tags #679983 confirmed pending
thanks
On Mon, Jul 02, 2012 at 09:45:25PM +0200, Michael Stapelberg wrote:
> I created a systemd service file for atop.
The latest upstream release uses a systemd service file, which,
however only calls the upstream start script.
Greetings
Marc
Hi Michael,
thanks for your interest in systemd and your efforts trying to get
packages to ship native systemd service files.
I became aware of this bug report via debian-release because Marc asked
for a freeze exception and would like to comment on a few issues:
- the .service file uses a depre
On Wed, Jul 11, 2012 at 10:21:13AM +0200, Michael Stapelberg wrote:
> Quoting Marc Haber (2012-07-09 10:03:46)
> > Looking forward to the patch set, welcome on board.
> You can find the patch attached. I decided to make it one single commit
> because all the changes go hand in hand. Therefore, you
Hi Marc,
Quoting Marc Haber (2012-07-09 10:03:46)
> Looking forward to the patch set, welcome on board.
You can find the patch attached. I decided to make it one single commit
because all the changes go hand in hand. Therefore, you never end up
with a broken atop, no matter which commit you check
On Mon, Jul 09, 2012 at 06:07:25AM +0200, Michael Stapelberg wrote:
> Quoting Marc Haber (2012-07-08 23:37:56)
> > On Debian/kFreeBSD?
> Of course not, but that’s not the point.
It is. Maintaining compatibility code for > 1 init system is bound to
be untested and bug prone.
Looking forward to the
Hi Marc,
Quoting Marc Haber (2012-07-08 23:37:56)
> On Debian/kFreeBSD?
Of course not, but that’s not the point.
> I would like to invite you to take part more directly. atop is
> maintained on alioth in the collab-maint project. Feel free to clone
> the git from there. While I would like to revi
On Sun, Jul 08, 2012 at 10:46:55PM +0200, Michael Stapelberg wrote:
> Quoting Marc Haber (2012-07-08 22:07:33)
> > > I’m not entirely sure what the 3-second delay is for (and it’s not
> > > documented), so I cannot comment on that.
> >
> > It's what upstream does. I guess that the sleep is there t
Hi Marc,
Quoting Marc Haber (2012-07-08 22:07:33)
> > I’m not entirely sure what the 3-second delay is for (and it’s not
> > documented), so I cannot comment on that.
>
> It's what upstream does. I guess that the sleep is there to allow the
> dying atop to close the log file.
That can be solved m
On Sun, Jul 08, 2012 at 04:16:44PM +0200, Michael Stapelberg wrote:
> So, _cron is essentially restart, but with a 3-second delay followed by
> deleting old atop_* files.
It is essentially what the upstream cron job did in a dedicated shell
script, using its own configuration and introducing possi
Hi Marc,
Quoting Marc Haber (2012-07-03 09:30:24)
> > No worries, there is a compatibility layer: /lib/lsb/init-functions
> > diverts the call to systemd, thus /etc/init.d/atop stop is equivalent to
> > directly doing systemctl stop atop.service, likewise for start and
> > restart.
>
> And for _c
On Mon, Jul 02, 2012 at 11:54:59PM +0200, Michael Stapelberg wrote:
> Quoting Marc Haber (2012-07-02 23:09:13)
> > the cron job calls the init script. Is this really the intended way to
> > interact with systemd? How is it made sure that the new atop process
> > is placed into the correct cgroup, h
Hi Marc,
Quoting Marc Haber (2012-07-02 23:09:13)
> the cron job calls the init script. Is this really the intended way to
> interact with systemd? How is it made sure that the new atop process
> is placed into the correct cgroup, how is a systemd-induced restart of
> systemd inhibited?
No worries
On Mon, Jul 02, 2012 at 11:04:43PM +0200, Michael Stapelberg wrote:
> Quoting Marc Haber (2012-07-02 22:53:09)
> > systemd needs to restart atop daily at midnight, the log rotation
> > mechanism depends on that.
> >
> > Please take a close look at how init script and daily cron job work
> > togeth
Hi Marc,
Quoting Marc Haber (2012-07-02 22:53:09)
> systemd needs to restart atop daily at midnight, the log rotation
> mechanism depends on that.
>
> Please take a close look at how init script and daily cron job work
> together.
I did. Why would this not work with systemd? logrotate will stop a
tags #679983 - patch
thanks
On Mon, Jul 02, 2012 at 09:45:25PM +0200, Michael Stapelberg wrote:
> [Unit]
> Description=advanced interactive monitor
> After=syslog.target
> Documentation=man:atop(1)
>
> [Service]
> ExecStart=/usr/bin/atop -a -w /var/log/atop.log 600
>
> [Install]
> WantedBy=multi
Package: atop
Severity: wishlist
Tags: patch
Hello,
I created a systemd service file for atop.
systemd is a sysvinit replacement (see [1] for more information).
Please include this file in /lib/systemd/system/ in the next upload of this
package.
I also sent this file upstream, so hopefully it
16 matches
Mail list logo
