[gentoo-user] systemd

[gentoo-user] systemd

[Stefan G. Weichinger]

Ok then, separate thread ;-)

I just watched this:

http://www.youtube.com/watch?v=TyMLi8QF6sw

while I continued testing systemd in my VM.

The VM runs ~amd64, so far only a few services started (I still get my
head around how to enable/disable specific services/targets), and it
boots really fast.

I am not the ricer-kind-of-gentoo-users, but it impresses me anyway.

And I clearly see the benefits of socket-based-activation,
cgroup-control and other stuff systemd brings.

AFAI see the wiki mentioned is a starting point only.
It provides the first steps, but not much more, at least to me, right now.

-

There is a layman-overlay "systemd" which brings units (service-files)
to your system. AFAI understand it is still up to the user to
enable/link these files into their runlevel?

If it is that way it feels like a bit of trial-and-error to me to get my
systemd-setup doing the same things my current openrc-system does.

Especially for not-so-trivial stuff like the network settings for KVM
(bridging, TUN/TAP ...)

My approach would be to "rc-config show", take that list and try to
enable the according services within systemd.

Maybe I am completely wrong, maybe not.

I'd be happy to discuss these things with you gentoo-users.

Thanks, greets, Stefan

Re: [gentoo-user] systemd

[Sebastian Beßler]

[-- Attachment #1: Type: text/plain, Size: 635 bytes --]

Am 17.08.2011 23:04, schrieb Stefan G. Weichinger:

> I just watched this:
> 
> http://www.youtube.com/watch?v=TyMLi8QF6sw

Great Video, thanks for the Link.
I wanted to try systemd for quite some time but now I think that I will
install it on my Desktop-PC tomorrow. The summer here this year is very
wet, so leaving the house is not really fun so I have time to kill.

> I'd be happy to discuss these things with you gentoo-users.

I will use that offer and will keep you, and everyone else here, up to
date and posted.

But now it is time for bed here.

Greetings and good night or good day

Sebastian Beßler


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 900 bytes --]

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 18.08.2011 00:54, schrieb Sebastian Beßler:

>> I'd be happy to discuss these things with you gentoo-users.
> 
> I will use that offer and will keep you, and everyone else here, up 
> to date and posted.

looking fwd to your report.
greets, Stefan

Re: [gentoo-user] systemd

[Sebastian Beßler]

[-- Attachment #1: Type: text/plain, Size: 323 bytes --]

As always when I want to do anything like this there comes something
more important along and occupies all of my time.

So migration to systemd is stoped for now.
Hope I will come to it soon.

Greets
Sebastian

Am 20.08.2011 22:22, schrieb Stefan G. Weichinger:
> looking fwd to your report.
> greets, Stefan


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 900 bytes --]

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 2011-08-20 22:54, schrieb Sebastian Beßler:
> As always when I want to do anything like this there comes something 
> more important along and occupies all of my time.
> 
> So migration to systemd is stoped for now. Hope I will come to it
> soon.


Continued playing and learning and enabled it on my ~amd64 thinkpad.

Networkmanager and gdm work already, sound as well, it was quite easy.
enabled syslog-ng and sshd.socket

Beauty issues:

I get dozens of
/sys/devices/virtual/tty/tty* services running.
Don't know if that has to be that way.

My encrypted /home is now mounted three times. This might be related to
suspend-to-ram and the fact that I maybe should look deeper into how to
cryptsetup within systemd.

So far it works, but it isn't correct and also seems to trigger problems
with waking up from hibernation (multiple password-requests ...).

I will continue exploring other services soon.
Especially how to set up KVM, vmware, and the needed network-bridging etc.

Stefan

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

next box tested.
Installed systemd on my main workstation now that I understood how to
easily flip back to booting w/ openrc in case of problems.

I heavily use LVM here and this gives me the following issues:

I use lvm.service from the gentoo-wiki:

http://en.gentoo-wiki.com/wiki/Systemd#LVM

For sure I enabled it ...

When I boot this machine it boots up to starting the LVM-devices and
waits for some time then writes something like:

welcome to emergency mode ...
Start of /dev/VG...something failed (due to some dependencies)

(for all the LVs)

and lets me login or press Ctrl-D to continue.

(I can't remember the exact words, don't know if they are logged somewhere)

When I press Ctrl-D all the LVs are mounted(!) and it boots up fine to
graphical login.

hmm.

It then tells me:

# systemctl status lvm.service
lvm.service - Linux Volume Manager
	  Loaded: loaded (/etc/systemd/system/lvm.service)
	  Active: active (exited) since Mon, 22 Aug 2011 09:55:36 +0200; 22min ago
	 Process: 5568 ExecStop=/sbin/lvchange --sysinit -a ln $(/sbin/vgs -o
vg_name --noheadings --nosuffix 2> /dev/null) (code=exited, status=3)
	 Process: 5861 ExecStart=/sbin/vgchange --sysinit -a ly (code=exited,
status=0/SUCCESS)
	 Process: 5738 ExecStart=/sbin/vgscan --mknodes --ignorelockingfailure
(code=exited, status=0/SUCCESS)
	 Process: 5654 ExecStart=/sbin/pvscan --ignorelockingfailure
(code=exited, status=0/SUCCESS)
	  CGroup: name=systemd:/system/lvm.service

Why does that ExecStop fail? Why is it called at boot anyway?

I also tried another lvm.service from the russian gentoo-wiki, that
servicefile just pulls in /etc/init.d/lvm, but that didn't help so I
went back to the mentioned file.

What I wonder: what changes between running into that timeout and my
pressing Ctrl-D?

Thanks for helpful comments, Stefan

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 22.08.2011 10:26, schrieb Stefan G. Weichinger:

> What I wonder: what changes between running into that timeout and my
> pressing Ctrl-D?

To me it seems that the underlying RAID-device (which is the PV inside
the LVM-VG) isn't up fast enough.

Trying to figure it out now.

Re: [gentoo-user] systemd

[Joost Roeleveld]

On Monday, August 22, 2011 11:54:48 AM Stefan G. Weichinger wrote:
> Am 22.08.2011 10:26, schrieb Stefan G. Weichinger:
> > What I wonder: what changes between running into that timeout and my
> > pressing Ctrl-D?
> 
> To me it seems that the underlying RAID-device (which is the PV inside
> the LVM-VG) isn't up fast enough.
> 
> Trying to figure it out now.

Are they actually started in the right order?
In other words, first RAID, then LVM?

--
Joost

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 22.08.2011 12:26, schrieb Joost Roeleveld:

> Are they actually started in the right order?
> In other words, first RAID, then LVM?

I don't know ;-)

I still try to understand all this.
There is no specific RAID-service-file, so it seems to be done by udev
and the related target/service somehow.


lvm.service says

After=udev-settle.service

so udev should detect all devices first, then lvm.service gets started.

But somehow this doesn't work here, only after hitting that timeout one
time.

Stefan

Re: [gentoo-user] systemd

[Joost Roeleveld]

On Monday, August 22, 2011 12:31:19 PM Stefan G. Weichinger wrote:
> Am 22.08.2011 12:26, schrieb Joost Roeleveld:
> > Are they actually started in the right order?
> > In other words, first RAID, then LVM?
> 
> I don't know ;-)
> 
> I still try to understand all this.
> There is no specific RAID-service-file, so it seems to be done by udev
> and the related target/service somehow.
> 
> 
> lvm.service says
> 
> After=udev-settle.service
> 
> so udev should detect all devices first, then lvm.service gets started.
> 
> But somehow this doesn't work here, only after hitting that timeout one
> time.

That's unfortunate. The stop-service might be started to try to clean up when 
it fails.

What kind of RAID are you using? Does it perhaps rely on a module that is 
loaded in the background?
In which case you could try adding that module as a dependency?

--
Joost

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 2011-08-22 13:42, schrieb Joost Roeleveld:

> That's unfortunate. The stop-service might be started to try to clean
> up when it fails.

Hm, yes, I understand.

> What kind of RAID are you using? Does it perhaps rely on a module
> that is loaded in the background? In which case you could try adding
> that module as a dependency?

This is plain RAID1 and the kernel has that built in, so no, there is no
module used here for RAID.

There are LVM-options as modules:

crypt target, snapshot target, mirror target ... but that should not
play a role here.

S

Re: [gentoo-user] systemd

[Joost Roeleveld]

On Monday, August 22, 2011 06:55:21 PM Stefan G. Weichinger wrote:
> Am 2011-08-22 13:42, schrieb Joost Roeleveld:
> > That's unfortunate. The stop-service might be started to try to clean
> > up when it fails.
> 
> Hm, yes, I understand.
> 
> > What kind of RAID are you using? Does it perhaps rely on a module
> > that is loaded in the background? In which case you could try adding
> > that module as a dependency?
> 
> This is plain RAID1 and the kernel has that built in, so no, there is no
> module used here for RAID.
> 
> There are LVM-options as modules:
> 
> crypt target, snapshot target, mirror target ... but that should not
> play a role here.

Not really, unless you actually have any of these defined...
I do hope you can figure this one out as I do like the idea of systemd. But I 
need RAID and LVM to work correctly for my system to boot.

--
Joost

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 22.08.2011 19:03, schrieb Joost Roeleveld:

> I do hope you can figure this one out as I do like the idea of systemd. But I 
> need RAID and LVM to work correctly for my system to boot.

Got it.

Compared this one:

https://github.com/falconindy/initscripts-systemd/blob/master/lvm.service

w/ the lvm.service in the gentoo wiki.

The line:

Requires=udev-settle.service

missed, I added it and now it boots up straight and fast.

Gotta re-test it right now ;-)

S

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 22.08.2011 20:24, schrieb Stefan G. Weichinger:

> The line:
> 
> Requires=udev-settle.service
> 
> missed, I added it and now it boots up straight and fast.

update: edited the example in the gentoo-wiki now.

S

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 22.08.2011 20:29, schrieb Stefan G. Weichinger:

> update: edited the example in the gentoo-wiki now.

replying to myself once more, which makes it feel more like a wiki or
blog than a mailing-list ;-)

additional thoughts:

* as there is readahead-support in systemd I assume I could get rid of
preload:

http://packages.gentoo.org/package/sys-apps/preload?arches=all

As it isn't maintained actively anymore it maybe isn't of much use
anymore anyway?

As there is no related service-file for preload here, it is deactivated
for now anyway (as long as I choose the systemd-using GRUB-line).

* remember those cgroup-hacks back then?

http://en.gentoo-wiki.com/wiki/Improve_responsiveness_with_cgroups

Is that stuff still valid?

With systemd the whole use of cgroups changes fundamentally, I don't
have the knowledge to decide if to use both in parallel.

For now I disabled the stuff from the wiki (stop sourcing
/etc/bash/local/cgrouprc) as it only gives me warnings ...

* found this blog-entry against systemd:

http://monolight.cc/2011/05/the-systemd-fallacy/

I agree, it might be more useful on desktops ... so far I am still
exploring and learning to get to the point to make a decision where and
if to use.

-

regards, Stefan

Re: [gentoo-user] systemd

[Joost Roeleveld]

On Monday, August 22, 2011 11:09:02 PM Stefan G. Weichinger wrote:
> Am 22.08.2011 20:29, schrieb Stefan G. Weichinger:
> > update: edited the example in the gentoo-wiki now.
> 
> replying to myself once more, which makes it feel more like a wiki or
> blog than a mailing-list ;-)

There wasn't much to add. You provided a solution and the only reply I could 
come up with "Well done" would sound condescending. Which is why I decided not 
to.

> additional thoughts:
> 
> * as there is readahead-support in systemd I assume I could get rid of
> preload:
> 
> http://packages.gentoo.org/package/sys-apps/preload?arches=all
> 
> As it isn't maintained actively anymore it maybe isn't of much use
> anymore anyway?

