apps/systemd-31 required by (gnome-base/gnome-control-center-3.12.1:2/2::gnome, installed) =sys-apps/systemd-208:0/2[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?,gudev,introspection?,static-libs(-)?] (=sys-apps/systemd-208:0/2[abi_x86_64(-),gudev,introspection]) required by (virtual/libgudev-208:0/0::gentoo, installed) sys-apps/systemd:0=[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?] (sys-apps/systemd:0=[abi_x86_64(-)]) required by (media-sound/pulseaudio-5.0-r1:0/0::gentoo, installed) =sys-apps/systemd-183:0= required by (gnome-base/gnome-session-3.12.1:0/0::gnome, installed) =sys-apps/systemd-208:0/2[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?,static-libs(-)?]
Re: [gentoo-user] problems getting systemd to work
2014-05-14 10:18 GMT-06:00 Canek Peláez Valdés can...@gmail.com: I will try your workaround, although my system is a little more complicated, since it uses RAID and LUKS besides LVM. That surely makes things more complicated, and I wouldn't expect something that simple to make the 'workaround', I would try to see what kernel command line dracut recommends, it might add something concerning RAID, the same it does with LVM.
Re: [gentoo-user] problems getting systemd to work
Am 14.05.2014 18:38, schrieb cov...@ccs.covici.com: Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:55 AM, Jc García jyo.gar...@gmail.com wrote: 2014-05-14 9:51 GMT-06:00 Canek Peláez Valdés can...@gmail.com: On Wed, May 14, 2014 at 10:42 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:05 AM, cov...@ccs.covici.com wrote: [snip] I am using version 037 and the command line was very simple As I told you in [1], the 037 version has problems with LVM; please also check the thread in [2], specially the post in [3]. Please try again with dracut-036-r4. Regards. [1] http://article.gmane.org/gmane.linux.gentoo.user/274340 [2] http://thread.gmane.org/gmane.linux.gentoo.user/274152 [3] http://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg145647.html (for some reason it doesn't appear in gmane). portage won't let me downgrade unless I downgrade systemd to 2.10 -- should I do that? You mean 210. I'm on 208; my LVM test system is basically stable. However Jc says he's got it to work with dracut 037-r1; perhaps try that version first? What I do know is that with dracut 037, me and Stefan ran intro troubles. I described it as a 'workaround' in my first post, aware of the fact that the behavior is not normal, but dracut acctually gives help finding this workaround, wich is telling the kernel what LVs to use at boot. Sorry, I forgot about that. I just tried my test LVM system with dracur-037-r1 and it didn't work. I have not tried yet your workaround, but there is obviously something fishy going on with the latest version of dracut; I would stay with 036-r4. I will try your workaround, although my system is a little more complicated, since it uses RAID and LUKS besides LVM. Here is what I got when I tried to emerge 036-r4 : Script started on Wed 14 May 2014 11:45:56 AM EDT These are the packages that would be merged, in order: Calculating dependencies . done! [ebuild UD ] sys-kernel/dracut-036-r4 [037] USE=systemd -debug (-selinux) 250 kB [blocks B ] =sys-apps/systemd-210 (=sys-apps/systemd-210 is blocking sys-kernel/dracut-036-r4) Total: 1 package (1 downgrade), Size of downloads: 250 kB Conflict: 1 block (1 unsatisfied) * Error: The above package list contains packages which cannot be * installed at the same time on the same system. (sys-apps/systemd-212-r4:0/2::gentoo, installed) pulled in by sys-apps/systemd:0/2= required by (gnome-base/gvfs-1.20.1:0/0::gnome, installed) =sys-apps/systemd-183:0/2= required by (net-misc/networkmanager-0.9.8.10:0/0::gnome, installed) sys-apps/systemd:0/2=[abi_x86_64(-)] required by (media-sound/pulseaudio-5.0-r1:0/0::gentoo, installed) =sys-apps/systemd-186:0= required by (sys-apps/accountsservice-0.6.37:0/0::gentoo, installed) =sys-apps/systemd-206 required by (sys-process/procps-3.3.9-r1:0/0::gentoo, installed) =sys-apps/systemd-183:0/2= required by (gnome-base/gnome-session-3.12.1:0/0::gnome, installed) =sys-apps/systemd-186:0/2= required by (sys-apps/accountsservice-0.6.37:0/0::gentoo, installed) =sys-apps/systemd-208 required by (sys-fs/udisks-2.1.3:2/2::gentoo, installed) sys-apps/systemd required by (net-wireless/bluez-5.18:0/3::gentoo, installed) =sys-apps/systemd-207 required by (sys-apps/gentoo-systemd-integration-4:0/0::gentoo, installed) =sys-apps/systemd-183:0= required by (net-misc/networkmanager-0.9.8.10:0/0::gnome, installed) =sys-apps/systemd-44:0= required by (x11-misc/colord-1.2.0:0/2::gentoo, installed) sys-apps/systemd:0= required by (sys-apps/dbus-1.8.2:0/0::gentoo, installed) =sys-apps/systemd-186:0=[pam] required by (gnome-base/gdm-3.12.1:0/0::gnome, installed) =sys-apps/systemd-31 required by (gnome-base/gnome-shell-3.12.1:0/0::gnome, installed) =sys-apps/systemd-44:0/2= required by (x11-misc/colord-1.2.0:0/2::gentoo, installed) sys-apps/systemd required by @selected sys-apps/systemd required by (gnome-base/gnome-settings-daemon-3.12.1:0/0::gnome, installed) =sys-apps/systemd-31 required by (gnome-base/gnome-control-center-3.12.1:2/2::gnome, installed) =sys-apps/systemd-208:0/2[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?,gudev,introspection?,static-libs(-)?] (=sys-apps/systemd-208:0/2[abi_x86_64(-),gudev,introspection]) required by (virtual/libgudev-208:0/0::gentoo, installed) sys-apps/systemd:0=[abi_x86_32(-)?,abi_x86_64(-)?,abi_x86_x32(-)?,abi_mips_n32(-)?,abi_mips_n64(-)?,abi_mips_o32(-)?] (sys-apps/systemd:0=[abi_x86_64(-)]) required by (media-sound/pulseaudio-5.0-r1:0/0::gentoo, installed) =sys-apps/systemd-183:0= required by (gnome-base/gnome-session-3.12.1:0/0::gnome, installed)
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:55 AM, Jc García jyo.gar...@gmail.com wrote: 2014-05-14 9:51 GMT-06:00 Canek Peláez Valdés can...@gmail.com: On Wed, May 14, 2014 at 10:42 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Wed, May 14, 2014 at 10:05 AM, cov...@ccs.covici.com wrote: [snip] I am using version 037 and the command line was very simple As I told you in [1], the 037 version has problems with LVM; please also check the thread in [2], specially the post in [3]. Please try again with dracut-036-r4. Regards. [1] http://article.gmane.org/gmane.linux.gentoo.user/274340 [2] http://thread.gmane.org/gmane.linux.gentoo.user/274152 [3] http://www.mail-archive.com/gentoo-user@lists.gentoo.org/msg145647.html (for some reason it doesn't appear in gmane). portage won't let me downgrade unless I downgrade systemd to 2.10 -- should I do that? You mean 210. I'm on 208; my LVM test system is basically stable. However Jc says he's got it to work with dracut 037-r1; perhaps try that version first? What I do know is that with dracut 037, me and Stefan ran intro troubles. I described it as a 'workaround' in my first post, aware of the fact that the behavior is not normal, but dracut acctually gives help finding this workaround, wich is telling the kernel what LVs to use at boot. Sorry, I forgot about that. I just tried my test LVM system with dracur-037-r1 and it didn't work. I have not tried yet your workaround, but there is obviously something fishy going on with the latest version of dracut; I would stay with 036-r4. I will try your workaround, although my system is a little more complicated, since it uses RAID and LUKS besides LVM. Well, the workaround sort of worked -- it went through the initrd -- I had debug in the kernel command line, but it did not stop for nothing! When it went to the real root, however it did not activate any of the lvm volumes I had except for what I specified in the kernel command line, causing things not to work well. Also, I noticed that if insisted on using the predictable network names, even though I have /etc/udev/rules.d/70-persistent-net.rules and /etc/udev/rules.d/80-name-slot.rules which work fine in openrc to give me back my eth* names. So all in all, it was a mess and took me to an emergency shell and that was the end of that. I did eventually activate some volumes by lvchange -aay, but obviously that would not work well. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On 14/05/14 19:40, cov...@ccs.covici.com wrote: Well, I rebooted under dracut, but it did not do the lvmscan and so the job trying to find the root file system timed out after 90 seconds. It took me to the emergency shell which I had specified, and I was able to do the lvm_scan and them magically root got mounted under sysroot, but I had no idea what to do next to maybe get things going. For what it's worth, I came across issues with my LVM/LUKS setup when I tried dracut, and in my searches came across [1] (fedoraproject.org). When dropped to the dracut emergency shell, the idea is to locate and flag your root volume in order to allow the boot process to continue. It varies between setups, but the idea is that you make your root volume accessible through whatever means (lvscan, cryptsetup, dm-crypt, whatever), then symlink it to /dev/root. Once that's done, you then `exit` the shell to allow the boot process to continue. [1] http://is.gd/bmzmNu Cheers. - -wraeth -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.22 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iF4EAREIAAYFAlNz+d4ACgkQXcRKerLZ91msnAD9GZ3oZ0rmQfeHx/yx6UlBn2U0 qkfzHR5uhvBnVK9Qi9IA/1VIVF3hYvYXUprWePQZcuLvewVzzW0xDVDFrLVgGoKo =FDs+ -END PGP SIGNATURE-
