Sync journal with fdatasync after 10s of inactivity (by default), or
after 10 messages (by default), or after 1h of last commit (by
default). Intervals configured via SyncIntervalIdleSec,
SyncIntervalMaxSec and SyncIntervalMsg options at journal.conf.
Manual sync can be performed via sending SI
> > +if (! (f->writable && f->fd >= 0))
> > +return -1;
> -1 means EPERM. Something different is needed.
EINVAL/EBUSY/EFAULT ?
> > +fdatasync(f->fd);
> Apparently the only error code which could happen here is EIO.
> Are you sure that we want to ignore it?
I think
On Tue, Mar 19, 2013 at 2:29 AM, Zbigniew Jędrzejewski-Szmek
wrote:
> ---
> A serie composed of three patches:
> 1. this one: make sd_journal_add_file available to the wide public
> 2. second one: allow opening of the journal in "stub" mode, without
>any files
> 3. third one: journalctl --file
This is useful for debugging and feels pretty natural. For example
answering the question "is this big .journal file worth keeping?"
is made easier.
---
man/journalctl.xml | 31 ++-
src/journal/journalctl.c | 33 -
2 files changed,
Explicit flag is required, since this is supposed to be used
mostly for debugging and it is better to catch the mistake
of not specyfing any files.
---
src/journal/sd-journal.c | 11 +++
src/systemd/sd-journal.h | 3 ++-
2 files changed, 9 insertions(+), 5 deletions(-)
diff --git a/src/j
---
A serie composed of three patches:
1. this one: make sd_journal_add_file available to the wide public
2. second one: allow opening of the journal in "stub" mode, without
any files
On Tue, Mar 19, 2013 at 4:33 AM, Kok, Auke-jan H
wrote:
> On Sat, Mar 16, 2013 at 8:10 AM, Cedric BAIL wrote:
>>As both of those process start visible application that should not
>> be killed on a restart of the initialization daemon carelessly, they
>> do have the same issue as Enlightenment
On Sat, 2013-03-16 at 18:45 +0200, Oleksii Shevchuk wrote:
> Sync journal with fdatasync after [...] 10 messages (by default)
That's going to be kind of painful for the current GNOME userspace, we
sadly emit a lot of crap at bootup =/ (I'm slowly trying to trim it,
but...it's slow going)
And boo
On Mon, Mar 18, 2013 at 02:38:48PM -0400, Colin Walters wrote:
> ---
> src/login/inhibit.c |9 ++---
> src/login/loginctl.c| 23 ---
> src/login/logind-dbus.c | 29 +
> src/login/logind-seat-dbus.c| 2
On Mon, Mar 18, 2013 at 02:38:24PM -0400, Colin Walters wrote:
> This is a followup to: commit 1a37b9b9043ef83e9900e460a9a1fccced3acf89
>
> It will fix denial messages from dbus-daemon between gdm and
> systemd-logind on logging into GNOME due to this.
Applied.
Zbyszek
___
On Mon, Mar 18, 2013 at 10:17:25PM +0200, Oleksii Shevchuk wrote:
> Sync journal with fdatasync after 10s of inactivity (by default), or
> after 10 messages (by default). Intervals configured
> via SyncIntervalSec and SyncIntervalMsg options at journal.conf.
> Manual sync can be performed via sendi
You are right, hopefully systemd will not turn into a direction
where only the ulra modern notebook or desktop can use it.
we see a lot of advantages to use systemd in embedded environments.
and we are talking here about_
- 180-680 MHh MIPS, ARM CPUs
- 4-32 MB Flash
- 32-128MB RAM
at the low end.
On Mon, Mar 18, 2013 at 11:57 AM, Reindl Harald wrote:
>
>
> Am 18.03.2013 19:45, schrieb Kay Sievers:
>> I put an SSD in that crappy box today; it's down from 25 to 7 sec on
>> the otherwise identical system. :)
>>
>> So I'll not need to debug any rotating media issues, I don't have any
>> of the
On Mon, Mar 18, 2013 at 2:30 PM, Holger Winkelmann wrote:
> Hi,
>
> Just a quick note, Systemd is a core function on a core OS.