I don't tend to use preload. Is it usefull in a non-systemd environment?

> As there is no related service-file for preload here, it is deactivated
> for now anyway (as long as I choose the systemd-using GRUB-line).
> 
> * remember those cgroup-hacks back then?
> 
> http://en.gentoo-wiki.com/wiki/Improve_responsiveness_with_cgroups
> 
> Is that stuff still valid?

Maybe, if you want to group stuff you're running yourself into seperate 
groups. The different services are grouped already.

> With systemd the whole use of cgroups changes fundamentally, I don't
> have the knowledge to decide if to use both in parallel.
> 
> For now I disabled the stuff from the wiki (stop sourcing
> /etc/bash/local/cgrouprc) as it only gives me warnings ...

What kind of warnings? Systemd already mounts the filesystem for it and starts 
poulating it. If your script does similar things, they might try to duplicate 
work?

> * found this blog-entry against systemd:
> 
> http://monolight.cc/2011/05/the-systemd-fallacy/
> 
> I agree, it might be more useful on desktops ... so far I am still
> exploring and learning to get to the point to make a decision where and
> if to use.

I think it is more useful on desktops and laptops, which get rebooted 
regularly. On a server that tends to run for months without a reboot, a fast 
init-system is important.

And I don't really see the point of D-BUS on a server either. All the services 
that need to talk to each other already have working communication paths.

I do intend to implement it on my desktop and netbook as I'd like to have 
those booting as fast as possible.

--
Joost

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 2011-08-23 08:27, schrieb Joost Roeleveld:
> On Monday, August 22, 2011 11:09:02 PM Stefan G. Weichinger wrote:
>> Am 22.08.2011 20:29, schrieb Stefan G. Weichinger:
>>> update: edited the example in the gentoo-wiki now.
>> 
>> replying to myself once more, which makes it feel more like a wiki
>> or blog than a mailing-list ;-)
> 
> There wasn't much to add. You provided a solution and the only reply
> I could come up with "Well done" would sound condescending. Which is
> why I decided not to.

ok, yes

> I don't tend to use preload. Is it usefull in a non-systemd
> environment?

I always had the impression that things started faster with preload,
yes. Might be less of an impact with the new SSD I have in my desktop
machine now.

I didn't really miss it when switching to systemd (where I don't have a
service-file for it yet).

>> http://en.gentoo-wiki.com/wiki/Improve_responsiveness_with_cgroups
>> 
>> Is that stuff still valid?
> 
> Maybe, if you want to group stuff you're running yourself into
> seperate groups. The different services are grouped already.
> 
>> With systemd the whole use of cgroups changes fundamentally, I
>> don't have the knowledge to decide if to use both in parallel.
>> 
>> For now I disabled the stuff from the wiki (stop sourcing 
>> /etc/bash/local/cgrouprc) as it only gives me warnings ...
> 
> What kind of warnings? Systemd already mounts the filesystem for it
> and starts poulating it. If your script does similar things, they
> might try to duplicate work?

The code tries to write to its own dir:

mkdir -p -m 0700 $cdir/user/$$ > /dev/null 2>&1
/bin/echo $$ > $cdir/user/$$/tasks
/bin/echo '1' > $cdir/user/$$/notify_on_release

But somehow the mkdir seems to fail as I get warnings from the two
echo-statements, that their "target-files" do not exist, which lead me
to the fact that $cdir/user/$$ does not exist.

> I think it is more useful on desktops and laptops, which get rebooted
>  regularly. On a server that tends to run for months without a
> reboot, a fast init-system is important.

You mean, "not so important" ?

> And I don't really see the point of D-BUS on a server either. All the
> services that need to talk to each other already have working
> communication paths.
> 
> I do intend to implement it on my desktop and netbook as I'd like to
> have those booting as fast as possible.

Yep, I agree.
Stefan

Re: [gentoo-user] systemd

[Joost Roeleveld]

On Tuesday, August 23, 2011 10:30:38 AM Stefan G. Weichinger wrote:
> Am 2011-08-23 08:27, schrieb Joost Roeleveld:
> > On Monday, August 22, 2011 11:09:02 PM Stefan G. Weichinger wrote:
> >> Am 22.08.2011 20:29, schrieb Stefan G. Weichinger:
> > I don't tend to use preload. Is it usefull in a non-systemd
> > environment?
> 
> I always had the impression that things started faster with preload,
> yes. Might be less of an impact with the new SSD I have in my desktop
> machine now.
> 
> I didn't really miss it when switching to systemd (where I don't have a
> service-file for it yet).

Guess it doesn't have much of an improvement anymore? :)

> >> http://en.gentoo-wiki.com/wiki/Improve_responsiveness_with_cgroups
> >> 
> >> Is that stuff still valid?
> > 
> > Maybe, if you want to group stuff you're running yourself into
> > seperate groups. The different services are grouped already.
> > 
> >> With systemd the whole use of cgroups changes fundamentally, I
> >> don't have the knowledge to decide if to use both in parallel.
> >> 
> >> For now I disabled the stuff from the wiki (stop sourcing
> >> /etc/bash/local/cgrouprc) as it only gives me warnings ...
> > 
> > What kind of warnings? Systemd already mounts the filesystem for it
> > and starts poulating it. If your script does similar things, they
> > might try to duplicate work?
> 
> The code tries to write to its own dir:
> 
> mkdir -p -m 0700 $cdir/user/$$ > /dev/null 2>&1
> /bin/echo $$ > $cdir/user/$$/tasks
> /bin/echo '1' > $cdir/user/$$/notify_on_release
> 
> But somehow the mkdir seems to fail as I get warnings from the two
> echo-statements, that their "target-files" do not exist, which lead me
> to the fact that $cdir/user/$$ does not exist.

You could try adding ls-statements to see if it can set that op?
Or try to run those commands.

Where is $cdir pointing to?

> > I think it is more useful on desktops and laptops, which get rebooted
> > 
> >  regularly. On a server that tends to run for months without a
> > 
> > reboot, a fast init-system is important.
> 
> You mean, "not so important" ?

Yes, that's what I meant :)

> > And I don't really see the point of D-BUS on a server either. All the
> > services that need to talk to each other already have working
> > communication paths.
> > 
> > I do intend to implement it on my desktop and netbook as I'd like to
> > have those booting as fast as possible.
> 
> Yep, I agree.
> Stefan

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 2011-08-23 11:04, schrieb Joost Roeleveld:
>> The code tries to write to its own dir:
>> 
>> mkdir -p -m 0700 $cdir/user/$$ > /dev/null 2>&1 /bin/echo $$ >
>> $cdir/user/$$/tasks /bin/echo '1' >
>> $cdir/user/$$/notify_on_release
>> 
>> But somehow the mkdir seems to fail as I get warnings from the two 
>> echo-statements, that their "target-files" do not exist, which lead
>> me to the fact that $cdir/user/$$ does not exist.
> 
> You could try adding ls-statements to see if it can set that op? Or
> try to run those commands.
> 
> Where is $cdir pointing to?

/sys/fs/cgroup which exists.

I removed the "> /dev/null..." part and now I see that I have a
permission problem, my user isn't allowed to mkdir there.

Will solve that ...

otoh it might be overkill to create my own cgroups as systemd does it
anyway, correct?

S

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 2011-08-23 11:17, schrieb Stefan G. Weichinger:

> I removed the "> /dev/null..." part and now I see that I have a 
> permission problem, my user isn't allowed to mkdir there.
> 
> Will solve that ...

Rather easy to see:

http://en.gentoo-wiki.com/wiki/Improve_responsiveness_with_cgroups

brings the script /usr/local/sbin/cgroup_start which is started by
openrc, but not by systemd. In there the perms would be set up for my
user ...


So the solution will be to teach systemd to start that script as well.

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 23.08.2011 11:22, schrieb Stefan G. Weichinger:

> http://en.gentoo-wiki.com/wiki/Improve_responsiveness_with_cgroups
> 
> brings the script /usr/local/sbin/cgroup_start which is started by
> openrc, but not by systemd. In there the perms would be set up for my
> user ...
> 
> 
> So the solution will be to teach systemd to start that script as well.

That was no big problem ... solved.

Interesting observation right now:

Wanted to extend a LV.

Unmounted it, "lvresize -L+5G ...", then "resize2fs ..."

resize2fs told me that the LV is mounted and that is has to do online
resizing.

whoa. I unmounted it before!

So it seems as if I would have to stop the related mount-service within
systemd first ...

That is an important thing to know IMO.

Stefan

Re: [gentoo-user] systemd

[Joost Roeleveld]

On Wednesday, August 24, 2011 12:00:17 AM Stefan G. Weichinger wrote:
> Am 23.08.2011 11:22, schrieb Stefan G. Weichinger:
> > http://en.gentoo-wiki.com/wiki/Improve_responsiveness_with_cgroups
> > 
> > brings the script /usr/local/sbin/cgroup_start which is started by
> > openrc, but not by systemd. In there the perms would be set up for my
> > user ...
> > 
> > 
> > So the solution will be to teach systemd to start that script as well.
> 
> That was no big problem ... solved.
> 
> Interesting observation right now:
> 
> Wanted to extend a LV.
> 
> Unmounted it, "lvresize -L+5G ...", then "resize2fs ..."
> 
> resize2fs told me that the LV is mounted and that is has to do online
> resizing.
> 
> whoa. I unmounted it before!
> 
> So it seems as if I would have to stop the related mount-service within
> systemd first ...
> 
> That is an important thing to know IMO.

You can resize a partition without having to umount it first.

That's been possible for a couple of years now. :)

--
Joost

Re: [gentoo-user] systemd

[Joost Roeleveld]

On Monday, August 22, 2011 11:09:02 PM Stefan G. Weichinger wrote:
> Am 22.08.2011 20:29, schrieb Stefan G. Weichinger:
> > update: edited the example in the gentoo-wiki now.
> 
> replying to myself once more, which makes it feel more like a wiki or
> blog than a mailing-list ;-)

There wasn't much to add. You provided a solution and the only reply I could 
come up with "Well done" would sound condescending. Which is why I decided not 
to.

> additional thoughts:
> 
> * as there is readahead-support in systemd I assume I could get rid of
> preload:
> 
> http://packages.gentoo.org/package/sys-apps/preload?arches=all
> 
> As it isn't maintained actively anymore it maybe isn't of much use
> anymore anyway?

I don't tend to use preload. Is it usefull in a non-systemd environment?

> As there is no related service-file for preload here, it is deactivated
> for now anyway (as long as I choose the systemd-using GRUB-line).
> 
> * remember those cgroup-hacks back then?
> 
> http://en.gentoo-wiki.com/wiki/Improve_responsiveness_with_cgroups
> 
> Is that stuff still valid?

Maybe, if you want to group stuff you're running yourself into seperate 
groups. The different services are grouped already.

> With systemd the whole use of cgroups changes fundamentally, I don't
> have the knowledge to decide if to use both in parallel.
> 
> For now I disabled the stuff from the wiki (stop sourcing
> /etc/bash/local/cgrouprc) as it only gives me warnings ...

What kind of warnings? Systemd already mounts the filesystem for it and starts 
poulating it. If your script does similar things, they might try to duplicate 
work?

> * found this blog-entry against systemd:
> 
> http://monolight.cc/2011/05/the-systemd-fallacy/
> 
> I agree, it might be more useful on desktops ... so far I am still
> exploring and learning to get to the point to make a decision where and
> if to use.

I think it is more useful on desktops and laptops, which get rebooted 
regularly. On a server that tends to run for months without a reboot, a fast 
init-system is important.

And I don't really see the point of D-BUS on a server either. All the services 
that need to talk to each other already have working communication paths.

I do intend to implement it on my desktop and netbook as I'd like to have 
those booting as fast as possible.

