[systemd-devel] Improved systemd service configuration from LSB init scripts?

Lennart Poettering lennart at poettering.net
Sat Jul 3 06:28:35 PDT 2010


On Mon, 07.06.10 20:17, Adam Spragg (adam at spra.gg) wrote:

> Hi there.

Heya,

> I was working on something earlier today which caused me to look at LSB init 
> script headers, and it made me think of the systemd ability to use "legacy" 
> LSB init scripts for "service" configuration, with the following query:
> 
> Does anyone think it would be useful to use the local extension facility 
> provided by LSB[0] to add support for "X-systemd-*" keywords in existing init 
> scripts?

I am not convinced that this is a good idea. Instead of having people
beef up their old SysV init scripts they should just write new and clean
systemd .service files. Especially if people want to use advanced
features of systemd they should use our native format instead of clunky
SysV scripts.

> That might give sysadmins/distros an easier way to configure, or provide an 
> upgrade path to, systemd, while still maintaining sysv as an
> option/fallback. 

We already provide that: if both a sysv init script and a native unit
file are installed the latter takes precedence. That allows people to
reliably boot up a system with either init system.

> It could also head off inevitable bug reports from users who are not 
> interested in systemd and do not have it installed, but nevertheless get 
> systemd config files dropped about the place, because the daemon package will 
> need to include/install them.

Well, I am not too worried about that.

In general, I am happy to adopt init script extensions that have been
introduced by distributions -- if they make sense and are actually
used. But on top of that I do not want to introduce any new extensions
for SysV init scripts. They are a dead end.

At the moment we already interpret quite a few extensions from various
distributions: from Debian we took a few facility definitions, from Suse
we have support for the special "boot" runlevel, and from Fedora we
chose to interpret the chkconfig header.

Lennart

-- 
Lennart Poettering - Red Hat, Inc.


More information about the systemd-devel mailing list