Systemd start that won't stop

classic Classic list List threaded Threaded
18 messages Options
Reply | Threaded
Open this post in threaded view
|

Systemd start that won't stop

Curt Howland
Hi. New Buster install.

While X11 is working just fine, the console is unusable, due to a
startup message that never stops:

[ ***] A start job is running for Hold until boot process finishes (2h
54m 38s / no limit)

This goes on until the laptop is shut down.

How do I find the errant process? The start-up messages fly by, but I
don't see any failures, or any indication of what process it is that
won't stop starting.

Curt-

Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Kenneth Parker-2


On Fri, Aug 9, 2019, 11:30 AM Curt Howland <[hidden email]> wrote:
Hi. New Buster install.

While X11 is working just fine, the console is unusable, due to a
startup message that never stops:

[ ***] A start job is running for Hold until boot process finishes (2h
54m 38s / no limit)

This goes on until the laptop is shut down.

How do I find the errant process? The start-up messages fly by, but I
don't see any failures, or any indication of what process it is that
won't stop starting.

Are any of your File Systems out of space?   

Good luck, by the way. 

Kenneth Parker 
Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Étienne Mollier
Kenneth Parker, on 2019-08-09:

> On Fri, Aug 9, 2019, 11:30 AM Curt Howland <[hidden email]> wrote:
> > Hi. New Buster install.
> >
> > While X11 is working just fine, the console is unusable, due to a
> > startup message that never stops:
> >
> > [ ***] A start job is running for Hold until boot process finishes (2h 54m 38s / no limit)
> >
> > This goes on until the laptop is shut down.
> >
> > How do I find the errant process? The start-up messages fly by, but I
> > don't see any failures, or any indication of what process it is that
> > won't stop starting.
Good day

You may be able to get a few information out of the following,
maybe see what's on hold:

        $ sudo systemctl list-units

But it is the first time I hear about this specific "start job"
"running for Hold" behaviour.  I'm not sure about what to look
at.  A blind shot at the following might bring some additional
information too:

        $ sudo journalctl -xe

In case you were encountering some hardware issue (but I doubt
so for the moment), perhaps a look at the following could be
welcome:

        $ sudo dmesg

And of course, from Kenneth:
> Are any of your File Systems out of space?

Kind regards,
--
Étienne Mollier <[hidden email]>
               5ab1 4edf 63bb ccff 8b54 2fa9 59da 56fe fff3 882d



signature.asc (673 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Curt Howland
In reply to this post by Curt Howland
On 8/9/19, Curt Howland <[hidden email]> wrote:
> Hi. New Buster install.
> [ ***] A start job is running for Hold until boot process finishes (2h
> 54m 38s / no limit)

Those asterisks are also red, and moving left to right, the same as
seen during shutdown when something won't politely die.

Only two things look interesting to me under "list units::
# systemctl list-units
[hidden email]                                loaded inactive
dead      start Getty on tty1
plymouth-quit-wait.service                        loaded activating
start     start Hold until boot process f

Since tty1 would be the console, it makes sense that it would be dead.
I have no idea what a "plymouth" is.

dmesg showed nothing that looked wrong.

There is no problem with partition space.

Curt-

Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Étienne Mollier
Curt Howland, on 2019-08-09:
> On 8/9/19, Curt Howland <[hidden email]> wrote:
> > Hi. New Buster install.
> > [ ***] A start job is running for Hold until boot process finishes (2h
> > 54m 38s / no limit)
>
> Those asterisks are also red, and moving left to right, the same as
> seen during shutdown when something won't politely die.

Just like the scanner in "Knight Rider", I see the picture.

> Only two things look interesting to me under "list units::
> # systemctl list-units
> [hidden email]                                loaded inactive
> dead      start Getty on tty1
> plymouth-quit-wait.service                        loaded activating
> start     start Hold until boot process f

This Plymouth thing is ringing me some bells, I'm almost certain
it has already been mentioned not very long ago, but I'm not
sure where or when.

What is the output of the following command?

        $ sudo systemctl status plymouth-quit-wait

> Since tty1 would be the console, it makes sense that it would be dead.
> I have no idea what a "plymouth" is.