> and a coreOS is likely to be started from flash drives. We have
> i.e. ATCA devices starting from eUSB Sticks or even SD cards.
>
> After they started the nodes could h
Am 18.03.2013 22:30, schrieb Holger Winkelmann:
> Just a quick note, Systemd is a core function on a core OS.
> and a coreOS is likely to be started from flash drives. We have
> i.e. ATCA devices starting from eUSB Sticks or even SD cards.
well, our ESXi starts also from a SD-card
/dev/md1
Hi,
Just a quick note, Systemd is a core function on a core OS.
and a coreOS is likely to be started from flash drives. We have
i.e. ATCA devices starting from eUSB Sticks or even SD cards.
After they started the nodes could have high perfomance SSD drives
mounted for payload data, or even not st
]] Reindl Harald
> but keep in mind that for professional environments for many
> years SSD is no option for some TB of data and even if
> the price falls down you have to calculate redundancy for
> RAID10 environemnts which can not be raplced by a SSD
> due lack of relieability and no real-world
Sync journal with fdatasync after 10s of inactivity (by default), or
after 10 messages (by default). Intervals configured
via SyncIntervalSec and SyncIntervalMsg options at journal.conf.
Manual sync can be performed via sending SIGUSR1.
---
src/journal/journal-file.c | 25 +++-
src/jour
On Sat, Mar 16, 2013 at 8:10 AM, Cedric BAIL wrote:
>As both of those process start visible application that should not
> be killed on a restart of the initialization daemon carelessly, they
> do have the same issue as Enlightenment.
>
>Something that would be nice is if we could instruct
Am 18.03.2013 19:45, schrieb Kay Sievers:
> I put an SSD in that crappy box today; it's down from 25 to 7 sec on
> the otherwise identical system. :)
>
> So I'll not need to debug any rotating media issues, I don't have any
> of them again. :)
this is nice for you
but keep in mind that for pro
On Mon, Mar 18, 2013 at 7:40 PM, Kok, Auke-jan H
wrote:
> On Sun, Mar 17, 2013 at 6:54 AM, Kay Sievers wrote:
>> Here is a chart:
>> http://people.freedesktop.org/~kay/bootchart-20130317-1434.svg
>>
>> Rotating media and really cheap hardware looks very sad, and we take
>> like 5 times longer t
On Sun, Mar 17, 2013 at 6:54 AM, Kay Sievers wrote:
> Here is a chart:
> http://people.freedesktop.org/~kay/bootchart-20130317-1434.svg
>
> Rotating media and really cheap hardware looks very sad, and we take
> like 5 times longer to boot than Windows 8.
I've seen many charts like that - in mos
On Mon, Mar 18, 2013 at 6:27 PM, Tom Gundersen wrote:
> On Mon, Mar 18, 2013 at 3:20 PM, Tom Gundersen wrote:
>> So perhaps "net.predictable-names"
>
> Pushed this. Feel free to rename.
Renamed to net.ifnames= to be more digestible for the kernel. If that
ever happens ...
I removed the initrd s
---
src/login/inhibit.c |9 ++---
src/login/loginctl.c| 23 ---
src/login/logind-dbus.c | 29 +
src/login/logind-seat-dbus.c| 22 +-
src/login/logind-session-dbus.c | 29 +++-
This is a followup to: commit 1a37b9b9043ef83e9900e460a9a1fccced3acf89
It will fix denial messages from dbus-daemon between gdm and
systemd-logind on logging into GNOME due to this.
See the previous commit for more details.
---
src/core/dbus-job.c |4 ++--
src/core/dbus-manager.c
'Twas brillig, and Britton Kerin at 18/03/13 17:49 did gyre and gimble:
> Hi everyone, I just had my first encounter with systemd and all in all I'm
> highly impressed.
>
> One thing saddened me a bit though: its not obvious what to do if I just want
> to start a service or do something *after* al
My service needs reasonably continuous mtimes on files, but at the
moment there is a big discontinuity soon after boot.
There are a number of web pages out there which tell you to put
stuff like this:
After=syslog.target network.target
WantedBy=multi-user.target
In your .service file lin
Am 18.03.2013 18:49, schrieb Britton Kerin:
> Hi everyone, I just had my first encounter with systemd and all in all I'm
> highly impressed.