Re: [gentoo-user] problems getting systemd to work
Am 13.05.2014 05:46, schrieb cov...@ccs.covici.com: Hi. Well, even with use_fstab=yes, it does not put one, just /etc/fstab.empty of 0 length -- how can I fix? I didn't read the full thread yet ... but let me get that straight for me to understand the status: * You want to have / and everything on LVM volumes (LVs) and /boot (not /boot/efi, right? No EFI here?) as a separate partition. Are you able to boot this via openrc? * What is the status now with dracut? The mentioned options ... I don't have them in my config (although my setup is now completely different from yours ... anyway). * Do you have systemd now installed in the default location? * global USE-flag systemd set? and recompiled stuff ... ? * grub-2, right? Stefan
Re: [gentoo-user] problems getting systemd to work
Stefan G. Weichinger li...@xunil.at wrote: Am 13.05.2014 05:46, schrieb cov...@ccs.covici.com: Hi. Well, even with use_fstab=yes, it does not put one, just /etc/fstab.empty of 0 length -- how can I fix? I didn't read the full thread yet ... but let me get that straight for me to understand the status: * You want to have / and everything on LVM volumes (LVs) and /boot (not /boot/efi, right? No EFI here?) as a separate partition. Correct. Are you able to boot this via openrc? Yes openrc works fine. * What is the status now with dracut? The mentioned options ... I don't have them in my config (although my setup is now completely different from yours ... anyway). I emerged it and want to use it to boot with systemd. * Do you have systemd now installed in the default location? Yes. * global USE-flag systemd set? and recompiled stuff ... ? Yes. * grub-2, right? Nope, lilo. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Am 13.05.2014 14:29, schrieb cov...@ccs.covici.com: * What is the status now with dracut? The mentioned options ... I don't have them in my config (although my setup is now completely different from yours ... anyway). I emerged it and want to use it to boot with systemd. Did you configure dracut as Canek's examples show? Adding in the modules etc ? Could you build the initramfs with it, matching your kernel? Do you successfully use that initramfs with openrc then? * grub-2, right? Nope, lilo. serious? Wow ... I have no experience with that combination ... So you have something like: append = quiet init=/usr/lib/systemd/systemd in your lilo.conf and applied it ... ? Stefan
Re: [gentoo-user] problems getting systemd to work
2014-05-12 21:46 GMT-06:00 cov...@ccs.covici.com: Hi. Well, even with use_fstab=yes, it does not put one, just /etc/fstab.empty of 0 length -- how can I fix? That is strange, never had this problem, actually adding this made it work for me, I assume you actually used the ' yes ', might be important for the syntax, I uncompressed my ramdisk a few minutes ago to actually verify the fstab, and it is in fact included, might be other parameters missing in cofiguration, all uncomented parameters in my dracut.conf are these in case it might help you: logfile=/var/log/dracut.log fileloglvl=6 add_dracutmodules+=lvm bash dm kernel-modules systemd lvmconf=yes use_fstab=yes host_cmdline=yes kernel_cmdline=cmdline... And when generating I just simply run: dracut --kver 'kernel_version' As you see my configuration is pretty simple, I suggest you to try forcing the inclussion of fstab(extract, and re-compress the image), to verify if it solves your booting problem, it might be a dracut bug(tough it seems a relatively simple feature to be that prone to bugs).
Re: [gentoo-user] problems getting systemd to work
2014-05-13 7:02 GMT-06:00 Jc García jyo.gar...@gmail.com: 2014-05-12 21:46 GMT-06:00 cov...@ccs.covici.com: Hi. Well, even with use_fstab=yes, it does not put one, just /etc/fstab.empty of 0 length -- how can I fix? That is strange, never had this problem, actually adding this made it work for me, I assume you actually used the ' yes ', might be important for the syntax, I uncompressed my ramdisk a few minutes ago to actually verify the fstab, and it is in fact included, might be other parameters missing in cofiguration, all uncomented parameters in my dracut.conf are these in case it might help you: Sorry I did a wrong cd(I still sleepy, is early morning here), I have the same fstab.empty with nothing inside. so I'm lost here.