It is supposed to be some sort of graphical boot screen.  I just
tried it on my Buster machine, but the only notable difference
that appeared. was that now I have asterisks when typing in the
passphrase of my ciphered drive.  I'm not even certain this is
related.

> dmesg showed nothing that looked wrong.
>
> There is no problem with partition space.

Good, I believe we can rule out the hardware side for now.

Kind regards,
--
Étienne Mollier <[hidden email]>
               5ab1 4edf 63bb ccff 8b54 2fa9 59da 56fe fff3 882d



signature.asc (673 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Sven Joachim
In reply to this post by Curt Howland
On 2019-08-09 13:53 -0400, Curt Howland wrote:

> On 8/9/19, Curt Howland <[hidden email]> wrote:
>> Hi. New Buster install.
>> [ ***] A start job is running for Hold until boot process finishes (2h
>> 54m 38s / no limit)
>
> Those asterisks are also red, and moving left to right, the same as
> seen during shutdown when something won't politely die.
>
> Only two things look interesting to me under "list units::
> # systemctl list-units
> [hidden email]                                loaded inactive
> dead      start Getty on tty1
> plymouth-quit-wait.service                        loaded activating
> start     start Hold until boot process f
>
> Since tty1 would be the console, it makes sense that it would be dead.
> I have no idea what a "plymouth" is.

So you probably have no idea why it is installed in the first place?
Just curious, because there is a huge rise for plymouth in the number of
popcon submissions[1] since the Buster release, and I wonder why.

Cheers,
       Sven


1. https://qa.debian.org/popcon.php?package=plymouth

Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Étienne Mollier
On 09/08/2019 21.15, Sven Joachim wrote:

> On 2019-08-09 13:53 -0400, Curt Howland wrote:
>
>> On 8/9/19, Curt Howland <[hidden email]> wrote:
>>> Hi. New Buster install.
>>> [ ***] A start job is running for Hold until boot process finishes (2h
>>> 54m 38s / no limit)
>> Those asterisks are also red, and moving left to right, the same as
>> seen during shutdown when something won't politely die.
>>
>> Only two things look interesting to me under "list units::
>> # systemctl list-units
>> [hidden email]                                loaded inactive
>> dead      start Getty on tty1
>> plymouth-quit-wait.service                        loaded activating
>> start     start Hold until boot process f
>>
>> Since tty1 would be the console, it makes sense that it would be dead.
>> I have no idea what a "plymouth" is.
> So you probably have no idea why it is installed in the first place?
> Just curious, because there is a huge rise for plymouth in the number of
> popcon submissions[1] since the Buster release, and I wonder why.
>
> Cheers,
>        Sven
Good day Sven,

It seems to be pulled by "task-gnome-desktop", either by
dependency, or mere recommendation, I don't know.  From the
quick test I did, it appeared in the list of packages to be
installed:

        $ sudo apt install task-gnome-desktop

This looks like a component of Debian default installation from
now on; unless I misunderstood the role of task-* packages.
Hence the sudden rise in popularity.

Kind regards,
--
Étienne Mollier <[hidden email]>
               5AB1 4EDF 63BB CCFF 8B54 2FA9 59DA 56FE FFF3 882D


signature.asc (673 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Greg Wooledge
On Fri, Aug 09, 2019 at 09:48:41PM +0200, Étienne Mollier wrote:

> It [plymouth] seems to be pulled by "task-gnome-desktop", either by
> dependency, or mere recommendation, I don't know.  From the
> quick test I did, it appeared in the list of packages to be
> installed:
>
> $ sudo apt install task-gnome-desktop
>
> This looks like a component of Debian default installation from
> now on; unless I misunderstood the role of task-* packages.
> Hence the sudden rise in popularity.