>
> One thing saddened me a bit though: its not obvious what to do if I just want
> to start a service or do something *after* all the stuff that shipped w
Hi everyone, I just had my first encounter with systemd and all in all I'm
highly impressed.
One thing saddened me a bit though: its not obvious what to do if I just want
to start a service or do something *after* all the stuff that shipped with ths
OS has happened and the system is fully operatio
On Mon, Mar 18, 2013 at 3:20 PM, Tom Gundersen wrote:
> So perhaps "net.predictable-names"
Pushed this. Feel free to rename.
-t
___
systemd-devel mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/systemd-dev
On Mon, Mar 18, 2013 at 5:09 PM, Kay Sievers wrote:
> On Mon, Mar 18, 2013 at 4:04 PM, Tom Gundersen wrote:
>> If not, I guess we could
>> rename it an use RUN, otherwise I suppose we'd have to stick with
>> IMPORT... Anyone else have an opinion?
>
> I would not expect any real problems; the modp
Le lundi 18 mars 2013 à 14:34 +0100, Lennart Poettering a écrit :
> On Mon, 18.03.13 09:39, Frederic Crozat ([email protected]) wrote:
>
> > Le dimanche 17 mars 2013 à 14:54 +0100, Kay Sievers a écrit :
> > > Here is a chart:
> > > http://people.freedesktop.org/~kay/bootchart-20130317-1434.svg
>
On Mon, Mar 18, 2013 at 4:04 PM, Tom Gundersen wrote:
> On Mon, Mar 18, 2013 at 3:31 PM, Kay Sievers wrote:
>> On Mon, Mar 18, 2013 at 3:24 PM, Tom Gundersen wrote:
>>> On Mon, Mar 18, 2013 at 3:01 PM, Kay Sievers wrote:
>>
IMPORT is used to be called immediately and not be queued up to th
On Mon, Mar 18, 2013 at 3:31 PM, Kay Sievers wrote:
> On Mon, Mar 18, 2013 at 3:24 PM, Tom Gundersen wrote:
>> On Mon, Mar 18, 2013 at 3:01 PM, Kay Sievers wrote:
>
>>> IMPORT is used to be called immediately and not be queued up to the
>>> end of the rules handling, so that later rules could as
On Mon, Mar 18, 2013 at 2:58 PM, Zbigniew Jędrzejewski-Szmek
wrote:
> On Mon, Mar 18, 2013 at 02:17:26PM +0100, Tom Gundersen wrote:
>>Every line in the rules file contains at least one key-value
>> pair.
>> - There are two kinds of keys: match and assignment.
>> - If all match
On Mon, Mar 18, 2013 at 3:24 PM, Tom Gundersen wrote:
> On Mon, Mar 18, 2013 at 3:01 PM, Kay Sievers wrote:
>> IMPORT is used to be called immediately and not be queued up to the
>> end of the rules handling, so that later rules could assume that this
>> rule has finished. PROGRAM would do the s
On Mon, Mar 18, 2013 at 3:01 PM, Kay Sievers wrote:
> On Mon, Mar 18, 2013 at 2:17 PM, Tom Gundersen wrote:
>> The 'kmod' and 'firmware' builtin's do not set any variables, so no need to
>> use
>> IMPORT. Use RUN as is already the case for the 'uaccess' builtin.
>
> IMPORT is used to be called i
On Mon, Mar 18, 2013 at 2:17 PM, Tom Gundersen wrote:
> Distros that whish to support old kernels should set
> --with-firmware-dirs="/usr/lib/firmware/updates:/usr/lib/firmware"
> to retain the old behaviour.
Applied.