--
Joost

Re: [gentoo-user] systemd

[Stroller]

On 23 August 2011, at 07:27, Joost Roeleveld wrote:
> ...
>> * found this blog-entry against systemd:
>> 
>> http://monolight.cc/2011/05/the-systemd-fallacy/
>> 
>> I agree, it might be more useful on desktops ... so far I am still
>> exploring and learning to get to the point to make a decision where and
>> if to use.
> 
> I think it is more useful on desktops and laptops, which get rebooted 
> regularly. On a server that tends to run for months without a reboot, a fast 
> init-system is important.
> 
> And I don't really see the point of D-BUS on a server either. All the services 
> that need to talk to each other already have working communication paths.

Reading that blog entry I found discouraging the idea that dbus might be required on my servers in the future, if systemd becomes popular with distros.

Stroller.

Re: [gentoo-user] systemd

[Canek Peláez Valdés]

On Tue, Aug 23, 2011 at 1:17 PM, Stroller
<stroller@stellar.eclipse.co.uk> wrote:
>
> On 23 August 2011, at 07:27, Joost Roeleveld wrote:
>> ...
>>> * found this blog-entry against systemd:
>>>
>>> http://monolight.cc/2011/05/the-systemd-fallacy/
>>>
>>> I agree, it might be more useful on desktops ... so far I am still
>>> exploring and learning to get to the point to make a decision where and
>>> if to use.
>>
>> I think it is more useful on desktops and laptops, which get rebooted
>> regularly. On a server that tends to run for months without a reboot, a fast
>> init-system is important.
>>
>> And I don't really see the point of D-BUS on a server either. All the services
>> that need to talk to each other already have working communication paths.
>
> Reading that blog entry I found discouraging the idea that dbus might be required on my servers in the future, if systemd becomes popular with distros.

I don't see the problem with D-Bus. It's small (the only hard
dependency it has is an XML parser), and it provides the Linux/UNIX
(de facto) standard interprocess communication system.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] systemd

[Alan McKinnon]

On Tue 23 August 2011 18:17:17 Stroller did opine thusly:
> On 23 August 2011, at 07:27, Joost Roeleveld wrote:
> > ...
> > 
> >> * found this blog-entry against systemd:
> >> 
> >> http://monolight.cc/2011/05/the-systemd-fallacy/
> >> 
> >> I agree, it might be more useful on desktops ... so far I am
> >> still exploring and learning to get to the point to make a
> >> decision where and if to use.
> > 
> > I think it is more useful on desktops and laptops, which get
> > rebooted regularly. On a server that tends to run for months
> > without a reboot, a fast init-system is important.
> > 
> > And I don't really see the point of D-BUS on a server either.
> > All the services that need to talk to each other already have
> > working communication paths.
> Reading that blog entry I found discouraging the idea that dbus
> might be required on my servers in the future, if systemd becomes
> popular with distros.

What's your objection to dbus? It gives you a standard message bus, is 
small, light, consumes minimal resources and provides a nice standard 
way to do IPC. Probably easier than reinventing the wheel with named 
pipes and other bits over and over.

Now if it had similarities to say hal, I would instantly understand. 
But dbus is good and useful in all the ways that hal isn't.

-- 
alan dot mckinnon at gmail dot com

Re: [gentoo-user] systemd

[Canek Peláez Valdés]

On Tue, Aug 23, 2011 at 2:57 PM, Alan McKinnon <alan.mckinnon@gmail.com> wrote:
> On Tue 23 August 2011 18:17:17 Stroller did opine thusly:
>> On 23 August 2011, at 07:27, Joost Roeleveld wrote:
>> > ...
>> >
>> >> * found this blog-entry against systemd:
>> >>
>> >> http://monolight.cc/2011/05/the-systemd-fallacy/
>> >>
>> >> I agree, it might be more useful on desktops ... so far I am
>> >> still exploring and learning to get to the point to make a
>> >> decision where and if to use.
>> >
>> > I think it is more useful on desktops and laptops, which get
>> > rebooted regularly. On a server that tends to run for months
>> > without a reboot, a fast init-system is important.
>> >
>> > And I don't really see the point of D-BUS on a server either.
>> > All the services that need to talk to each other already have
>> > working communication paths.
>> Reading that blog entry I found discouraging the idea that dbus
>> might be required on my servers in the future, if systemd becomes
>> popular with distros.
>
> What's your objection to dbus? It gives you a standard message bus, is
> small, light, consumes minimal resources and provides a nice standard
> way to do IPC. Probably easier than reinventing the wheel with named
> pipes and other bits over and over.
>
> Now if it had similarities to say hal, I would instantly understand.
> But dbus is good and useful in all the ways that hal isn't.

Wasn't. HAL is dead. From http://www.freedesktop.org/wiki/Software/hal

"HAL is in maintenance mode - no new features are added. All future
development focuses on udisks, upower and other parts of the stack.
See Software/DeviceKit for more information."

HAL was an experiment, and it failed. At some point, dbus was an
experiment, and I think it succeeded. Most technologies (KDE, GNOME,
pulseaudio, udev, devfs, ALSA, systemd, upstart) start as experiments.
Some fail, some succeed, and some are replaced by even newer
experiments.

It's the only way new and interesting stuff gets created.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] systemd

[Alan McKinnon]

On Tue 23 August 2011 15:06:25 Canek Peláez Valdés did opine thusly:
> > Now if it had similarities to say hal, I would instantly
> > understand. But dbus is good and useful in all the ways that
> > hal isn't.
> Wasn't. HAL is dead. From
> http://www.freedesktop.org/wiki/Software/hal

Sadly, HAL is not yet dead. It lives still.

It lives on the production database server I just happen to be 
rebooting as I type this (another story for another time) and will 
continue to live here for a very very long time indeed.

Dale can confirm this. Dale will swear in a court of law with hand on 
bible than hal lives on in zombie form, infesting all the matter of 
his house and computers, infecting them with their undead zombieness.

Ye gods, it's been a long hard day....



-- 
alan dot mckinnon at gmail dot com

Re: [gentoo-user] systemd

[Canek Peláez Valdés]

On Tue, Aug 23, 2011 at 3:43 PM, Alan McKinnon <alan.mckinnon@gmail.com> wrote:
> On Tue 23 August 2011 15:06:25 Canek Peláez Valdés did opine thusly:
>> > Now if it had similarities to say hal, I would instantly
>> > understand. But dbus is good and useful in all the ways that
>> > hal isn't.
>> Wasn't. HAL is dead. From
>> http://www.freedesktop.org/wiki/Software/hal
>
> Sadly, HAL is not yet dead. It lives still.
>
> It lives on the production database server I just happen to be
> rebooting as I type this (another story for another time) and will
> continue to live here for a very very long time indeed.
>
> Dale can confirm this. Dale will swear in a court of law with hand on
> bible than hal lives on in zombie form, infesting all the matter of
> his house and computers, infecting them with their undead zombieness.
>
> Ye gods, it's been a long hard day....

I remember getting rid of HAL in one weekend, from all my computers.
It was a long weekend, but it was not as bad as getting rid of Qt from
all the computers in my office some years ago.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] systemd

[Alan McKinnon]

On Tue 23 August 2011 15:50:24 Canek Peláez Valdés did opine thusly:
> On Tue, Aug 23, 2011 at 3:43 PM, Alan McKinnon 
<alan.mckinnon@gmail.com> wrote:
> > On Tue 23 August 2011 15:06:25 Canek Peláez Valdés did opine 
thusly:
> >> > Now if it had similarities to say hal, I would instantly
> >> > understand. But dbus is good and useful in all the ways
> >> > that
> >> > hal isn't.
> >> 
> >> Wasn't. HAL is dead. From
> >> http://www.freedesktop.org/wiki/Software/hal
> > 
> > Sadly, HAL is not yet dead. It lives still.
> > 
> > It lives on the production database server I just happen to be
> > rebooting as I type this (another story for another time) and
> > will continue to live here for a very very long time indeed.
> > 
> > Dale can confirm this. Dale will swear in a court of law with
> > hand on bible than hal lives on in zombie form, infesting all
> > the matter of his house and computers, infecting them with
> > their undead zombieness.
> > 
> > Ye gods, it's been a long hard day....
> 
> I remember getting rid of HAL in one weekend, from all my computers.
> It was a long weekend, but it was not as bad as getting rid of Qt
> from all the computers in my office some years ago.

Come to my work place, I have the perfect task for you:

to excise perl-5.8.0 from all the many machines it's on, plus the 
atrocious in-house coding using it that no-one left understands, and 
all running on hardware that no-one can replace.


-- 
alan dot mckinnon at gmail dot com

Re: [gentoo-user] systemd

[Canek Peláez Valdés]

On Tue, Aug 23, 2011 at 4:19 PM, Alan McKinnon <alan.mckinnon@gmail.com> wrote:
> On Tue 23 August 2011 15:50:24 Canek Peláez Valdés did opine thusly:
>> On Tue, Aug 23, 2011 at 3:43 PM, Alan McKinnon
> <alan.mckinnon@gmail.com> wrote:
>> > On Tue 23 August 2011 15:06:25 Canek Peláez Valdés did opine
> thusly:
>> >> > Now if it had similarities to say hal, I would instantly
>> >> > understand. But dbus is good and useful in all the ways
>> >> > that
>> >> > hal isn't.
>> >>
>> >> Wasn't. HAL is dead. From
>> >> http://www.freedesktop.org/wiki/Software/hal
>> >
>> > Sadly, HAL is not yet dead. It lives still.
>> >
>> > It lives on the production database server I just happen to be
>> > rebooting as I type this (another story for another time) and
>> > will continue to live here for a very very long time indeed.
>> >
>> > Dale can confirm this. Dale will swear in a court of law with
>> > hand on bible than hal lives on in zombie form, infesting all
>> > the matter of his house and computers, infecting them with
>> > their undead zombieness.
>> >
>> > Ye gods, it's been a long hard day....
>>
>> I remember getting rid of HAL in one weekend, from all my computers.
>> It was a long weekend, but it was not as bad as getting rid of Qt
>> from all the computers in my office some years ago.
>
> Come to my work place, I have the perfect task for you:
>
> to excise perl-5.8.0 from all the many machines it's on, plus the
> atrocious in-house coding using it that no-one left understands, and
> all running on hardware that no-one can replace.

Been there, done that. One of the reasons I got back to school to get
my Computer Science PhD.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

[gentoo-user] Re: systemd

[walt]

On 08/23/2011 01:19 PM, Alan McKinnon wrote:

> all running on hardware that no-one can replace.

Okay, I give.  Why not?

Re: [gentoo-user] Re: systemd

[Alan McKinnon]

On Wed 24 August 2011 14:19:56 walt did opine thusly:
> On 08/23/2011 01:19 PM, Alan McKinnon wrote:
> > all running on hardware that no-one can replace.
> 
> Okay, I give.  Why not?

Dell 23xx and 24xx generation hardware


-- 
alan dot mckinnon at gmail dot com

Re: [gentoo-user] systemd

[Sebastian Beßler]

[-- Attachment #1: Type: text/plain, Size: 503 bytes --]

Am 23.08.2011 21:43, schrieb Alan McKinnon:

> Sadly, HAL is not yet dead. It lives still.
> 
> It lives on the production database server I just happen to be 
> rebooting as I type this (another story for another time) and will 
> continue to live here for a very very long time indeed.

WHY is HAL installed on a database server?
I still see desktop systems with HAL, last on an newish kubuntu of a
friend, but on a server? For what is HAL needed there?