Re: [gentoo-user] problems getting systemd to work
Stefan G. Weichinger li...@xunil.at wrote: Am 13.05.2014 14:29, schrieb cov...@ccs.covici.com: * What is the status now with dracut? The mentioned options ... I don't have them in my config (although my setup is now completely different from yours ... anyway). I emerged it and want to use it to boot with systemd. Did you configure dracut as Canek's examples show? Adding in the modules etc ? Could you build the initramfs with it, matching your kernel? Do you successfully use that initramfs with openrc then? * grub-2, right? Nope, lilo. serious? Wow ... I have no experience with that combination ... So you have something like: append = quiet init=/usr/lib/systemd/systemd in your lilo.conf and applied it ... ? Nope, dracut does things different, so I am still working on the kernel command line. My question was about the /etc/fstab.empty problem. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
2014-05-13 7:18 GMT-06:00 cov...@ccs.covici.com: Nope, dracut does things different, so I am still working on the kernel command line. My question was about the /etc/fstab.empty problem. In the documentation it says, enabling this, uses the fstab instead of /proc/self/mountinfo, when generating the image, it does not say it includes it, so I think that is working properly. Have you tried booting manually from a grub command line(I read you use lilo, I don't know if it has this feature, but the idea is manually write your kernel command line before boot ), this has helped me several times when finding booting problems, you can use anything with grub in it (livecd, usb, etc...) so you don't have to install it, in case you are not familiar with grub.
Re: [gentoo-user] problems getting systemd to work
Jc García jyo.gar...@gmail.com wrote: 2014-05-12 21:46 GMT-06:00 cov...@ccs.covici.com: Hi. Well, even with use_fstab=yes, it does not put one, just /etc/fstab.empty of 0 length -- how can I fix? That is strange, never had this problem, actually adding this made it work for me, I assume you actually used the ' yes ', might be important for the syntax, I uncompressed my ramdisk a few minutes ago to actually verify the fstab, and it is in fact included, might be other parameters missing in cofiguration, all uncomented parameters in my dracut.conf are these in case it might help you: logfile=/var/log/dracut.log fileloglvl=6 add_dracutmodules+=lvm bash dm kernel-modules systemd lvmconf=yes use_fstab=yes host_cmdline=yes kernel_cmdline=cmdline... And when generating I just simply run: dracut --kver 'kernel_version' As you see my configuration is pretty simple, I suggest you to try forcing the inclussion of fstab(extract, and re-compress the image), to verify if it solves your booting problem, it might be a dracut bug(tough it seems a relatively simple feature to be that prone to bugs). hmmm, do we really need add_dracutmodules+=lvm bash dm kernel-modules systemd I had the fstab with and without the quotes, but no difference. Maybe I need to include the thing individually? -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Jc García jyo.gar...@gmail.com wrote: 2014-05-13 7:18 GMT-06:00 cov...@ccs.covici.com: Nope, dracut does things different, so I am still working on the kernel command line. My question was about the /etc/fstab.empty problem. In the documentation it says, enabling this, uses the fstab instead of /proc/self/mountinfo, when generating the image, it does not say it includes it, so I think that is working properly. Have you tried booting manually from a grub command line(I read you use lilo, I don't know if it has this feature, but the idea is manually write your kernel command line before boot ), this has helped me several times when finding booting problems, you can use anything with grub in it (livecd, usb, etc...) so you don't have to install it, in case you are not familiar with grub. Well, if this is OK, then I can proceed, I thought that the systemd generator needed the actual /etc/fstab file to generate the correct mount events, so I was concerned that this would not happen, but maybe the initrd does not need to do that so much. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
2014-05-13 7:43 GMT-06:00 cov...@ccs.covici.com: Jc García jyo.gar...@gmail.com wrote: 2014-05-12 21:46 GMT-06:00 cov...@ccs.covici.com: Hi. Well, even with use_fstab=yes, it does not put one, just /etc/fstab.empty of 0 length -- how can I fix? That is strange, never had this problem, actually adding this made it work for me, I assume you actually used the ' yes ', might be important for the syntax, I uncompressed my ramdisk a few minutes ago to actually verify the fstab, and it is in fact included, might be other parameters missing in cofiguration, all uncomented parameters in my dracut.conf are these in case it might help you: logfile=/var/log/dracut.log fileloglvl=6 add_dracutmodules+=lvm bash dm kernel-modules systemd lvmconf=yes use_fstab=yes host_cmdline=yes kernel_cmdline=cmdline... And when generating I just simply run: dracut --kver 'kernel_version' As you see my configuration is pretty simple, I suggest you to try forcing the inclussion of fstab(extract, and re-compress the image), to verify if it solves your booting problem, it might be a dracut bug(tough it seems a relatively simple feature to be that prone to bugs). hmmm, do we really need add_dracutmodules+=lvm bash dm kernel-modules systemd It's not necessary indeed, the Idea was you to compare this with yours, since its working here, for example bash is actually a thing of mine wanting a nice shell even in the initrd. you include what fits your needs. I had the fstab with and without the quotes, but no difference. Maybe I need to include the thing individually? -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
