Re: Resolved (without understanding ; -) (was: Re: Error while trying to install openssh-server on Buster)

2020-07-24 Thread Andrei POPESCU
On Vi, 24 iul 20, 10:58:24, to...@tuxteam.de wrote:
> 
> It's as if they were copying the disruptive antipatterns of proprietary
> software companies. But we don't need those antipatterns in the free
> software context, do we?

One person's bug is another's feature.

Kind regards,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser


signature.asc
Description: PGP signature


Re: Resolved (without understanding ; -) (was: Re: Error while trying to install openssh-server on Buster)

2020-07-24 Thread tomas
On Thu, Jul 23, 2020 at 06:48:10PM -0500, David Wright wrote:
> On Thu 23 Jul 2020 at 10:12:09 (+0200), to...@tuxteam.de wrote:

[...]

> > Hours of fun :-)
> 
> Sure, I agree. But they're hours I don't really have. That's one
> reason why I don't run a DE: I just don't understand what's going on
> behind my back, and I don't enjoy that.

Same here. Actually, even the DBus daemon is banned from my box,
which, as a (welcome) side effect keeps miscellaneous other stuff
at a safe distance ;-)

> To start at the beginning; I looked up [...]

Good analysis.

> I don't think it's a bad idea: it seems better to have people posting
> "Where's my stick?" when their newly plugged-in USB stick fails to get
> automounted, rather than "Where's the partition with all my data on it?"

Definitely, yes.

What miffs me most about those DEs is, that they fall into that antipattern
of "reinventing the world" instead of trying the slower path and negotiate
with what's there. Kind of xkcd 927 over and over again.

It's as if they were copying the disruptive antipatterns of proprietary
software companies. But we don't need those antipatterns in the free
software context, do we?

Cheers
-- t


signature.asc
Description: Digital signature


Re: Resolved (without understanding ; -) (was: Re: Error while trying to install openssh-server on Buster)

2020-07-23 Thread David Wright
On Thu 23 Jul 2020 at 10:12:09 (+0200), to...@tuxteam.de wrote:
> On Wed, Jul 22, 2020 at 02:22:32PM -0500, David Wright wrote:
> > On Wed 22 Jul 2020 at 14:23:48 (-0400), rhkramer wrote:
> 
> [...]
> 
> > > The basic solution involved stopping gparted [...]
> 
> > AIUI gparted locks up the disks when you run it [...]
> 
> Whatever "locking up the disks" means, in this context.
> 
> Not criticising you, don't get me wrong. Just pointing out what
> kind of Rube Goldbergian dystopia we are heading to. There are
> many layers involved, and each one has a different notion of
> what a "disk" is and what "locking up" means. Under some moon
> phases, some of the layers mix.
> 
> Hours of fun :-)

Sure, I agree. But they're hours I don't really have. That's one
reason why I don't run a DE: I just don't understand what's going on
behind my back, and I don't enjoy that.

To start at the beginning; I looked up   man gparted   (it's installed,
but I don't use it) and saw that, unlike {f,g}disk, the device argument
is optional. I assumed that it would be unset when the program is run
from an icon. And assuming the OP was still running, IIRC, KDE, I put
two and two together.

Further down the page, it says "You can help to reduce the risk of
data loss by not mounting or unmounting partitions outside of the
gparted application while gparted is running." I assumed that KDE
might take it upon itself to prevent that happening.

I then searched for "lock" in gparted's Changes file. It mentions
that hal-lock is now used instead of fdi-policy. So I looked up
man hal-lock   and it says "hal-lock can be used to acquire a lock
on a given interface either on a given device or globally." So that
gave a mechanism.

I don't think it's a bad idea: it seems better to have people posting
"Where's my stick?" when their newly plugged-in USB stick fails to get
automounted, rather than "Where's the partition with all my data on it?"
But I prefer, as a sysadmin, to take such responsibilities on myself.

I did wonder about  apt-get install  and mount, but as I don't have
anything to install at the moment, I left on the side. Perhaps I'll
wrap the command in strace next time I use it.

Of course, you can always read my "AIUI" to mean I don't understand it
very well.

Cheers,
David.



Re: Resolved (without understanding ; -) (was: Re: Error while trying to install openssh-server on Buster)

2020-07-23 Thread tomas
On Wed, Jul 22, 2020 at 02:22:32PM -0500, David Wright wrote:
> On Wed 22 Jul 2020 at 14:23:48 (-0400), rhkramer wrote:

[...]

> > The basic solution involved stopping gparted [...]

> AIUI gparted locks up the disks when you run it [...]

Whatever "locking up the disks" means, in this context.

Not criticising you, don't get me wrong. Just pointing out what
kind of Rube Goldbergian dystopia we are heading to. There are
many layers involved, and each one has a different notion of
what a "disk" is and what "locking up" means. Under some moon
phases, some of the layers mix.