Greetings

Sebastian Beßler


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 900 bytes --]

Re: [gentoo-user] systemd

[Alan McKinnon]

On Tue 23 August 2011 22:16:30 Sebastian Beßler did opine thusly:
> Am 23.08.2011 21:43, schrieb Alan McKinnon:
> > Sadly, HAL is not yet dead. It lives still.
> > 
> > It lives on the production database server I just happen to be
> > rebooting as I type this (another story for another time) and
> > will continue to live here for a very very long time indeed.
> 
> WHY is HAL installed on a database server?
> I still see desktop systems with HAL, last on an newish kubuntu of a
> friend, but on a server? For what is HAL needed there?

I wish I knew why. The fellow that did the install might know. I'm 
betting it's because he clicked yes, yes, yes, yes, ok on the RHEL 
install CD dialogs.

I can't fix it without running afoul of the Change Management process, 
and today's emergency reboot didn't leave me any time to poke around 
and determine the effect of removing hal.

This is how life in corporate IT works....

-- 
alan dot mckinnon at gmail dot com

Re: [gentoo-user] systemd

[kashani]

On 8/23/2011 1:43 PM, Alan McKinnon wrote:
>
> I can't fix it without running afoul of the Change Management process,
> and today's emergency reboot didn't leave me any time to poke around
> and determine the effect of removing hal.
>
> This is how life in corporate IT works....
>

	I hate Corp CM and it's one of the reasons I stay in startups. It's job 
is to slow normal change down so much so that every change becomes an 
emergency.

	However next time I have to deal with one I am shoving mathematical 
proof of "there is no rollback in systems" down there throats. 
http://www.iu.hio.no/~mark/papers/totalfield.pdf

For those that aren't ginormous systems nerds this bit sums it up nicely.

	"There is a deeper issue with roll-back in partial systems. If a system 
is in contact with another system, e.g. receiving data, or if we have 
partitioned a system into loosely coupled pieces only one of which is 
being changed, then the other system becomes a part of the total system 
and we must write a hypothetical journal for the entire system in order 
to achieve a consistent rollback."

kashani

Re: [gentoo-user] systemd

[Alan McKinnon]

On Tue 23 August 2011 14:10:57 kashani did opine thusly:
> On 8/23/2011 1:43 PM, Alan McKinnon wrote:
> > I can't fix it without running afoul of the Change Management
> > process, and today's emergency reboot didn't leave me any time
> > to poke around and determine the effect of removing hal.
> > 
> > This is how life in corporate IT works....
> 
> 	I hate Corp CM and it's one of the reasons I stay in startups. 
It's
> job is to slow normal change down so much so that every change
> becomes an emergency.
> 
> 	However next time I have to deal with one I am shoving 
mathematical
> proof of "there is no rollback in systems" down there throats.
> http://www.iu.hio.no/~mark/papers/totalfield.pdf

Haven't read the pdf yet, but I just have to share this joke.

Tonight's CM was an unscheduled emergency reboot. This gave me 
opportunity to do something I've been dying to do for ages, enter 
this:

Install plan: reboot server
Test plan:    ping server
Backout plan: unreboot server                 <====== :-)

On the whole our CM process is sane. The manager knows how 
infrastructure works:

If that undersea optical link goes down, I'm fixing it right now and 
to hell with the paperwork and process.

Contrast with my gf's job at the bank. That one truly is a case where 
to change anything, she has to invent imaginary catastrophic 
emergencies. More often than not, she causes them in undetectable ways 
just to get her job done.


> 
> For those that aren't ginormous systems nerds this bit sums it up
> nicely.
> 
> 	"There is a deeper issue with roll-back in partial systems. If a
> system is in contact with another system, e.g. receiving data, or
> if we have partitioned a system into loosely coupled pieces only
> one of which is being changed, then the other system becomes a part
> of the total system and we must write a hypothetical journal for
> the entire system in order to achieve a consistent rollback."
> 
> kashani
-- 
alan dot mckinnon at gmail dot com

Re: [gentoo-user] systemd

[Dale]

Alan McKinnon wrote:
> On Tue 23 August 2011 15:06:25 Canek Peláez Valdés did opine thusly:
>    
>>> Now if it had similarities to say hal, I would instantly
>>> understand. But dbus is good and useful in all the ways that
>>> hal isn't.
>>>        
>> Wasn't. HAL is dead. From
>> http://www.freedesktop.org/wiki/Software/hal
>>      
> Sadly, HAL is not yet dead. It lives still.
>
> It lives on the production database server I just happen to be
> rebooting as I type this (another story for another time) and will
> continue to live here for a very very long time indeed.
>
> Dale can confirm this. Dale will swear in a court of law with hand on
> bible than hal lives on in zombie form, infesting all the matter of
> his house and computers, infecting them with their undead zombieness.
>
> Ye gods, it's been a long hard day....
>
>    

Not here.  I shot hal with a silver bullet and drove a stake into it a 
long time ago.  If that thing even twitches, I'll go Navy Seals on it.  
O_O  Man I love the 2nd amendment we have.  ;-)  Even the NSA wouldn't 
be able to bring that back.

Dale

:-)  :-)

Re: [gentoo-user] systemd

[Joost Roeleveld]

On Tuesday, August 23, 2011 11:15:01 PM Dale wrote:
> Alan McKinnon wrote:
> > On Tue 23 August 2011 15:06:25 Canek Peláez Valdés did opine thusly:
> >>> Now if it had similarities to say hal, I would instantly
> >>> understand. But dbus is good and useful in all the ways that
> >>> hal isn't.
> >> 
> >> Wasn't. HAL is dead. From
> >> http://www.freedesktop.org/wiki/Software/hal
> > 
> > Sadly, HAL is not yet dead. It lives still.
> > 
> > It lives on the production database server I just happen to be
> > rebooting as I type this (another story for another time) and will
> > continue to live here for a very very long time indeed.
> > 
> > Dale can confirm this. Dale will swear in a court of law with hand on
> > bible than hal lives on in zombie form, infesting all the matter of
> > his house and computers, infecting them with their undead zombieness.
> > 
> > Ye gods, it's been a long hard day....
> 
> Not here.  I shot hal with a silver bullet and drove a stake into it a
> long time ago.  If that thing even twitches, I'll go Navy Seals on it.
> O_O  Man I love the 2nd amendment we have.  ;-)  Even the NSA wouldn't
> be able to bring that back.

I'm missing the exorcism in there, it's probably still floating around as a 
non-corporeal lifeform causing all kinds of strange issues....

--
Joost

Re: [gentoo-user] systemd

[Alex Schuster]

Alan McKinnon writes:

> On Tue 23 August 2011 18:17:17 Stroller did opine thusly:
> > On 23 August 2011, at 07:27, Joost Roeleveld wrote:
[...]
> > > And I don't really see the point of D-BUS on a server either.
> > > All the services that need to talk to each other already have
> > > working communication paths.
> > 
> > Reading that blog entry I found discouraging the idea that dbus
> > might be required on my servers in the future, if systemd becomes
> > popular with distros.
> 
> What's your objection to dbus? It gives you a standard message bus, is
> small, light, consumes minimal resources and provides a nice standard
> way to do IPC. Probably easier than reinventing the wheel with named
> pipes and other bits over and over.

Except for me. dbus-daemon often uses 10-20% of my CPU according to top. And 
this morning, it was using about 750M of memory. Which is less than kwin's 
and Kontact's usage, but still.
But I think the problem is on my side, I run KDE4 with only 8G of memory, no 
wonder I need 1.7G of swap right now.
</rant>

	Wonko

Re: [gentoo-user] systemd

[Michael Schreckenbauer]

Hi,

Am Dienstag, 30. August 2011, 13:56:44 schrieb Alex Schuster:
> Alan McKinnon writes:
> > On Tue 23 August 2011 18:17:17 Stroller did opine thusly:
> > > On 23 August 2011, at 07:27, Joost Roeleveld wrote:
> [...]
> 
> > > > And I don't really see the point of D-BUS on a server either.
> > > > All the services that need to talk to each other already have
> > > > working communication paths.
> > > 
> > > Reading that blog entry I found discouraging the idea that dbus
> > > might be required on my servers in the future, if systemd becomes
> > > popular with distros.
> > 
> > What's your objection to dbus? It gives you a standard message bus, is
> > small, light, consumes minimal resources and provides a nice standard
> > way to do IPC. Probably easier than reinventing the wheel with named
> > pipes and other bits over and over.
> 
> Except for me. dbus-daemon often uses 10-20% of my CPU according to top.

Mine idles most of the time, no CPU is used. My computer is running for ~6h 
now, dbus-daemon used less than 1.5s CPU time.

> And
> this morning, it was using about 750M of memory. Which is less than kwin's
> and Kontact's usage, but still.

Strange. Mine uses only ~20MB.

> But I think the problem is on my side, I run KDE4 with only 8G of memory, no
> wonder I need 1.7G of swap right now.
> </rant>

I have only 4GB of memory, run kde4, swap is not used at all most of the time. 
There are still ~512MB free with ~1,3GB cached currently.
I do have programs running :) firefox with some tabs, kdevelop with a project 
(~100.000 LOC), kmail, LibreOffice and 3 konsoles, each with some tabs open.
I know, I am of no help at all, but I really wonder, why your numbers differ so 
significantly from mine.

> 	Wonko

Michael

Re: [gentoo-user] systemd

[Canek Peláez Valdés]

On Tue, Aug 30, 2011 at 8:13 AM, Michael Schreckenbauer <grimlog@gmx.de> wrote:
> Hi,
>
> Am Dienstag, 30. August 2011, 13:56:44 schrieb Alex Schuster:
>> Alan McKinnon writes:
>> > On Tue 23 August 2011 18:17:17 Stroller did opine thusly:
>> > > On 23 August 2011, at 07:27, Joost Roeleveld wrote:
>> [...]
>>
>> > > > And I don't really see the point of D-BUS on a server either.
>> > > > All the services that need to talk to each other already have
>> > > > working communication paths.
>> > >
>> > > Reading that blog entry I found discouraging the idea that dbus
>> > > might be required on my servers in the future, if systemd becomes
>> > > popular with distros.
>> >
>> > What's your objection to dbus? It gives you a standard message bus, is
>> > small, light, consumes minimal resources and provides a nice standard
>> > way to do IPC. Probably easier than reinventing the wheel with named
>> > pipes and other bits over and over.
>>
>> Except for me. dbus-daemon often uses 10-20% of my CPU according to top.
>
> Mine idles most of the time, no CPU is used. My computer is running for ~6h
> now, dbus-daemon used less than 1.5s CPU time.

Same here:

canek@negra ~ $ uptime
 11:01:52 up 5 days, 20:13,  1 user,  load average: 0.06, 0.43, 0.50

(It's a laptop that I usually suspend at night).

>> And
>> this morning, it was using about 750M of memory. Which is less than kwin's
>> and Kontact's usage, but still.
>
> Strange. Mine uses only ~20MB.

Same here:

top - 11:02:40 up 5 days, 20:14,  1 user,  load average: 0.12, 0.39, 0.49
Tasks: 163 total,   1 running, 158 sleeping,   0 stopped,   4 zombie
Cpu(s):  7.3%us,  2.0%sy,  0.0%ni, 90.2%id,  0.0%wa,  0.0%hi,  0.5%si,  0.0%st
Mem:   3891064k total,  3214892k used,   676172k free,    36072k buffers
Swap:  4192960k total,   708604k used,  3484356k free,   863416k cached

PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
631 messageb  20   0 20548 2404 1040 S    0  0.1   1:47.94 dbus-daemon


>> But I think the problem is on my side, I run KDE4 with only 8G of memory, no
>> wonder I need 1.7G of swap right now.
>> </rant>
>
> I have only 4GB of memory, run kde4, swap is not used at all most of the time.
> There are still ~512MB free with ~1,3GB cached currently.
> I do have programs running :) firefox with some tabs, kdevelop with a project
> (~100.000 LOC), kmail, LibreOffice and 3 konsoles, each with some tabs open.
> I know, I am of no help at all, but I really wonder, why your numbers differ so
> significantly from mine.