> The hack which makes firmware events skip the dependency tracking is still
On Mon, Mar 18, 2013 at 3:07 PM, Kay Sievers wrote:
> On Mon, Mar 18, 2013 at 2:17 PM, Tom Gundersen wrote:
>> The properties will still be set in the udev database, but they will not be
>> used
>> for setting the interface names. As for the other kernel commandline
>> switches,
>> we allow it
On Mon, Mar 18, 2013 at 3:07 PM, Tom Gundersen wrote:
> On Mon, Mar 18, 2013 at 3:02 PM, Kay Sievers wrote:
>> On Mon, Mar 18, 2013 at 2:17 PM, Tom Gundersen wrote:
>>> ---
>>> man/udev.xml | 43 ---
>>> 1 file changed, 36 insertions(+), 7 deletions(-)
>>
On Mon, Mar 18, 2013 at 2:17 PM, Tom Gundersen wrote:
> The properties will still be set in the udev database, but they will not be
> used
> for setting the interface names. As for the other kernel commandline switches,
> we allow it to be prefixed by 'rd.' to only apply in the initrd.
> +IMPORT
On Mon, Mar 18, 2013 at 3:02 PM, Kay Sievers wrote:
> On Mon, Mar 18, 2013 at 2:17 PM, Tom Gundersen wrote:
>> ---
>> man/udev.xml | 43 ---
>> 1 file changed, 36 insertions(+), 7 deletions(-)
>
> Hmm, they are documented in systemd-udevd, what's the right
On Mon, Mar 18, 2013 at 2:17 PM, Tom Gundersen wrote:
> ---
> man/udev.xml | 43 ---
> 1 file changed, 36 insertions(+), 7 deletions(-)
Hmm, they are documented in systemd-udevd, what's the right place?
Kay
___
On Mon, Mar 18, 2013 at 2:17 PM, Tom Gundersen wrote:
> The 'kmod' and 'firmware' builtin's do not set any variables, so no need to
> use
> IMPORT. Use RUN as is already the case for the 'uaccess' builtin.
IMPORT is used to be called immediately and not be queued up to the
end of the rules handl
On Mon, Mar 18, 2013 at 02:17:26PM +0100, Tom Gundersen wrote:
> ---
> man/udev.xml | 43 ---
> 1 file changed, 36 insertions(+), 7 deletions(-)
>
> diff --git a/man/udev.xml b/man/udev.xml
> index 34b1e6f..fed8a5e 100644
> --- a/man/udev.xml
> +++ b/man/ud
On Mon, Mar 18, 2013 at 2:17 PM, Tom Gundersen wrote:
> The new IMPORT{builtin} and RUN{builtin} were not documented. Also make it
> clear that RUN= is
> really an alias for RUN{program}=.
Applied.
Thanks,
Kay
___
systemd-devel mailing list
systemd-de
On Mon, 18.03.13 09:39, Frederic Crozat ([email protected]) wrote:
> Le dimanche 17 mars 2013 à 14:54 +0100, Kay Sievers a écrit :
> > Here is a chart:
> > http://people.freedesktop.org/~kay/bootchart-20130317-1434.svg
> >
> > Rotating media and really cheap hardware looks very sad, and we take
On Sat, 16.03.13 18:45, Oleksii Shevchuk ([email protected]) wrote:
> +int journal_file_sync(JournalFile *f) {
> +if (! (f->writable && f->fd >= 0))
> +return -1;
> +
> +if (fdatasync(f->fd))
> +return -errno;
> +
> +return 0;
> +}
I'd really
Distros that whish to support old kernels should set
--with-firmware-dirs="/usr/lib/firmware/updates:/usr/lib/firmware"
to retain the old behaviour.
The hack which makes firmware events skip the dependency tracking is still
preserved, this would allow adding back firmware handling by a custom ud
The properties will still be set in the udev database, but they will not be used
for setting the interface names. As for the other kernel commandline switches,
we allow it to be prefixed by 'rd.' to only apply in the initrd.
---
TODO | 2 --
man/udev.xml | 6
---
man/udev.xml | 43 ---
1 file changed, 36 insertions(+), 7 deletions(-)
diff --git a/man/udev.xml b/man/udev.xml
index 34b1e6f..fed8a5e 100644
--- a/man/udev.xml
+++ b/man/udev.xml
@@ -55,14 +55,12 @@
Configuration
-udev configuration files
The 'kmod' and 'firmware' builtin's do not set any variables, so no need to use
IMPORT. Use RUN as is already the case for the 'uaccess' builtin.