Hours of fun :-)

Cheers
-- t


signature.asc
Description: Digital signature


Re: Resolved (without understanding ;-) (was: Re: Error while trying to install openssh-server on Buster)

2020-07-22 Thread rhkramer
On Wednesday, July 22, 2020 02:54:27 PM Greg Wooledge wrote:
> On Wed, Jul 22, 2020 at 02:23:48PM -0400, rhkramer wrote:
> > The basic solution involved stopping gparted (I don't know why, but
> > google found a page that described the same problem I had and the page
> > said it occurred with gparted running and went away with gparted
> > stopped.
> 
> Fascinating.
> 
>  was my first Google result when I looked
> for the key words from your explanation.

Thanks, yes, fascinating -- my googling turned up a different link, without as 
much discussion / explanation.



Re: Resolved (without understanding ; -) (was: Re: Error while trying to install openssh-server on Buster)

2020-07-22 Thread David Wright
On Wed 22 Jul 2020 at 14:23:48 (-0400), rhkramer wrote:
> On Wednesday, July 22, 2020 7:08:40 AM EDT Andrew Cater wrote:
> > It should "just work" - if you can "ssh localhost" - the server is running.
> 
> Thanks for the reply.  The problem is solved, and I'll mention the solution 
> here with maybe more details in replies to Greg Wooledge and/or  Jonathan 
> Dowland.  
> 
> The basic solution involved stopping gparted (I don't know why, but google 
> found a page that described the same problem I had and the page said it 
> occurred with gparted running and went away with gparted stopped.

AIUI gparted locks up the disks when you run it. Perhaps it's
something to do with its ability to move and resize partitions.
I certainly wouldn't try to carry out other administration tasks
while I was running a partitioner, particular one that doesn't focus
on a particular device.

> Anyway, what I wanted to say here is that it did not "just work" while I had 
> that _.mount error, but I probably should have quoted a little more of the 
> output from apt-get install before I stopped gparted.

Yes. Obviously.

> Here it is
> 
> 
> Created symlink /etc/systemd/system/multi-user.target.wants/ssh.service → /
> lib/systemd/system/ssh.service.
> rescue-ssh.target is a disabled or a static unit, not starting it.
> Failed to start ssh.service: Unit -.mount is masked.
> invoke-rc.d: initscript ssh, action "start" failed.
> ● ssh.service - OpenBSD Secure Shell server
>Loaded: loaded (/lib/systemd/system/ssh.service; enabled; vendor preset: 
> enabled)
>Active: inactive (dead)
>  Docs: man:sshd(8)
>man:sshd_config(5)
> dpkg: error processing package openssh-server (--configure):
>  installed openssh-server package post-installation script subprocess 
> returned 
> error exit status 1
> Processing triggers for man-db (2.8.5-2) ...
> Processing triggers for systemd (241-7~deb10u4) ...
> Errors were encountered while processing:
>  openssh-server
> Error: GDBus.Error:org.freedesktop.systemd1.UnitMasked: Unit -.mount is 
> masked.
> E: Sub-process /usr/bin/dpkg returned an error code (1)
> 
> 
> After shutting down gparted, I ran apt-get install openssh-server again, and 
> did not get the message about _.mount being masked, and openssh-server is now 
> running.  (With a slight problem which I'll look into (I can't ssh in as root 
> -- but I assume I can find that problem -- I suspect something is preventing 
> remote logins by root -- I know I've seen an option or such to do that, so I 
> think I just have to find it again).)
> 
> Hmm, I wrote enough here that I might as well try to finish the story, with 
> what I might have written to Greg and/or Johnathan, which is basically the 
> following:
> 
> It's peachy that the problem is gone, but I really don't understand anything 
> about what the problem was or why shutting down gparted solved it.
> 
> I may (or may not) spend a little time digging into it (starting with finding 
> out what a unit is in systemd, and does that somehow relate to a mount point, 
> maybe in root, or if not, what specifically does -.mount refer to, and why it 
> has to use a - (in fact, a double -) instead of a /).  And I will rant, just 
> a 
> little bit ;-)
> 
> Slight rant (something to ignore): I mean, it is frustrating that I now have 
> to learn something new to do the same things I used to be able to do without 
> learning that new thing.  (The new thing being systemd.  I might consider 
> switching to a distro that does not use systemd, but I suppose that is just 
> burying my head in the sand.)

Yes, normally you can't login as root remotely. The usual way round
this is to exchange ssh keys so you can use passwordless login.
How you exchange said keys is left as an exercise to think about
for a while.

> > On Wed, Jul 22, 2020 at 2:06 AM rhkramer  wrote:
> > > I get this error when trying to apt-get install openssh-server on my (up
> > > to
> > > date) Buster system:
> > > 
> > > Error: GDBus.Error:org.freedesktop.systemd1.UnitMasked: Unit -.mount is
> > > masked.
> > > 
> > > I tried (based on the reference below):
> > > 
> > > root@s32:/# systemctl unmask org.freedesktop.systemd1.UnitMasked
> > > Unit org.freedesktop.systemd1.UnitMasked.service does not exist,
> > > proceeding
> > > anyway.
> > > 
> > > I don't really have a clue.  One googled page suggests
> > > 
> > > 
> > > Situation
> > > 
> > > Checking the status of a service shows it is masked.
> > > 
> > > Running systemctl unmask  doesn't change the status.
> > > 
> > > 
> > > Resolution
> > > The systemd unit file is empty.  Replace it in /usr/lib/systemd/system by
> > > reinstalling the package in which the unit file was contained.
> > > 

Cheers,
David.