Kinda similar here: GNOME 3.0, Emacs with several LaTeX articles,
Evince, Evolution, Rhythmbox, Chromium with like 20 tabs (my 4 zombie
processes are Chromium tabs), and the heaviest of all, Inkscape with 6
different SVG pictures.

There is something really wrong with Alex D-Bus; but I don't think
it's the bus. Probably some program is spamming the bus, making it use
that much memory, but I don't know for sure.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] systemd

[Alex Schuster]

Michael Schreckenbauer writes:

> Hi,
> 
> Am Dienstag, 30. August 2011, 13:56:44 schrieb Alex Schuster:
> > Alan McKinnon writes:
[...]
> > > What's your objection to dbus? It gives you a standard message
> > > bus, is small, light, consumes minimal resources and provides a
> > > nice standard way to do IPC. Probably easier than reinventing the
> > > wheel with named pipes and other bits over and over.
> > 
> > Except for me. dbus-daemon often uses 10-20% of my CPU according to
> > top.
> Mine idles most of the time, no CPU is used. My computer is running for
> ~6h now, dbus-daemon used less than 1.5s CPU time.

After a relogin, this is also true here at this moment, dbus-daemon uses 
about 4% of one of my two cores. After 26 days of uptime, its total CPU time 
is 1380 minutes, Followed by udisks-daemon with 990 minutes, and then mysqld 
with 25 minutes.

> > And
> > this morning, it was using about 750M of memory. Which is less than
> > kwin's and Kontact's usage, but still.
> 
> Strange. Mine uses only ~20MB.

Right now I have three dbus-daemon processes (one owned by messagebus, two 
owned by my user), with a total of 4.5M only. The excessive usage of 750M (I 
noticed this for the first time) probably was a memory leak. Like with KWin, 
where it always happens after some days of being logged in.

> > But I think the problem is on my side, I run KDE4 with only 8G of
> > memory, no wonder I need 1.7G of swap right now.
> > </rant>
> 
> I have only 4GB of memory, run kde4, swap is not used at all most of the
> time. There are still ~512MB free with ~1,3GB cached currently.
> I do have programs running :) firefox with some tabs, kdevelop with a
> project (~100.000 LOC), kmail, LibreOffice and 3 konsoles, each with some
> tabs open. I know, I am of no help at all, but I really wonder, why your
> numbers differ so significantly from mine.

I run some more applications. 9 Konsole tabs, two Dolphins, a Konqueror as 
file manager, Amarok, TV-Browser, Kontact, Chromium with 15 tabs, KMyMoney. 
That's what comes up after login, after a while of being logged in more 
stuff is running. Yesterday I had a Windows running in vmplayer, that may 
use 512M.
Right now, 4.7G of memory is needed (free -m, -/+ buffers/cache entry).

[Later]

Whoops, forgot to actually send this mail. Ten hours later I have another 
dbus-daemon process, owned by root, but memory their usage is the same. 
5250M of RAM are needed now altogether.

	Wonko

Re: [gentoo-user] systemd

[Michael Mol]

On Tue, Aug 30, 2011 at 7:01 PM, Alex Schuster <wonko@wonkology.org> wrote:
> Michael Schreckenbauer writes:
>
>> Hi,
>>
>> Am Dienstag, 30. August 2011, 13:56:44 schrieb Alex Schuster:
>> > Alan McKinnon writes:
> [...]
>> > > What's your objection to dbus? It gives you a standard message
>> > > bus, is small, light, consumes minimal resources and provides a
>> > > nice standard way to do IPC. Probably easier than reinventing the
>> > > wheel with named pipes and other bits over and over.
>> >
>> > Except for me. dbus-daemon often uses 10-20% of my CPU according to
>> > top.
>> Mine idles most of the time, no CPU is used. My computer is running for
>> ~6h now, dbus-daemon used less than 1.5s CPU time.
>
> After a relogin, this is also true here at this moment, dbus-daemon uses
> about 4% of one of my two cores. After 26 days of uptime, its total CPU time
> is 1380 minutes, Followed by udisks-daemon with 990 minutes, and then mysqld
> with 25 minutes.
>
>> > And
>> > this morning, it was using about 750M of memory. Which is less than
>> > kwin's and Kontact's usage, but still.
>>
>> Strange. Mine uses only ~20MB.
>
> Right now I have three dbus-daemon processes (one owned by messagebus, two
> owned by my user), with a total of 4.5M only. The excessive usage of 750M (I
> noticed this for the first time) probably was a memory leak. Like with KWin,
> where it always happens after some days of being logged in.
>
>> > But I think the problem is on my side, I run KDE4 with only 8G of
>> > memory, no wonder I need 1.7G of swap right now.
>> > </rant>
>>
>> I have only 4GB of memory, run kde4, swap is not used at all most of the
>> time. There are still ~512MB free with ~1,3GB cached currently.
>> I do have programs running :) firefox with some tabs, kdevelop with a
>> project (~100.000 LOC), kmail, LibreOffice and 3 konsoles, each with some
>> tabs open. I know, I am of no help at all, but I really wonder, why your
>> numbers differ so significantly from mine.
>
> I run some more applications. 9 Konsole tabs, two Dolphins, a Konqueror as
> file manager, Amarok, TV-Browser, Kontact, Chromium with 15 tabs, KMyMoney.
> That's what comes up after login, after a while of being logged in more
> stuff is running. Yesterday I had a Windows running in vmplayer, that may
> use 512M.
> Right now, 4.7G of memory is needed (free -m, -/+ buffers/cache entry).
>
> [Later]
>
> Whoops, forgot to actually send this mail. Ten hours later I have another
> dbus-daemon process, owned by root, but memory their usage is the same.
> 5250M of RAM are needed now altogether.

Very, very weird. You all seem to have some weird issues with
dbus-daemon that I don't have.

(times retrieved with ps axS, memory consumption retrieved with htop)

On my 77-day uptime server runing Debian 5, dbus's total time is 0:00.
(That's with ps axS) virtual memory of 21M, resident of 900K.

On my gentoo desktop, 8 days' uptime, I show two dbus-daemon
processes, both with 0:00. Both with virtual of 19M. One with resident
of 984K, one with resident of 808K.

On my router, Debian 6, with 4 days' uptime, ps axS shows 0:00 for
consumed time. 23M virtual, 968K resident.


-- 
:wq

Re: [gentoo-user] systemd

[Alan McKinnon]

On Tue, 30 Aug 2011 14:13:53 +0200
Michael Schreckenbauer <grimlog@gmx.de> wrote:

> Hi,
> 
> Am Dienstag, 30. August 2011, 13:56:44 schrieb Alex Schuster:
> > Alan McKinnon writes:
> > > On Tue 23 August 2011 18:17:17 Stroller did opine thusly:
> > > > On 23 August 2011, at 07:27, Joost Roeleveld wrote:
> > [...]
> > 
> > > > > And I don't really see the point of D-BUS on a server either.
> > > > > All the services that need to talk to each other already have
> > > > > working communication paths.
> > > > 
> > > > Reading that blog entry I found discouraging the idea that dbus
> > > > might be required on my servers in the future, if systemd
> > > > becomes popular with distros.
> > > 
> > > What's your objection to dbus? It gives you a standard message
> > > bus, is small, light, consumes minimal resources and provides a
> > > nice standard way to do IPC. Probably easier than reinventing the
> > > wheel with named pipes and other bits over and over.
> > 
> > Except for me. dbus-daemon often uses 10-20% of my CPU according to
> > top.
> 
> Mine idles most of the time, no CPU is used. My computer is running
> for ~6h now, dbus-daemon used less than 1.5s CPU time.
> 
> > And
> > this morning, it was using about 750M of memory. Which is less than
> > kwin's and Kontact's usage, but still.
> 
> Strange. Mine uses only ~20MB.
> 
> > But I think the problem is on my side, I run KDE4 with only 8G of
> > memory, no wonder I need 1.7G of swap right now.
> > </rant>
> 
> I have only 4GB of memory, run kde4, swap is not used at all most of
> the time. There are still ~512MB free with ~1,3GB cached currently.
> I do have programs running :) firefox with some tabs, kdevelop with a
> project (~100.000 LOC), kmail, LibreOffice and 3 konsoles, each with
> some tabs open. I know, I am of no help at all, but I really wonder,
> why your numbers differ so significantly from mine.

He probably has the same problem as I - something badly wrong in the
roll-your-own config.

I had a 4G Dell laptop where KDE would start and instantly consume at
least 1.5G just for it's various bits. Akonadi, Nepomuk, Virtuoso were
the usual culprits. Oddly, dbus would often rise to 700M (!). And
forget about actually emerging something - the first sniff that gcc was
running and the machine would thrash like mad and 4G swap would fill up
in no time at all. Less sweap wasn't an option - the battery is dud so
I needed hibernate.

Sadly (or not, depending on your viewpoint), that machine died on
Monday morning - suspect graphics card. I can't complain - it ran flat
out 24/7/365 and I treated it like one of the servers that I could
carry around. So I can't even troubleshoot what I configured how to
make performance behave like it did.

Happily, there was a nice pretty lady from Samsung in the office 3
months ago wanting to sell the 900X Macbook Air knock-off into the
company. The IT manager didn't know what to do with the demo she left
behind so I knicked it for myself (sans paperwork of course. Makes it
easier to prolong how long it takes to test properly) and it's running
Ubuntu. Memory issues are a thing of the past and everything behaves
just like it should. Even <gasp> flash.


-- 
Alan McKinnnon
alan.mckinnon@gmail.com

Re: [gentoo-user] systemd

[Canek Peláez Valdés]

On Thu, Sep 1, 2011 at 5:48 PM, Alan McKinnon <alan.mckinnon@gmail.com> wrote:
> On Tue, 30 Aug 2011 14:13:53 +0200
> Michael Schreckenbauer <grimlog@gmx.de> wrote:
>
>> Hi,
>>
>> Am Dienstag, 30. August 2011, 13:56:44 schrieb Alex Schuster:
>> > Alan McKinnon writes:
>> > > On Tue 23 August 2011 18:17:17 Stroller did opine thusly:
>> > > > On 23 August 2011, at 07:27, Joost Roeleveld wrote:
>> > [...]
>> >
>> > > > > And I don't really see the point of D-BUS on a server either.
>> > > > > All the services that need to talk to each other already have
>> > > > > working communication paths.
>> > > >
>> > > > Reading that blog entry I found discouraging the idea that dbus
>> > > > might be required on my servers in the future, if systemd
>> > > > becomes popular with distros.
>> > >
>> > > What's your objection to dbus? It gives you a standard message
>> > > bus, is small, light, consumes minimal resources and provides a
>> > > nice standard way to do IPC. Probably easier than reinventing the
>> > > wheel with named pipes and other bits over and over.
>> >
>> > Except for me. dbus-daemon often uses 10-20% of my CPU according to
>> > top.
>>
>> Mine idles most of the time, no CPU is used. My computer is running
>> for ~6h now, dbus-daemon used less than 1.5s CPU time.
>>
>> > And
>> > this morning, it was using about 750M of memory. Which is less than
>> > kwin's and Kontact's usage, but still.
>>
>> Strange. Mine uses only ~20MB.
>>
>> > But I think the problem is on my side, I run KDE4 with only 8G of
>> > memory, no wonder I need 1.7G of swap right now.
>> > </rant>
>>
>> I have only 4GB of memory, run kde4, swap is not used at all most of
>> the time. There are still ~512MB free with ~1,3GB cached currently.
>> I do have programs running :) firefox with some tabs, kdevelop with a
>> project (~100.000 LOC), kmail, LibreOffice and 3 konsoles, each with
>> some tabs open. I know, I am of no help at all, but I really wonder,
>> why your numbers differ so significantly from mine.
>
> He probably has the same problem as I - something badly wrong in the
> roll-your-own config.
>
> I had a 4G Dell laptop where KDE would start and instantly consume at
> least 1.5G just for it's various bits. Akonadi, Nepomuk, Virtuoso were
> the usual culprits. Oddly, dbus would often rise to 700M (!). And
> forget about actually emerging something - the first sniff that gcc was
> running and the machine would thrash like mad and 4G swap would fill up
> in no time at all. Less sweap wasn't an option - the battery is dud so
> I needed hibernate.
>
> Sadly (or not, depending on your viewpoint), that machine died on
> Monday morning - suspect graphics card. I can't complain - it ran flat
> out 24/7/365 and I treated it like one of the servers that I could
> carry around. So I can't even troubleshoot what I configured how to
> make performance behave like it did.
>
> Happily, there was a nice pretty lady from Samsung in the office 3
> months ago wanting to sell the 900X Macbook Air knock-off into the
> company. The IT manager didn't know what to do with the demo she left
> behind so I knicked it for myself (sans paperwork of course. Makes it
> easier to prolong how long it takes to test properly) and it's running
> Ubuntu. Memory issues are a thing of the past and everything behaves
> just like it should. Even <gasp> flash.