[gentoo-user] problems getting systemd to work
Hi. I have been trying to get systemd to boot, but I have run into several problems and need some help. I am using everything but /boot as lvm's, with a separate user partition. I had to copy systemd to /sbin because the initrd looks for the realinit too soon, but that is maybe another matter. I had set confirm_spawn=y in the kernel command line, but it only waits a short time and then says assuming positive response and tries to continue -- how can I get it to wait for me? Also, even so, it died on mounting of my lvms, saying there was some kind of timeout and came to a complete halt (maybe it was a shell, but no prompt) after all those failed, so I could do nothing much. Openrc works fine, but I was trying to get gnome to work, so I was trying to use systemd. It saved no logs (none I can find), but then again /var was not mounted. Any help with this would be appreciated. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
On Mon, May 12, 2014 at 2:22 AM, cov...@ccs.covici.com wrote: Hi. I have been trying to get systemd to boot, but I have run into several problems and need some help. I am using everything but /boot as lvm's, with a separate user partition. I had to copy systemd to /sbin because the initrd looks for the realinit too soon, but that is maybe another matter. Moving systemd to /sbin sounds like it's not going to work. Run readelf -d /usr/lib/systemd/systemd; all the NEEDED libraries on /usr/lib should be available to the binary at the time it's being executed. How did you get your initramfs? dracut? genkernel? Roll your own? I had set confirm_spawn=y in the kernel command line, but it only waits a short time and then says assuming positive response and tries to continue -- how can I get it to wait for me? Also, even so, it died on mounting of my lvms, saying there was some kind of timeout and came to a complete halt (maybe it was a shell, but no prompt) after all those failed, so I could do nothing much. Openrc works fine, but I was trying to get gnome to work, so I was trying to use systemd. It saved no logs (none I can find), but then again /var was not mounted. Any help with this would be appreciated. I use dracut for my initramfs; I would recommend you to try it. However, last time I tried to use it with LVM (a few days ago), the last version (037) failed, but 036-r4 worked perfectly. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 2:22 AM, cov...@ccs.covici.com wrote: Hi. I have been trying to get systemd to boot, but I have run into several problems and need some help. I am using everything but /boot as lvm's, with a separate user partition. I had to copy systemd to /sbin because the initrd looks for the realinit too soon, but that is maybe another matter. Moving systemd to /sbin sounds like it's not going to work. Run readelf -d /usr/lib/systemd/systemd; all the NEEDED libraries on /usr/lib should be available to the binary at the time it's being executed. How can I do this, genkernel looks for its init before it mounts /usr and genkernel-next will not mount the separate /usr at all. My latest initrd is from the very latest genkernel. But how to get a complete history of systemd actions in the order that they are done, I thought the confirm_spawn would do this for me -- at least for my initial debugging. How did you get your initramfs? dracut? genkernel? Roll your own? I had set confirm_spawn=y in the kernel command line, but it only waits a short time and then says assuming positive response and tries to continue -- how can I get it to wait for me? Also, even so, it died on mounting of my lvms, saying there was some kind of timeout and came to a complete halt (maybe it was a shell, but no prompt) after all those failed, so I could do nothing much. Openrc works fine, but I was trying to get gnome to work, so I was trying to use systemd. It saved no logs (none I can find), but then again /var was not mounted. Any help with this would be appreciated. I use dracut for my initramfs; I would recommend you to try it. However, last time I tried to use it with LVM (a few days ago), the last version (037) failed, but 036-r4 worked perfectly. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