wooledg:~$ apt-cache show task-gnome-desktop | egrep '(Depends|Recommends|Suggets):'
Depends: tasksel (= 3.53), task-desktop, gnome-core
Recommends: gnome, libreoffice-gnome, libreoffice-writer, libreoffice-calc, libreoffice-impress, libreoffice-help-en-us, mythes-en-us, hunspell-en-us, hyphen-en-us, network-manager-gnome

wooledg:~$ apt-cache show task-desktop | egrep '(Depends|Recommends|Suggets):'  Depends: tasksel (= 3.53), xorg, xserver-xorg-video-all, xserver-xorg-input-all, desktop-base
Recommends: task-gnome-desktop | task-xfce-desktop | task-kde-desktop | task-lxde-desktop | task-cinnamon-desktop | task-mate-desktop | task-lxqt-desktop, xdg-utils, avahi-daemon, libnss-mdns, anacron, eject, iw, alsa-utils, libu2f-udev, sudo, firefox | firefox-esr

wooledg:~$ apt-cache show desktop-base | egrep '(Depends|Recommends|Suggets):'
Depends: librsvg2-common, fonts-quicksand
Recommends: plymouth-label

wooledg:~$ apt-cache show plymouth-label | egrep '(Depends|Recommends|Suggets):'Depends: plymouth (= 0.9.4-1.1), libc6 (>= 2.4), libcairo2 (>= 1.14.0), libglib2.0-0 (>= 2.12.0), libpango-1.0-0 (>= 1.14.0), libpangocairo-1.0-0 (>= 1.14.0), libplymouth4 (>= 0.9.3)


... so, probably not *just* GNOME, then.

Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Sven Joachim
In reply to this post by Étienne Mollier
On 2019-08-09 21:48 +0200, Étienne Mollier wrote:

> It seems to be pulled by "task-gnome-desktop", either by
> dependency, or mere recommendation, I don't know.  From the
> quick test I did, it appeared in the list of packages to be
> installed:
>
> $ sudo apt install task-gnome-desktop
>
> This looks like a component of Debian default installation from
> now on; unless I misunderstood the role of task-* packages.
> Hence the sudden rise in popularity.

Thanks.  As Greg has already pointed out, it is actually desktop-base
which recommends plymouth-label.  In the changelog I found the following
explanation:

,----
| desktop-base (10.0.1) unstable; urgency=medium
| [...]
|   [ Jonathan Carter ]
|   * Add plymouth-label as recommends (Closes: #927077)
|
|  -- Jonathan Carter <[hidden email]>  Thu, 04 Apr 2019 17:27:08 +0200
`----

Reading the log of bug #927077 I have my doubts that this was a good
idea, but I'm a layman when it comes to modern desktops.

Cheers,
       Sven

Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Étienne Mollier
In reply to this post by Greg Wooledge
Greg Wooledge, on 2019-08-09:

> On Fri, Aug 09, 2019 at 09:48:41PM +0200, Étienne Mollier wrote:
> > It [plymouth] seems to be pulled by "task-gnome-desktop", either by
> > dependency, or mere recommendation, I don't know.  From the
> > quick test I did, it appeared in the list of packages to be
> > installed:
> >
> > $ sudo apt install task-gnome-desktop
> >
> > This looks like a component of Debian default installation from
> > now on; unless I misunderstood the role of task-* packages.
> > Hence the sudden rise in popularity.
>
> wooledg:~$ apt-cache show task-gnome-desktop | egrep '(Depends|Recommends|Suggets):'
> Depends: tasksel (= 3.53), task-desktop, gnome-core
> Recommends: gnome, libreoffice-gnome, libreoffice-writer, libreoffice-calc, libreoffice-impress, libreoffice-help-en-us, mythes-en-us, hunspell-en-us, hyphen-en-us, network-manager-gnome
>
> wooledg:~$ apt-cache show task-desktop | egrep '(Depends|Recommends|Suggets):'
> Depends: tasksel (= 3.53), xorg, xserver-xorg-video-all, xserver-xorg-input-all, desktop-base
> Recommends: task-gnome-desktop | task-xfce-desktop | task-kde-desktop | task-lxde-desktop | task-cinnamon-desktop | task-mate-desktop | task-lxqt-desktop, xdg-utils, avahi-daemon, libnss-mdns, anacron, eject, iw, alsa-utils, libu2f-udev, sudo, firefox | firefox-esr
>
> wooledg:~$ apt-cache show desktop-base | egrep '(Depends|Recommends|Suggets):'
> Depends: librsvg2-common, fonts-quicksand
> Recommends: plymouth-label
>
> wooledg:~$ apt-cache show plymouth-label | egrep '(Depends|Recommends|Suggets):'
> Depends: plymouth (= 0.9.4-1.1), libc6 (>= 2.4), libcairo2 (>= 1.14.0), libglib2.0-0 (>= 2.12.0), libpango-1.0-0 (>= 1.14.0), libpangocairo-1.0-0 (>= 1.14.0), libplymouth4 (>= 0.9.3)
>
>
> ... so, probably not *just* GNOME, then.
Whoah!  Indeed!  The presence of task-desktop in the chain
suggests that this would be the case for any of the possible
desktop available.  It seems that we will have to get used to
it.