That saddens me a little. Of all the friends and coworkers I have, I'm
the only one left using Gentoo. All of them switched to Fedora, or
Ubuntu, or OpenSuSE.

My three years old laptop runs up-to-date Gentoo with systemd, plus
the GNOME overlay so I can use GNOME 3. Absolutely everything works
with the laptop: the Wi-Fi with NetworkManager, the sound with
PulseAudio, suspend/resume, all the Fn keys, wathever. Even the
ridiculous touchstrip fingerprint sensor, though I don't really use
it.

Even more: every damn GUI program does what it should, so even though
I'm able to configure and use everything with the command line, I
don't have to, because I can use the pretty graphic programs... if at
all, 'cause nowadays everything usually "just works". And the same
it's true for all my other machines.

But I'm also aware that all of this is possible because I have the
knowledge (and the patience) to detect and fix any problem that I may
encounter when updating my machine. I don't need a robust QA from my
distribution, but I'm the exception. And this is the kind of QA
problems that make people to replace Gentoo for Ubuntu, or Fedora, or
whatever.

I whish I knew how to solve this so everybody could use Gentoo, but I
don't. I don't think I will ever use any other distro (although I've
toyed with the idea of trying exherbo), but I cannot in good
conscience recommend it to "normal" users, unless I'm willing to be
their tech support forever.

I love Gentoo, been using it since 2004, but I'm the first one to
admit it's not for everyone.

And that saddens me a little.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] systemd

[Alan McKinnon]

On Thu, 1 Sep 2011 18:32:11 -0400
Canek Peláez Valdés <caneko@gmail.com> wrote:

> > Happily, there was a nice pretty lady from Samsung in the office 3
> > months ago wanting to sell the 900X Macbook Air knock-off into the
> > company. The IT manager didn't know what to do with the demo she
> > left behind so I knicked it for myself (sans paperwork of course.
> > Makes it easier to prolong how long it takes to test properly) and
> > it's running Ubuntu. Memory issues are a thing of the past and
> > everything behaves just like it should. Even <gasp> flash.  
> 
> That saddens me a little. Of all the friends and coworkers I have, I'm
> the only one left using Gentoo. All of them switched to Fedora, or
> Ubuntu, or OpenSuSE.

I completely understand how you feel. But, I'm enjoying this break from
Gentoo. Note I said "break", not "leave behind".

It will take 3 to 6 weeks for me to make up my mind what monster Dell I
want next, assault Purchasing to make them approve it then have it
built in Ireland and shipped to ZA.

That's about enough time I think to run into the Ubuntu "you will do it
our way with the deps we want you to have" philosophy enough times to
drive me back to gentoo. But the next machine will not have KDE on it -
all this RAM nonsense started by trying to get Akonadi to work, for very
loose definitions of "work", such as "show me my mail sometime today"

e17 beckons.


-- 
Alan McKinnnon
alan.mckinnon@gmail.com

Re: [gentoo-user] systemd

[Mick]

[-- Attachment #1: Type: Text/Plain, Size: 2501 bytes --]

On Friday 02 Sep 2011 00:27:58 Alan McKinnon wrote:
> On Thu, 1 Sep 2011 18:32:11 -0400
> 
> Canek Peláez Valdés <caneko@gmail.com> wrote:
> > > Happily, there was a nice pretty lady from Samsung in the office 3
> > > months ago wanting to sell the 900X Macbook Air knock-off into the
> > > company. The IT manager didn't know what to do with the demo she
> > > left behind so I knicked it for myself (sans paperwork of course.
> > > Makes it easier to prolong how long it takes to test properly) and
> > > it's running Ubuntu. Memory issues are a thing of the past and
> > > everything behaves just like it should. Even <gasp> flash.
> > 
> > That saddens me a little. Of all the friends and coworkers I have, I'm
> > the only one left using Gentoo. All of them switched to Fedora, or
> > Ubuntu, or OpenSuSE.
> 
> I completely understand how you feel. But, I'm enjoying this break from
> Gentoo. Note I said "break", not "leave behind".
> 
> It will take 3 to 6 weeks for me to make up my mind what monster Dell I
> want next, assault Purchasing to make them approve it then have it
> built in Ireland and shipped to ZA.

I must have words with Dell!!!

Mine was built in China and shipped from China ... twice, because the 
horrendous courier they use did not deliver it.


> That's about enough time I think to run into the Ubuntu "you will do it
> our way with the deps we want you to have" philosophy enough times to
> drive me back to gentoo. 

I have a laptop which is only used occasionally and (K)Ubuntu is a good use 
case for it.  Keeping it up to date does not take long with more or less 
vanilla Kubuntu settings and because the user needs are not particularly 
demanding Ubuntu has filled the OS role admirably.  Doing this with Gentoo 
would require much more attention and time from me.  However, I would not use 
Ubuntu for my needs.  Very much like Alan says, there is an annoying 
underlying feeling of "all your OS belongs to us".

> But the next machine will not have KDE on it -
> all this RAM nonsense started by trying to get Akonadi to work, for very
> loose definitions of "work", such as "show me my mail sometime today"
> 
> e17 beckons.

Let us know which mail client you end up with.  Although I now use e17 and 
before that I used a rather edited Fluxbox (I avoid bloatware DEs if I can 
help it) I am still using kmail because attempts to adopt other mail clients 
ended up in disappointment.
-- 
Regards,
Mick

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] systemd

[Alan McKinnon]

On Fri, 2 Sep 2011 08:33:56 +0100
Mick <michaelkintzios@gmail.com> wrote:

> > But the next machine will not have KDE on it -
> > all this RAM nonsense started by trying to get Akonadi to work, for
> > very loose definitions of "work", such as "show me my mail sometime
> > today"
> > 
> > e17 beckons.  
> 
> Let us know which mail client you end up with.  Although I now use
> e17 and before that I used a rather edited Fluxbox (I avoid bloatware
> DEs if I can help it) I am still using kmail because attempts to
> adopt other mail clients ended up in disappointment.

At the moment I'm using claws and it seems quite fine. It has it's
funnies (mostly just different behaviour from kmail actually) so I need
to rewire my brain a bit.

It wants to wait for me to manually expunge folders, and I haven't quite
figured out how to always get it to use the correct .sig (a few times
my work .sig got used on mailing lists where I always use gmail), but
claws has these two amazing features that make it a killer app for me:

- it will not ever compose html mail but does a decent job of display it
when received
- it actually sends, receives and stores mail (akonadi users will
  appreciate how a mailer can get to make that insanely more complex
  than it needs to be)



-- 
Alan McKinnnon
alan.mckinnon@gmail.com

Re: [gentoo-user] systemd

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 1364 bytes --]

On Fri, 2 Sep 2011 17:28:21 +0200, Alan McKinnon wrote:

> At the moment I'm using claws and it seems quite fine. It has it's
> funnies (mostly just different behaviour from kmail actually) so I need
> to rewire my brain a bit.

I *really* like Claws Mail, I never liked KMail. In fact, removing KDEPIM
is the best way to improve KDE.

> It wants to wait for me to manually expunge folders, and I haven't quite

Account Preferences>Advanced>Move deleted mail to trash and expunge
immediately

> figured out how to always get it to use the correct .sig (a few times
> my work .sig got used on mailing lists where I always use gmail), but
> claws has these two amazing features that make it a killer app for me:

Set up a different (SMTP-only if necessary) account for each address and
specify the sig. You can set the default account to use for a folder,
which means you always use the correct address on a mailing list. If no
default account is set, Claws will default to using the account that the
mail you are replying to was sent to.

> - it actually sends, receives and stores mail (akonadi users will
>   appreciate how a mailer can get to make that insanely more complex
>   than it needs to be)

I'm so glad I ditched KMail before akonadi and friends came along.


-- 
Neil Bothwick

Why is the word abbreviation so long?

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] systemd

[Mick]

[-- Attachment #1: Type: Text/Plain, Size: 1233 bytes --]

On Saturday 03 Sep 2011 01:05:33 Neil Bothwick wrote:
> On Fri, 2 Sep 2011 17:28:21 +0200, Alan McKinnon wrote:
> > At the moment I'm using claws and it seems quite fine. It has it's
> > funnies (mostly just different behaviour from kmail actually) so I need
> > to rewire my brain a bit.
> 
> I *really* like Claws Mail, I never liked KMail. In fact, removing KDEPIM
> is the best way to improve KDE.

Interesting!  With me it has been the opposite.  I didn't make a list of the 
things that really bothered me with Claws, but after a dozen use cases of me 
saying "why can't Claws behave like Kmail?" I gave up and went back to Kmail 
...


> I'm so glad I ditched KMail before akonadi and friends came along.

Akonadi and all the bloatware that came with KDE4 has been a major 
disappointment and cause of annoyance for me.  Thankfully, after some initial 
teething problems with sqlite (I don't use mysql on my laptop) akonadi has not 
given me any trouble - but I am worried about what is coming when reading 
Alan's experience!

Have I understood this correctly that Kmail2 will no longer store messages in 
conventional maildir files and it will all be stored in database tables?
-- 
Regards,
Mick

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] systemd

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 594 bytes --]

On Sat, 3 Sep 2011 08:56:15 +0100, Mick wrote:

> > I *really* like Claws Mail, I never liked KMail. In fact, removing
> > KDEPIM is the best way to improve KDE.  
> 
> Interesting!  With me it has been the opposite.  I didn't make a list
> of the things that really bothered me with Claws, but after a dozen use
> cases of me saying "why can't Claws behave like Kmail?" I gave up and
> went back to Kmail

Well, if you want a program that behaves just like KMail, KMail is
probably the best choice...


-- 
Neil Bothwick

ALZHEIMER.COM found . . . Out of . . . something . .

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] systemd

[Alan McKinnon]

On Sat, 3 Sep 2011 08:56:15 +0100
Mick <michaelkintzios@gmail.com> wrote:

> > I'm so glad I ditched KMail before akonadi and friends came along.  
> 
> Akonadi and all the bloatware that came with KDE4 has been a major 
> disappointment and cause of annoyance for me.  Thankfully, after some
> initial teething problems with sqlite (I don't use mysql on my
> laptop) akonadi has not given me any trouble - but I am worried about
> what is coming when reading Alan's experience!

To be fair my problems started after migrating a setup that had come
along since early-KDE3 days and went through many pre-release versions
of akonadi. Almost everyone I found with major problems was in the same
boat while people who did fresh installs on SuSE and Fedora were fine.
Observe that clkean new install is not a valid use-case for a gentoo
user

> Have I understood this correctly that Kmail2 will no longer store
> messages in conventional maildir files and it will all be stored in
> database tables?