2014-05-12 4:15 GMT-06:00 cov...@ccs.covici.com: How can I do this, genkernel looks for its init before it mounts /usr and genkernel-next will not mount the separate /usr at all. My latest initrd is from the very latest genkernel. But how to get a complete history of systemd actions in the order that they are done, I thought the confirm_spawn would do this for me -- at least for my initial debugging. I have had this trouble too, and a very similar setup than you, and after a few workarounds I got to boot with a genkernel and a dracut generated initramfs, so it can be done both ways, but i would recommend dracut, since is more straight forward in practice, and you can setup once and then just generate initramfs that surely will work. The most important part is your kernel boot comand line, giving instructions so your system specific lvm volumes (root, usr and var if separated). mine looks like this rd.lvm rd.lvm.vg=gentoovg rd.lvm.lv=gentoovg/root rd.lvm.lv=gentoovg/usr root=/dev/mapper/gentoovg-root ccinit=/usr/lib/systemd/systemd quiet A little too long in my opinion, but works, and the rd.lvm.lv parts result redundant if rd.lvm.vg is already set, i think, it worked when I tested, but I kept the redundancy just in case. this can be setup in sevaral ways, directly when compiling the kernel, using dracut the config file, or the bootloader, I used dracut since I wanted to centralize the boot process configuration as much as possible. Also be sure that the lvm binaries are included in the initramfs, if you will be using dracut you would need to add to /etc/dracut.conf: use_fstab=yes host_cmdline=yes kernel_cmdline=your_cmd_line
Re: [gentoo-user] problems getting systemd to work
On Mon, May 12, 2014 at 5:15 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 2:22 AM, cov...@ccs.covici.com wrote: Hi. I have been trying to get systemd to boot, but I have run into several problems and need some help. I am using everything but /boot as lvm's, with a separate user partition. I had to copy systemd to /sbin because the initrd looks for the realinit too soon, but that is maybe another matter. Moving systemd to /sbin sounds like it's not going to work. Run readelf -d /usr/lib/systemd/systemd; all the NEEDED libraries on /usr/lib should be available to the binary at the time it's being executed. How can I do this, genkernel looks for its init before it mounts /usr and genkernel-next will not mount the separate /usr at all. My latest initrd is from the very latest genkernel. With genkernel, I don't know; I never used it. On the other hand, dracut is designed to work with systemd; if you use the systemd USE flag and the systemd module, it even uses systemd *inside* the initramfs. But how to get a complete history of systemd actions in the order that they are done, I thought the confirm_spawn would do this for me -- at least for my initial debugging. The problem obviously is not in systemd, but in the integration of genkernel+systemd. I repeat, I never used genkernel, so I don't know what you can do. That being said, get a complete history of systemd actions in the order that they are done will not tell you much: systemd uses heavy parallelization, so in some runs the order in which actions are performed will be different from others. The problem is that if systemd is installed into /usr/lib (which is Gentoo's case), then /usr should be mounted before systemd starts. That's responsibility of the initramfs, not of systemd, and the solution lies in the initramfs, not in systemd. My only possible recommendation would be for you to try dracut. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Jc García jyo.gar...@gmail.com wrote: 2014-05-12 4:15 GMT-06:00 cov...@ccs.covici.com: How can I do this, genkernel looks for its init before it mounts /usr and genkernel-next will not mount the separate /usr at all. My latest initrd is from the very latest genkernel. But how to get a complete history of systemd actions in the order that they are done, I thought the confirm_spawn would do this for me -- at least for my initial debugging. I have had this trouble too, and a very similar setup than you, and after a few workarounds I got to boot with a genkernel and a dracut generated initramfs, so it can be done both ways, but i would recommend dracut, since is more straight forward in practice, and you can setup once and then just generate initramfs that surely will work. The most important part is your kernel boot comand line, giving instructions so your system specific lvm volumes (root, usr and var if separated). mine looks like this rd.lvm rd.lvm.vg=gentoovg rd.lvm.lv=gentoovg/root rd.lvm.lv=gentoovg/usr root=/dev/mapper/gentoovg-root ccinit=/usr/lib/systemd/systemd quiet A little too long in my opinion, but works, and the rd.lvm.lv parts result redundant if rd.lvm.vg is already set, i think, it worked when I tested, but I kept the redundancy just in case. this can be setup in sevaral ways, directly when compiling the kernel, using dracut the config file, or the bootloader, I used dracut since I wanted to centralize the boot process configuration as much as possible. Also be sure that the lvm binaries are included in the initramfs, if you will be using dracut you would need to add to /etc/dracut.conf: use_fstab=yes host_cmdline=yes kernel_cmdline=your_cmd_line My kernel command line is like this: init=/linuxrc ramdisk=8192 real_root=/dev/mapper/linux--files-64--root udev video=uvesafb:1280x1024 speakup.synth=spkout vmalloc=256M dolvm rootfstype=ext4 real_init=/sbin/systemd systemd.confirm_spawn=yes I thought the dolvm would take care of all lvm related stuff, I don't understand the rd.lvm parts at all, I have never seen such. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 5:15 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 2:22 AM, cov...@ccs.covici.com wrote: Hi. I have been trying to get systemd to boot, but I have run into several problems and need some help. I am using everything but /boot as lvm's, with a separate user partition. I had to copy systemd to /sbin because the initrd looks for the realinit too soon, but that is maybe another matter. Moving systemd to /sbin sounds like it's not going to work. Run readelf -d /usr/lib/systemd/systemd; all the NEEDED libraries on /usr/lib should be available to the binary at the time it's being executed. How can I do this, genkernel looks for its init before it mounts /usr and genkernel-next will not mount the separate /usr at all. My latest initrd is from the very latest genkernel. With genkernel, I don't know; I never used it. On the other hand, dracut is designed to work with systemd; if you use the systemd USE flag and the systemd module, it even uses systemd *inside* the initramfs. But how to get a complete history of systemd actions in the order that they are done, I thought the confirm_spawn would do this for me -- at least for my initial debugging. The problem obviously is not in systemd, but in the integration of genkernel+systemd. I repeat, I never used genkernel, so I don't know what you can do. That being said, get a complete history of systemd actions in the order that they are done will not tell you much: systemd uses heavy parallelization, so in some runs the order in which actions are performed will be different from others. The problem is that if systemd is installed into /usr/lib (which is Gentoo's case), then /usr should be mounted before systemd starts. That's responsibility of the initramfs, not of systemd, and the solution lies in the initramfs, not in systemd. My only possible recommendation would be for you to try dracut. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México OK, I will try dracut, but I still want to know what systemd is doing, what processes its spawning, etc. -- how can I find this out -- I thought to use the confirm_spawn, but it times out and keeps going, what can I do instead? Thanks people for all your responses, this is a great list. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
On Mon, May 12, 2014 at 11:31 AM, cov...@ccs.covici.com wrote: [snip] OK, I will try dracut, I hope it works with dracut. This is my kernel command line and RAID/LVM related stuff from GRUB2: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash GRUB_PRELOAD_MODULES=lvm mdraid1x And this is my dracut.conf (minus comments): add_dracutmodules+=crypt lvm mdraid systemd add_drivers+=autofs4 ipv6 dm-crypt aes sha256 fscks=umount mount /sbin/fsck* e2fsck That's it. I didn't touched anything else to make dracut+systemd work with LVM and RAID (and LUKS, but that doesn't matter). Also, dracut comes with extensive and very clear documentation; check the man pages included. but I still want to know what systemd is doing, what processes its spawning, etc. -- how can I find this out -- I thought to use the confirm_spawn, but it times out and keeps going, what can I do instead? You can use bootchart: man 1 systemd-bootchart It will produce a chart with all the processes, and how long it takes for every one of them. But remember, the order depends on which one finishes before, and that can change from boot to boot. Thanks people for all your responses, this is a great list. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:31 AM, cov...@ccs.covici.com wrote: [snip] OK, I will try dracut, I hope it works with dracut. This is my kernel command line and RAID/LVM related stuff from GRUB2: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash GRUB_PRELOAD_MODULES=lvm mdraid1x And this is my dracut.conf (minus comments): add_dracutmodules+=crypt lvm mdraid systemd add_drivers+=autofs4 ipv6 dm-crypt aes sha256 fscks=umount mount /sbin/fsck* e2fsck That's it. I didn't touched anything else to make dracut+systemd work with LVM and RAID (and LUKS, but that doesn't matter). Also, dracut comes with extensive and very clear documentation; check the man pages included. but I still want to know what systemd is doing, what processes its spawning, etc. -- how can I find this out -- I thought to use the confirm_spawn, but it times out and keeps going, what can I do instead? You can use bootchart: man 1 systemd-bootchart It will produce a chart with all the processes, and how long it takes for every one of them. But remember, the order depends on which one finishes before, and that can change from boot to boot. Thanks people for all your responses, this is a great list. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Well, since I am unable to see, the graph would not do me any good, any way to get it in text form? What I want to see (and I know the order may change) is which starts first and so on, to make sure targets, etc. do what I want them to do. -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