Thanks for having spotted this!

Personally, I missed it because I stick to the "Base utilities"
and "SSH server" options when setting up a machine for myself
using Debian Installer ISO.

Kind regards,
--
Étienne Mollier <[hidden email]>
               5ab1 4edf 63bb ccff 8b54 2fa9 59da 56fe fff3 882d



signature.asc (673 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Felix Miata-3
In reply to this post by Curt Howland
Curt Howland composed on 2019-08-09 13:53 (UTC-0400):

> plymouth-quit-wait.service
...
> I have no idea what a "plymouth" is.

Several things it brings to the table:
1-avoids /dastardly/ "flicker" on mode switching during startup
2-bling/eye candy during startup
3-bloats initrd
4-encryption handling
5-yet another tool for systemd to disrupt startup/shutdown

Its progenitor seems to be Fedora/Gnome:
https://fedoraproject.org/wiki/Features/BetterStartup

On my own installations I always block its installation, or remove it, or disable
it, depending on how and when it comes to my attention.

https://en.wikipedia.org/wiki/Plymouth_(software)
--
Evolution as taught in public schools is religion, not science.

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/

Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Richard Hector
In reply to this post by Étienne Mollier
On 10/08/19 6:39 AM, Étienne Mollier wrote:
>> Those asterisks are also red, and moving left to right, the same as
>> seen during shutdown when something won't politely die.
> Just like the scanner in "Knight Rider", I see the picture.
>

Except I believe that was a Pontiac, not a Plymouth.

Sorry,

Richard


signature.asc (849 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Mart van de Wege
In reply to this post by Felix Miata-3
Felix Miata <[hidden email]> writes:

> Curt Howland composed on 2019-08-09 13:53 (UTC-0400):
>
>> plymouth-quit-wait.service
> ...
>> I have no idea what a "plymouth" is.
>
> Several things it brings to the table:
> 1-avoids /dastardly/ "flicker" on mode switching during startup
> 2-bling/eye candy during startup
> 3-bloats initrd
> 4-encryption handling
> 5-yet another tool for systemd to disrupt startup/shutdown
>
And beyond the FUD:

6. Make sure that the console is sane when you try to provide things
like LUKS passphrases; the default bootup console will happily overwrite
the prompt, making it completely unclear why your boot is
hanging. Plymouth will make sure that when you are prompted, nothing
else will muck up your console. This is why it does not even have to run
as a graphical bootloader, but can handle a text console as well.

Plymouth provides a pretty bootscreen as a side effect. It actually
provides a useful service.

On the other hand, it should die when booting is done, if the
plymouth-quit-wait.service hangs, something *else* is wrong.

Regards,

Mart

--
"We will need a longer wall when the revolution comes."
--- AJS, quoting an uncertain source.

Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

tomas@tuxteam.de
On Sun, Aug 11, 2019 at 07:11:19PM +0200, Mart van de Wege wrote:

> Felix Miata <[hidden email]> writes:
>
> > Curt Howland composed on 2019-08-09 13:53 (UTC-0400):
> >
> >> plymouth-quit-wait.service
> > ...
> >> I have no idea what a "plymouth" is.
> >
> > Several things it brings to the table:
> > 1-avoids /dastardly/ "flicker" on mode switching during startup
> > 2-bling/eye candy during startup
> > 3-bloats initrd
> > 4-encryption handling
> > 5-yet another tool for systemd to disrupt startup/shutdown
> >
> And beyond the FUD:
>
> 6. Make sure that the console is sane when you try to provide things
> like LUKS passphrases; the default bootup console will happily overwrite
> the prompt [...]
Huh? Under what conditions does this happen?