---
rules/50-udev-default.rules | 2 +-
rules/80-drivers.rules | 12 ++--
2 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/r
The new IMPORT{builtin} and RUN{builtin} were not documented. Also make it
clear that RUN= is
really an alias for RUN{program}=.
---
man/udev.xml | 36 ++--
1 file changed, 30 insertions(+), 6 deletions(-)
diff --git a/man/udev.xml b/man/udev.xml
index 415cf8e..34
Am 18.03.2013 14:08, schrieb Zbigniew Jędrzejewski-Szmek:
> On Mon, Mar 18, 2013 at 02:04:39PM +0100, [email protected] wrote:
>> From: Harald Hoyer
>>
>> Instead of typing the rather unusual:
>> $ journalctl _SYSTEMD_UNIT=sshd.service
> What about -u sshd.service? If anything, I think this should
On Mon, Mar 18, 2013 at 02:04:39PM +0100, [email protected] wrote:
> From: Harald Hoyer
>
> Instead of typing the rather unusual:
> $ journalctl _SYSTEMD_UNIT=sshd.service
What about -u sshd.service? If anything, I think this should follow
the same rules as -u.
Zbyszek
__
From: Harald Hoyer
Instead of typing the rather unusual:
$ journalctl _SYSTEMD_UNIT=sshd.service
one can now type
$ journalctl sshd.service
-- Logs begin at So 2013-02-24 20:54:44 CET, end at Mo 2013-03-18 14:01:01 CET.
--
Mär 18 07:48:26 lenovo sshd[400]: Server listening on 0.0.0.0 port 22.
On 16/03/13 15:10, Cedric BAIL wrote:
> I think I am a little bit late about integrating systemd user
> session in a desktop
Not really; as far as I can see, non-trivial systemd user sessions under
X11 need some more thought, and some more code.
Specifically, they need at least a change to lo
And make acceptance tests on such machines ;-)
- Original Message -
> 'Twas brillig, and Kay Sievers at 18/03/13 10:42 did gyre and gimble:
> > On Mon, Mar 18, 2013 at 5:06 AM, Zbigniew Jędrzejewski-Szmek
> > wrote:
> >> On Sun, Mar 17, 2013 at 02:54:01PM +0100, Kay Sievers wrote:
> >>> H
'Twas brillig, and Kay Sievers at 18/03/13 10:42 did gyre and gimble:
> On Mon, Mar 18, 2013 at 5:06 AM, Zbigniew Jędrzejewski-Szmek
> wrote:
>> On Sun, Mar 17, 2013 at 02:54:01PM +0100, Kay Sievers wrote:
>>> Here is a chart:
>>> http://people.freedesktop.org/~kay/bootchart-20130317-1434.svg
>>
Am 17.03.2013 14:54, schrieb Kay Sievers:
> Here is a chart:
> http://people.freedesktop.org/~kay/bootchart-20130317-1434.svg
>
> Rotating media and really cheap hardware looks very sad, and we take
> like 5 times longer to boot than Windows 8.
>
> Why does bootchart stop before all the *really
Am 18.03.2013 11:42, schrieb Kay Sievers:
> On Mon, Mar 18, 2013 at 5:06 AM, Zbigniew Jędrzejewski-Szmek
> wrote:
>> On Sun, Mar 17, 2013 at 02:54:01PM +0100, Kay Sievers wrote:
>>> Here is a chart:
>>> http://people.freedesktop.org/~kay/bootchart-20130317-1434.svg
>>>
>>> Rotating media and r
On Mon, Mar 18, 2013 at 5:06 AM, Zbigniew Jędrzejewski-Szmek
wrote:
> On Sun, Mar 17, 2013 at 02:54:01PM +0100, Kay Sievers wrote:
>> Here is a chart:
>> http://people.freedesktop.org/~kay/bootchart-20130317-1434.svg
>>
>> Rotating media and really cheap hardware looks very sad, and we take
>> l
Le dimanche 17 mars 2013 à 14:54 +0100, Kay Sievers a écrit :
> Here is a chart:
> http://people.freedesktop.org/~kay/bootchart-20130317-1434.svg
>
> Rotating media and really cheap hardware looks very sad, and we take
> like 5 times longer to boot than Windows 8.
>
> Why does bootchart stop be
63 matches
Mail list logo