On Mon, May 12, 2014 at 11:52 AM, cov...@ccs.covici.com wrote: Canek Peláez Valdés can...@gmail.com wrote: On Mon, May 12, 2014 at 11:31 AM, cov...@ccs.covici.com wrote: [snip] OK, I will try dracut, I hope it works with dracut. This is my kernel command line and RAID/LVM related stuff from GRUB2: GRUB_CMDLINE_LINUX=init=/usr/lib/systemd/systemd quiet nosplash GRUB_PRELOAD_MODULES=lvm mdraid1x And this is my dracut.conf (minus comments): add_dracutmodules+=crypt lvm mdraid systemd add_drivers+=autofs4 ipv6 dm-crypt aes sha256 fscks=umount mount /sbin/fsck* e2fsck That's it. I didn't touched anything else to make dracut+systemd work with LVM and RAID (and LUKS, but that doesn't matter). Also, dracut comes with extensive and very clear documentation; check the man pages included. but I still want to know what systemd is doing, what processes its spawning, etc. -- how can I find this out -- I thought to use the confirm_spawn, but it times out and keeps going, what can I do instead? You can use bootchart: man 1 systemd-bootchart It will produce a chart with all the processes, and how long it takes for every one of them. But remember, the order depends on which one finishes before, and that can change from boot to boot. Thanks people for all your responses, this is a great list. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Well, since I am unable to see, the graph would not do me any good, any way to get it in text form? What I want to see (and I know the order may change) is which starts first and so on, to make sure targets, etc. do what I want them to do. Try adding this to your kernel command line: systemd.log_target=console systemd.log_level=debug. It will add a lot of output, including what is being executed. Everything is documented in the man page: man 1 systemd. Regards. -- Canek Peláez Valdés Profesor de asignatura, Facultad de Ciencias Universidad Nacional Autónoma de México
Re: [gentoo-user] problems getting systemd to work
2014-05-12 10:22 GMT-06:00 cov...@ccs.covici.com: My kernel command line is like this: init=/linuxrc ramdisk=8192 real_root=/dev/mapper/linux--files-64--root udev video=uvesafb:1280x1024 speakup.synth=spkout vmalloc=256M dolvm rootfstype=ext4 real_init=/sbin/systemd systemd.confirm_spawn=yes I thought the dolvm would take care of all lvm related stuff, I don't understand the rd.lvm parts at all, I have never seen such. I tried several times with only dolvm, but that didn't work for me, I found the documentation for rd.lvm 'man 7 dracut.kernel', in the LVM section.
Re: [gentoo-user] problems getting systemd to work
Jc García jyo.gar...@gmail.com wrote: 2014-05-12 10:22 GMT-06:00 cov...@ccs.covici.com: My kernel command line is like this: init=/linuxrc ramdisk=8192 real_root=/dev/mapper/linux--files-64--root udev video=uvesafb:1280x1024 speakup.synth=spkout vmalloc=256M dolvm rootfstype=ext4 real_init=/sbin/systemd systemd.confirm_spawn=yes I thought the dolvm would take care of all lvm related stuff, I don't understand the rd.lvm parts at all, I have never seen such. I tried several times with only dolvm, but that didn't work for me, I found the documentation for rd.lvm 'man 7 dracut.kernel', in the LVM section. Ahhh, these are dracut specific, that is why I had never heard of them before. So I am off to read the dracut docs before I change anything! -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
Re: [gentoo-user] problems getting systemd to work
Jc García jyo.gar...@gmail.com wrote: 2014-05-12 4:15 GMT-06:00 cov...@ccs.covici.com: How can I do this, genkernel looks for its init before it mounts /usr and genkernel-next will not mount the separate /usr at all. My latest initrd is from the very latest genkernel. But how to get a complete history of systemd actions in the order that they are done, I thought the confirm_spawn would do this for me -- at least for my initial debugging. I have had this trouble too, and a very similar setup than you, and after a few workarounds I got to boot with a genkernel and a dracut generated initramfs, so it can be done both ways, but i would recommend dracut, since is more straight forward in practice, and you can setup once and then just generate initramfs that surely will work. The most important part is your kernel boot comand line, giving instructions so your system specific lvm volumes (root, usr and var if separated). mine looks like this rd.lvm rd.lvm.vg=gentoovg rd.lvm.lv=gentoovg/root rd.lvm.lv=gentoovg/usr root=/dev/mapper/gentoovg-root ccinit=/usr/lib/systemd/systemd quiet A little too long in my opinion, but works, and the rd.lvm.lv parts result redundant if rd.lvm.vg is already set, i think, it worked when I tested, but I kept the redundancy just in case. this can be setup in sevaral ways, directly when compiling the kernel, using dracut the config file, or the bootloader, I used dracut since I wanted to centralize the boot process configuration as much as possible. Also be sure that the lvm binaries are included in the initramfs, if you will be using dracut you would need to add to /etc/dracut.conf: use_fstab=yes host_cmdline=yes kernel_cmdline=your_cmd_line Hi. Well, even with use_fstab=yes, it does not put one, just /etc/fstab.empty of 0 length -- how can I fix? -- Your life is like a penny. You're going to lose it. The question is: How do you spend it? John Covici cov...@ccs.covici.com