Cheers
-- t

signature.asc (205 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Mart van de Wege
<[hidden email]> writes:

> On Sun, Aug 11, 2019 at 07:11:19PM +0200, Mart van de Wege wrote:
>> Felix Miata <[hidden email]> writes:
>>
>> > Curt Howland composed on 2019-08-09 13:53 (UTC-0400):
>> >
>> >> plymouth-quit-wait.service
>> > ...
>> >> I have no idea what a "plymouth" is.
>> >
>> > Several things it brings to the table:
>> > 1-avoids /dastardly/ "flicker" on mode switching during startup
>> > 2-bling/eye candy during startup
>> > 3-bloats initrd
>> > 4-encryption handling
>> > 5-yet another tool for systemd to disrupt startup/shutdown
>> >
>> And beyond the FUD:
>>
>> 6. Make sure that the console is sane when you try to provide things
>> like LUKS passphrases; the default bootup console will happily overwrite
>> the prompt [...]
>
> Huh? Under what conditions does this happen?
>
Look it up. Plymouth was only introduced in wheezy. I'm not going to do
your homework for you.

Regards,

Mart

--
"We will need a longer wall when the revolution comes."
--- AJS, quoting an uncertain source.

Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

Reco
In reply to this post by tomas@tuxteam.de
        Hi.

On Sun, Aug 11, 2019 at 07:55:12PM +0200, [hidden email] wrote:

> On Sun, Aug 11, 2019 at 07:11:19PM +0200, Mart van de Wege wrote:
> > Felix Miata <[hidden email]> writes:
> >
> > > Curt Howland composed on 2019-08-09 13:53 (UTC-0400):
> > >
> > >> plymouth-quit-wait.service
> > > ...
> > >> I have no idea what a "plymouth" is.
> > >
> > > Several things it brings to the table:
> > > 1-avoids /dastardly/ "flicker" on mode switching during startup
> > > 2-bling/eye candy during startup
> > > 3-bloats initrd
> > > 4-encryption handling
> > > 5-yet another tool for systemd to disrupt startup/shutdown
> > >
> > And beyond the FUD:
> >
> > 6. Make sure that the console is sane when you try to provide things
> > like LUKS passphrases; the default bootup console will happily overwrite
> > the prompt [...]
>
> Huh? Under what conditions does this happen?

Unusual ones. Put several encrypted devices into crypttab(5) with "none"
option *and* make sure that none of them are mounted by initramfs
scripts. Or put systemd into initramfs like they do in RHEL.
In such setup systemd will ask you to enter several passphrases at same
time.

If you're using a single encrypted device which is setup by conventional
initramfs scripts - there's no need for plymouth as there's nothing to
fix. Everything works as intended. For me, at least.

In short, plymouth can fix problems, but you have to work hard to get
them.

Reco

Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

tomas@tuxteam.de
In reply to this post by Mart van de Wege
On Sun, Aug 11, 2019 at 10:18:27PM +0200, Mart van de Wege wrote:
> <[hidden email]> writes:
>
> > On Sun, Aug 11, 2019 at 07:11:19PM +0200, Mart van de Wege wrote:

[...]

> > Huh? Under what conditions does this happen?
> >
> Look it up. Plymouth was only introduced in wheezy. I'm not going to do
> your homework for you.

Look. I don't need any homework done. I'm fine with cryptsetup and
without Plymouth.

-- t

signature.asc (205 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Systemd start that won't stop

tomas@tuxteam.de
In reply to this post by Reco
On Sun, Aug 11, 2019 at 11:27:31PM +0300, Reco wrote:
> Hi.
>
> On Sun, Aug 11, 2019 at 07:55:12PM +0200, [hidden email] wrote:

[...]

> > Huh? Under what conditions does this happen?
>
> Unusual ones. Put several encrypted devices into crypttab(5) with "none"
> option *and* make sure that none of them are mounted by initramfs
> scripts. Or put systemd into initramfs like they do in RHEL.
> In such setup systemd will ask you to enter several passphrases at same
> time.

Ah, thank, Reco. That makes sense...

> If you're using a single encrypted device which is setup by conventional
> initramfs scripts - there's no need for plymouth as there's nothing to
> fix. Everything works as intended. For me, at least.

Yes, I now remember some folks here complaining that they couldn't enter
their passphrase...

> In short, plymouth can fix problems, but you have to work hard to get
> them.

I see ;-)

Cheers
-- t

signature.asc (205 bytes) Download Attachment