Re: Resolved (without understanding ;-) (was: Re: Error while trying to install openssh-server on Buster)

2020-07-22 Thread Greg Wooledge
On Wed, Jul 22, 2020 at 02:23:48PM -0400, rhkramer wrote:
> The basic solution involved stopping gparted (I don't know why, but google 
> found a page that described the same problem I had and the page said it 
> occurred with gparted running and went away with gparted stopped.

Fascinating.

 was my first Google result when I looked
for the key words from your explanation.



Resolved (without understanding ;-) (was: Re: Error while trying to install openssh-server on Buster)

2020-07-22 Thread rhkramer
On Wednesday, July 22, 2020 7:08:40 AM EDT Andrew Cater wrote:
> It should "just work" - if you can "ssh localhost" - the server is running.

Thanks for the reply.  The problem is solved, and I'll mention the solution 
here with maybe more details in replies to Greg Wooledge and/or  Jonathan 
Dowland.  

The basic solution involved stopping gparted (I don't know why, but google 
found a page that described the same problem I had and the page said it 
occurred with gparted running and went away with gparted stopped.

Anyway, what I wanted to say here is that it did not "just work" while I had 
that _.mount error, but I probably should have quoted a little more of the 
output from apt-get install before I stopped gparted.  Here it is


Created symlink /etc/systemd/system/multi-user.target.wants/ssh.service → /
lib/systemd/system/ssh.service.
rescue-ssh.target is a disabled or a static unit, not starting it.
Failed to start ssh.service: Unit -.mount is masked.
invoke-rc.d: initscript ssh, action "start" failed.
● ssh.service - OpenBSD Secure Shell server
   Loaded: loaded (/lib/systemd/system/ssh.service; enabled; vendor preset: 
enabled)
   Active: inactive (dead)
 Docs: man:sshd(8)
   man:sshd_config(5)
dpkg: error processing package openssh-server (--configure):
 installed openssh-server package post-installation script subprocess returned 
error exit status 1
Processing triggers for man-db (2.8.5-2) ...
Processing triggers for systemd (241-7~deb10u4) ...
Errors were encountered while processing:
 openssh-server
Error: GDBus.Error:org.freedesktop.systemd1.UnitMasked: Unit -.mount is 
masked.
E: Sub-process /usr/bin/dpkg returned an error code (1)


After shutting down gparted, I ran apt-get install openssh-server again, and 
did not get the message about _.mount being masked, and openssh-server is now 
running.  (With a slight problem which I'll look into (I can't ssh in as root 
-- but I assume I can find that problem -- I suspect something is preventing 
remote logins by root -- I know I've seen an option or such to do that, so I 
think I just have to find it again).)

Hmm, I wrote enough here that I might as well try to finish the story, with 
what I might have written to Greg and/or Johnathan, which is basically the 
following:

It's peachy that the problem is gone, but I really don't understand anything 
about what the problem was or why shutting down gparted solved it.

I may (or may not) spend a little time digging into it (starting with finding 
out what a unit is in systemd, and does that somehow relate to a mount point, 
maybe in root, or if not, what specifically does -.mount refer to, and why it 
has to use a - (in fact, a double -) instead of a /).  And I will rant, just a 
little bit ;-)

Slight rant (something to ignore): I mean, it is frustrating that I now have 
to learn something new to do the same things I used to be able to do without 
learning that new thing.  (The new thing being systemd.  I might consider 
switching to a distro that does not use systemd, but I suppose that is just 
burying my head in the sand.)





> 
> On Wed, Jul 22, 2020 at 2:06 AM rhkramer  wrote:
> > I get this error when trying to apt-get install openssh-server on my (up
> > to
> > date) Buster system:
> > 
> > Error: GDBus.Error:org.freedesktop.systemd1.UnitMasked: Unit -.mount is
> > masked.
> > 
> > I tried (based on the reference below):
> > 
> > root@s32:/# systemctl unmask org.freedesktop.systemd1.UnitMasked
> > Unit org.freedesktop.systemd1.UnitMasked.service does not exist,
> > proceeding
> > anyway.
> > 
> > I don't really have a clue.  One googled page suggests
> > 
> > 
> > Situation
> > 
> > Checking the status of a service shows it is masked.
> > 
> > Running systemctl unmask  doesn't change the status.
> > 
> > 
> > Resolution
> > The systemd unit file is empty.  Replace it in /usr/lib/systemd/system by
> > reinstalling the package in which the unit file was contained.
> >