No, akonadi functions as a giant cache. The backing store for mail is
still mbox, MailDir, IMAP or whatever and used as normal. Akonadi
stores meta data about the mail (maildir index files are not in files
for example, they are in Akonadi). Contacts and calendars are still on
disk with a cached copy in Akonadi. 

Full text searches of mail bodies is stored in Nepomuk.


-- 
Alan McKinnnon
alan.mckinnon@gmail.com

Re: [gentoo-user] systemd

[Alan McKinnon]

On Sat, 3 Sep 2011 01:05:33 +0100
Neil Bothwick <neil@digimed.co.uk> wrote:

> On Fri, 2 Sep 2011 17:28:21 +0200, Alan McKinnon wrote:
> 
> > At the moment I'm using claws and it seems quite fine. It has it's
> > funnies (mostly just different behaviour from kmail actually) so I
> > need to rewire my brain a bit.
> 
> I *really* like Claws Mail, I never liked KMail. In fact, removing
> KDEPIM is the best way to improve KDE.

:-)
 
> > It wants to wait for me to manually expunge folders, and I haven't
> > quite
> 
> Account Preferences>Advanced>Move deleted mail to trash and expunge
> immediately

When I'd got around to reading the FAQ, I found this very tidbit of
useful info but thanks

> > figured out how to always get it to use the correct .sig (a few
> > times my work .sig got used on mailing lists where I always use
> > gmail), but claws has these two amazing features that make it a
> > killer app for me:
> 
> Set up a different (SMTP-only if necessary) account for each address
> and specify the sig. You can set the default account to use for a
> folder, which means you always use the correct address on a mailing
> list. If no default account is set, Claws will default to using the
> account that the mail you are replying to was sent to.

I like to use local IMAP folder and route all mail into those,
organized by work and other. I found that I can set prefs for a top
folders and all children folders which handled this one nicely.
> 
> > - it actually sends, receives and stores mail (akonadi users will
> >   appreciate how a mailer can get to make that insanely more complex
> >   than it needs to be)
> 
> I'm so glad I ditched KMail before akonadi and friends came along.

"The mythical man-month" by Brooks perfectly describes akonadi, they
went right ahead and made every mistake in the book, especially all the
ones solved in the 70s already,...


-- 
Alan McKinnnon
alan.mckinnon@gmail.com

Re: [gentoo-user] systemd

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 834 bytes --]

On Sat, 3 Sep 2011 11:40:52 +0200, Alan McKinnon wrote:

> > Set up a different (SMTP-only if necessary) account for each address
> > and specify the sig. You can set the default account to use for a
> > folder, which means you always use the correct address on a mailing
> > list. If no default account is set, Claws will default to using the
> > account that the mail you are replying to was sent to.  
> 
> I like to use local IMAP folder and route all mail into those,
> organized by work and other. I found that I can set prefs for a top
> folders and all children folders which handled this one nicely.

So do I. In Claws, folders and accounts are entirely separate - you can
have mail from various accounts in the same IMAP folder tree.


-- 
Neil Bothwick

Always be sincere... whether you mean it or not!

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] systemd

[Peter Humphrey]

On Saturday 03 September 2011 10:40:52 Alan McKinnon wrote:
> On Sat, 3 Sep 2011 01:05:33 +0100
> 
> Neil Bothwick <neil@digimed.co.uk> wrote:

> > I'm so glad I ditched KMail before akonadi and friends came along.

> "The mythical man-month" by Brooks perfectly describes akonadi, they
> went right ahead and made every mistake in the book, especially all the
> ones solved in the 70s already,...

What? Is that old dog still around? I read it about 20 years ago.

-- 
Rgds
Peter		Linux Counter 5290, 1994-04-23

Re: [gentoo-user] systemd

[Alan McKinnon]

On Sat, 3 Sep 2011 14:09:15 +0100
Peter Humphrey <peter@humphrey.ukfsn.org> wrote:

> On Saturday 03 September 2011 10:40:52 Alan McKinnon wrote:
> > On Sat, 3 Sep 2011 01:05:33 +0100
> > 
> > Neil Bothwick <neil@digimed.co.uk> wrote:
> 
> > > I'm so glad I ditched KMail before akonadi and friends came along.
> 
> > "The mythical man-month" by Brooks perfectly describes akonadi, they
> > went right ahead and made every mistake in the book, especially all
> > the ones solved in the 70s already,...
> 
> What? Is that old dog still around? I read it about 20 years ago.
 
It might be old but every word in it is still just as true as the day
it was written. Nothing else has come close to doing it like Brooks.

I found my copy in a tiny out of the way junk shop in a place called
Memel (at the arse end of the world), priced at about a cup of coffee.
A treasured find indeed :-)




-- 
Alan McKinnnon
alan.mckinnon@gmail.com

Re: [gentoo-user] systemd

[Peter Humphrey]

On Saturday 03 September 2011 01:05:33 Neil Bothwick wrote:

> I'm so glad I ditched KMail before akonadi and friends came along.

You've got me intrigued now. I haven't used Claws since an early, unstable 
version, but it looks as though it might suit me now.

First, though, does it have an easy bulk import mechanism from KMail format? 
I have over 10,000 mails here in about 50 folders, and I really wouldn't 
want to go through that lot one by one.

-- 
Rgds
Peter		Linux Counter 5290, 1994-04-23

Re: [gentoo-user] systemd

[Alan McKinnon]

On Sun, 4 Sep 2011 00:52:39 +0100
Peter Humphrey <peter@humphrey.ukfsn.org> wrote:

> On Saturday 03 September 2011 01:05:33 Neil Bothwick wrote:
> 
> > I'm so glad I ditched KMail before akonadi and friends came along.
> 
> You've got me intrigued now. I haven't used Claws since an early,
> unstable version, but it looks as though it might suit me now.
> 
> First, though, does it have an easy bulk import mechanism from KMail
> format? I have over 10,000 mails here in about 50 folders, and I
> really wouldn't want to go through that lot one by one.
> 

It doesn't support KMail folders directly (Claws doesn't do traditional
MailDir, it's native format is MH) but you can find import scripts here:

http://www.claws-mail.org/tools.php?section=downloads

There are kmail-specific scripts listed

-- 
Alan McKinnnon
alan.mckinnon@gmail.com

Re: [gentoo-user] systemd

[Peter Humphrey]

On Sunday 04 September 2011 01:19:51 Alan McKinnon wrote:

> It doesn't support KMail folders directly (Claws doesn't do traditional
> MailDir, it's native format is MH) but you can find import scripts here:
> 
> http://www.claws-mail.org/tools.php?section=downloads
> 
> There are kmail-specific scripts listed

Good stuff! Thanks Alan.

-- 
Rgds
Peter		Linux Counter 5290, 1994-04-23

Re: [gentoo-user] systemd

[Peter Humphrey]

On Sunday 04 September 2011 01:44:31 I wrote:
> On Sunday 04 September 2011 01:19:51 Alan McKinnon wrote:
> > It doesn't support KMail folders directly (Claws doesn't do traditional
> > MailDir, it's native format is MH) but you can find import scripts
> > here:
> > 
> > http://www.claws-mail.org/tools.php?section=downloads
> > 
> > There are kmail-specific scripts listed
> 
> Good stuff! Thanks Alan.

Except that, when I ran the converter script and restarted Claws, the folder 
structure had been duplicated but they were all empty. I have another head-
scratching opportunity...

-- 
Rgds
Peter		Linux Counter 5290, 1994-04-23

[gentoo-user] Re: systemd

[walt]

On 08/30/2011 04:56 AM, Alex Schuster wrote:
> Alan McKinnon writes:

> dbus-daemon often uses 10-20% of my CPU according to top.

Have you tried using dbus-monitor?  It may tell you if some app is
being inappropriate.

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Didn't do much research around this lately.

Today I revived my SSD (we'll see) and therefore fell over systemd when
I edited grub.conf

Where would/should I put stuff from /etc/local.d/ with systemd?

I have some commands there setting parameters for ssd-usage and those
would be skipped (not executed) with systemd.

Pls don't misunderstand, this is not gentoo ricer stuff, I am quite
happy w/ openrc, especially w/ the ssd.

Stefan

Re: [gentoo-user] systemd

[Canek Peláez Valdés]

On Tue, Oct 11, 2011 at 1:27 PM, Stefan G. Weichinger <lists@xunil.at> wrote:
>
>
> Didn't do much research around this lately.
>
> Today I revived my SSD (we'll see) and therefore fell over systemd when
> I edited grub.conf
>
> Where would/should I put stuff from /etc/local.d/ with systemd?
>
> I have some commands there setting parameters for ssd-usage and those
> would be skipped (not executed) with systemd.

Those command are on a script, right? Lets say the script is called
ssd-thingies; you put this on
/etc/systemd/system/ssd-thingies.service:

----------------------------------------------
[Unit]
Description=SSD thingies
After=basic.target

[Service]
Type=oneshot
ExecStart=/my/path/to/ssd-thingies

[Install]
WantedBy=multi-user.target
-----------------------------------------------

Then you run:

systemctl daemon-reload
systemctl enable ssd-thingies.service

Next time you reboot, the service will be called after the
basic.target has been completed. You can look at the status of the
script afterwards with

systemctl status ssd-thingies.service

If everything went OK, it should have a line like this:

Process: 1234 ExecStart=/my/path/to/ssd-thingies (code=exited, status=0/SUCCESS)

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 11.10.2011 23:04, schrieb Canek Peláez Valdés:

[...]

> systemctl status ssd-thingies.service
> 
> If everything went OK, it should have a line like this:
> 
> Process: 1234 ExecStart=/my/path/to/ssd-thingies (code=exited, status=0/SUCCESS)
> 
> Regards.

Thanks for the explanation!

I tried it right now, unfortunately I get:

# systemctl status ssd-thingies.service
ssd-thingies.service - SSD thingies
	  Loaded: loaded (/etc/systemd/system/ssd-thingies.service)
	  Active: failed since Tue, 11 Oct 2011 23:28:05 +0200; 21s ago
	 Process: 6696 ExecStart=/etc/local.d/stefan.start (code=exited,
status=203/EXEC)
	  CGroup: name=systemd:/system/ssd-thingies.service

Is it a permission-issue? AFAIK systemd runs w/ root?


# cat /etc/systemd/system/ssd-thingies.service

[Unit]
Description=SSD thingies
After=basic.target

[Service]
Type=oneshot
ExecStart=/etc/local.d/stefan.start

[Install]
WantedBy=multi-user.target


# ll /etc/local.d/stefan.start
-rwxr-xr-x 1 root root 795 11. Okt 16:47 /etc/local.d/stefan.start

---

Thanks, greets, Stefan

Re: [gentoo-user] systemd

[Canek Peláez Valdés]

On Tue, Oct 11, 2011 at 2:33 PM, Stefan G. Weichinger <lists@xunil.at> wrote:
> Am 11.10.2011 23:04, schrieb Canek Peláez Valdés:
>
> [...]
>
>> systemctl status ssd-thingies.service
>>
>> If everything went OK, it should have a line like this:
>>
>> Process: 1234 ExecStart=/my/path/to/ssd-thingies (code=exited, status=0/SUCCESS)
>>
>> Regards.
>
> Thanks for the explanation!
>
> I tried it right now, unfortunately I get:
>
> # systemctl status ssd-thingies.service
> ssd-thingies.service - SSD thingies
>          Loaded: loaded (/etc/systemd/system/ssd-thingies.service)
>          Active: failed since Tue, 11 Oct 2011 23:28:05 +0200; 21s ago
>         Process: 6696 ExecStart=/etc/local.d/stefan.start (code=exited,
> status=203/EXEC)
>          CGroup: name=systemd:/system/ssd-thingies.service
>
> Is it a permission-issue? AFAIK systemd runs w/ root?

It runs as root, but it's already telling you the problem:

> Process: 6696 ExecStart=/etc/local.d/stefan.start (code=exited, status=203/EXEC)

Your script (I believe) does not have execution perms. All the
commands for ExecStart (and ExecStop) need to be executable, so do a

chmod +x /etc/local.d/stefan.start

Also, if your scripts does not return 0 (or the last command it
executes does not return 0), it will tell you with the status= flag.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 12.10.2011 00:23, schrieb Canek Peláez Valdés:

> Your script (I believe) does not have execution perms. All the
> commands for ExecStart (and ExecStop) need to be executable, so do a
> 
> chmod +x /etc/local.d/stefan.start

I showed you before:

# ll /etc/local.d/stefan.start
-rwxr-xr-x 1 root root 795 11. Okt 16:47 /etc/local.d/stefan.start

> Also, if your scripts does not return 0 (or the last command it
> executes does not return 0), it will tell you with the status= flag.

Will check this.

Thanks, S

Re: [gentoo-user] systemd

[Canek Peláez Valdés]

On Tue, Oct 11, 2011 at 3:39 PM, Stefan G. Weichinger <lists@xunil.at> wrote:
> Am 12.10.2011 00:23, schrieb Canek Peláez Valdés:
>
>> Your script (I believe) does not have execution perms. All the
>> commands for ExecStart (and ExecStop) need to be executable, so do a
>>
>> chmod +x /etc/local.d/stefan.start
>
> I showed you before:
>
> # ll /etc/local.d/stefan.start
> -rwxr-xr-x 1 root root 795 11. Okt 16:47 /etc/local.d/stefan.start

Sorry, didn't see it. Can you execute it calling it directly? Maybe
it's missing the proper shebang.

>> Also, if your scripts does not return 0 (or the last command it
>> executes does not return 0), it will tell you with the status= flag.
>
> Will check this.

I really think is that systemd cannot execute it, but check that also.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 12.10.2011 00:47, schrieb Canek Peláez Valdés:

>> # ll /etc/local.d/stefan.start
>> -rwxr-xr-x 1 root root 795 11. Okt 16:47 /etc/local.d/stefan.start
> 
> Sorry, didn't see it. Can you execute it calling it directly? Maybe
> it's missing the proper shebang.

The shebang did the trick!

Thanks a lot, Stefan

[gentoo-user] Systemd

[siefke_listen]

[-- Attachment #1: Type: text/plain, Size: 244 bytes --]

Hello, 

I have a short question to systemd. I would like to ask your experience 
in the changeover. Was it easy? Were there problems?
Change or reinstall? What mean the profis here?

Thank you for help and have nice weekend.

Silvio 

[-- Attachment #2: Type: application/pgp-signature, Size: 488 bytes --]

Re: [gentoo-user] Systemd

[Neil Bothwick]

It's pretty straightforward, just follow the wiki. The profile takes care of setting the appropriate USE flags. One thing to watch out for is that systemd will not pick up your openrc startup services so save the output from rc-update - s then enable the services after changeover . 

On 4 November 2017 16:15:39 GMT, "siefke_listen@web.de" <siefke_listen@web.de> wrote:
>Hello, 
>
>I have a short question to systemd. I would like to ask your experience
>
>in the changeover. Was it easy? Were there problems?
>Change or reinstall? What mean the profis here?
>
>Thank you for help and have nice weekend.
>
>Silvio 

-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.

Re: [gentoo-user] Systemd

[siefke_listen]

On Sat, 04 Nov 2017 16:58:16 +0000
Neil Bothwick <neil@stfw.net> wrote:

> It's pretty straightforward, just follow the wiki. The profile takes care of setting the appropriate USE flags. One thing to watch out for is that systemd will not pick up your openrc startup services so save the output from rc-update - s then enable the services after changeover . 

That sounds good then I will do it. Thank you. 

Silvio

[gentoo-user] OT: but cool - NASDAQ is gentoo powered

[Adam Carter]

"Linux also offered financial firms the ability to modify the source
code to further speed performance, Lameter said. "It depends on how
daring the exchange is," Lameter said, noting that NASDAQ uses a
modified version of the Gentoo Linux distribution. "

http://www.itworld.com/open-source/193823/how-linux-mastered-wall-street

Re: [gentoo-user] OT: but cool - NASDAQ is gentoo powered

[Michael Mol]

On Mon, Aug 15, 2011 at 8:28 PM, Adam Carter <adamcarter3@gmail.com> wrote:
> http://www.itworld.com/open-source/193823/how-linux-mastered-wall-street

This is related to a question I wanted to poll the list with. How does
everybody here use Gentoo? For personal use? Production use? For
server, desktop or embedded roles? What's your most interesting setup
or use case?

I had Gentoo on both my desktop and HTPC, but I had to cannibalize the
HTPC for parts, so now it's just on my primary desktop box.

-- 
:wq

Re: [gentoo-user] OT: but cool - NASDAQ is gentoo powered

[Canek Peláez Valdés]

On Mon, Aug 15, 2011 at 9:48 PM, Michael Mol <mikemol@gmail.com> wrote:
> On Mon, Aug 15, 2011 at 8:28 PM, Adam Carter <adamcarter3@gmail.com> wrote:
>> http://www.itworld.com/open-source/193823/how-linux-mastered-wall-street
>
> This is related to a question I wanted to poll the list with. How does
> everybody here use Gentoo? For personal use? Production use? For
> server, desktop or embedded roles? What's your most interesting setup
> or use case?

I have one laptop, three desktops (two at my university), one HTPC
(with an ION Zotac mobo), and 2 servers, for very different purposes
each. The servers are in production.

Some years ago I administered the desktop machines in my work: Ten
gentoo boxen compiling in parallel with distcc; that was awesome.

My most interesting setup by far is the HTPC, I guess: it boots really
quickly (thanks to systemd), and it has a lot of little modifications
so I don't need to ssh into it to do anything: Everything is done
through the remote control.

In all my computers (i.e., not the two desktops at uni, since they are
not "mine") I use systemd (which thankfully has entered the portage
tree), and in my desktop and laptop I use GNOME 3 from the GNOME
overlay. It all works basically flawless.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] OT: but cool - NASDAQ is gentoo powered

[Stefan G. Weichinger]

Am 16.08.2011 19:06, schrieb Canek Peláez Valdés:

> In all my computers (i.e., not the two desktops at uni, since they are
> not "mine") I use systemd (which thankfully has entered the portage
> tree)

systemd sounds like a nice-to-have project for me.

Which howto did you follow, what do you recommend me to read to start
using it (with gentoo, sure) ?

Thanks, Stefan

Re: [gentoo-user] OT: but cool - NASDAQ is gentoo powered

[Paul Hartman]

On Tue, Aug 16, 2011 at 3:58 PM, Stefan G. Weichinger <lists@xunil.at> wrote:
> Am 16.08.2011 19:06, schrieb Canek Peláez Valdés:
>
>> In all my computers (i.e., not the two desktops at uni, since they are
>> not "mine") I use systemd (which thankfully has entered the portage
>> tree)
>
> systemd sounds like a nice-to-have project for me.
>
> Which howto did you follow, what do you recommend me to read to start
> using it (with gentoo, sure) ?

I became afraid after reading this wiki page, especially the part
about removing openrc and making your own init.d and conf.d entries:

http://en.gentoo-wiki.com/wiki/Systemd

:)

Re: [gentoo-user] OT: but cool - NASDAQ is gentoo powered

[Canek Peláez Valdés]

On Tue, Aug 16, 2011 at 5:06 PM, Paul Hartman
<paul.hartman+gentoo@gmail.com> wrote:
> On Tue, Aug 16, 2011 at 3:58 PM, Stefan G. Weichinger <lists@xunil.at> wrote:
>> Am 16.08.2011 19:06, schrieb Canek Peláez Valdés:
>>
>>> In all my computers (i.e., not the two desktops at uni, since they are
>>> not "mine") I use systemd (which thankfully has entered the portage
>>> tree)
>>
>> systemd sounds like a nice-to-have project for me.
>>
>> Which howto did you follow, what do you recommend me to read to start
>> using it (with gentoo, sure) ?
>
> I became afraid after reading this wiki page, especially the part
> about removing openrc and making your own init.d and conf.d entries:
>
> http://en.gentoo-wiki.com/wiki/Systemd

I don't know about the wiki (I didn't use it to install systemd), and
as I said, I think it works out-of-the-box now, and you can safely go
back to OpenRC if you want to.

Regards.
-- 
Canek Peláez Valdés
Posgrado en Ciencia e Ingeniería de la Computación
Universidad Nacional Autónoma de México

Re: [gentoo-user] OT: but cool - NASDAQ is gentoo powered

[Stefan G. Weichinger]

Am 17.08.2011 01:24, schrieb Canek Peláez Valdés:

>> http://en.gentoo-wiki.com/wiki/Systemd
> 
> I don't know about the wiki (I didn't use it to install systemd), and
> as I said, I think it works out-of-the-box now, and you can safely go
> back to OpenRC if you want to.

Installed it in a VM now, and followed the wiki ...

I don't even get a console so far, some strange timeouts somewhere.
getty@tty1.service depends on something I can't see (no way to scroll
up) ... still some fiddling needed here.

S

[gentoo-user] systemd (was: NASDAQ is gentoo powered)

[Stefan G. Weichinger]

Am 17.08.2011 16:04, schrieb Stefan G. Weichinger:
> Am 17.08.2011 01:24, schrieb Canek Peláez Valdés:
> 
>>> http://en.gentoo-wiki.com/wiki/Systemd
>>
>> I don't know about the wiki (I didn't use it to install systemd), and
>> as I said, I think it works out-of-the-box now, and you can safely go
>> back to OpenRC if you want to.
> 
> Installed it in a VM now, and followed the wiki ...
> 
> I don't even get a console so far, some strange timeouts somewhere.
> getty@tty1.service depends on something I can't see (no way to scroll
> up) ... still some fiddling needed here.

In rescue mode I read something about /etc/mtab, google says "ignore that".

Sigh. Still no ttys here.
"out of the box" should feel different.

Will dig more ...

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 17.08.2011 16:47, schrieb Stefan G. Weichinger:

> Sigh. Still no ttys here.
> "out of the box" should feel different.
> 
> Will dig more ...

Sorry for the noise, got it now. Too old udev etc before.

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 17.08.2011 17:00, schrieb Stefan G. Weichinger:
> Am 17.08.2011 16:47, schrieb Stefan G. Weichinger:
> 
>> Sigh. Still no ttys here.
>> "out of the box" should feel different.
>>
>> Will dig more ...
> 
> Sorry for the noise, got it now. Too old udev etc before.

First steps:

added network.service and sshd.service

following

http://en.gentoo-wiki.com/wiki/Systemd#Services

sshd.service gets started at boot, network.service not ...

I linked multi-user.target to /etc/systemd/system/default.target, didn't
help.

Do I need that link?

Starting network.service with systemctl works, so no typo hidden here,
as it seems.

Stefan

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Am 17.08.2011 18:00, schrieb Stefan G. Weichinger:

> sshd.service gets started at boot, network.service not ...
> 
> I linked multi-user.target to /etc/systemd/system/default.target,
> didn't help.
> 
> Do I need that link?

Solved, but dunno if done correctly.

ln -sf /etc/systemd/system/network.service
/etc/systemd/system/multi-user.target.wants

ln -sf /lib/systemd/system/multi-user.target
/etc/systemd/system/default.target

S

Re: [gentoo-user] systemd

[Stefan G. Weichinger]

Yes, I know, I should start a new thread.

So far I got the impression that it would take quite some time and work
to get my machines and their services configured correctly.

For now I will keep it inside that ~amd64-VM and continue to test and learn.

Stefan