Bug#768178: systemd: sysvinit wrapper breaks newly-installed services

[Ximin Luo]

On 05/11/14 21:52, Ximin Luo wrote:
> On 05/11/14 20:54, Zbigniew Jędrzejewski-Szmek wrote:
>> On Wed, Nov 05, 2014 at 08:12:39PM +0000, Ximin Luo wrote:
>>> All I care is that "service x start" works. It does not. This is
>>> correctly called "systemd breaks existing software" - it is breaking
>>> the sysvinit behaviour.
>>
>> Let's look what LSB says:
>>
>> http://refspecs.linuxbase.org/LSB_3.1.1/LSB-Core-generic/LSB-Core-generic/iniscrptact.html
>>
>>> For all other
>> [meaning everything apart from 'status']
>>> init-script actions, the init script shall return an exit status of
>>> zero if the action was successful. Otherwise, the exit status shall
>>> be non-zero
>> [...]
>>> 6	program is not configured
>>
>> Esentially, your scripts tells systemd "action was sucessful", and
>> systemd has no reason (or way) to doubt that.
>>
> 
> Sure, I get that my script doesn't work perfectly in this area. I will fix all of this stuff.
> 
> However, systemd is *also* at fault for not being able to deal with this. With sysvinit, a "start" action is executed regardless of any external considerations - it's just a shell script with a "start" clause. This behaviour should be preserved by systemd; however what actually happens is that this behaviour is overridden quite obnoxiously.
> 
> Nowhere does the LSB mention "scripts should not expect to work in the future, if they output a non-standard return code previously". This is what systemd does to my script (and other scripts that use dh_installinit).
> 
> So there is still a bug in /lib/lsb/init-functions.d/40-systemd.
> 

Take another example: even if my script did work perfectly according to the LSB, a service might stop *itself* a short time after "service x start" completes with return code 0. systemd will be unable to detect this [1], and a subsequent "service x start" will not work under systemd (because it thinks it's active, and treats the "start" subcommand as a no-op), even though this would work under plain sysvinit.

X

[1] without providing metadata in a pseudo-header - but we're working under the assumption that initscript writers should not have to care about systemd.

-- 
GPG: 4096R/1318EFAC5FBBDBCE
git://github.com/infinity0/pubkeys.git

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <http://alioth-lists.debian.net/pipermail/pkg-systemd-maintainers/attachments/20141105/6062b8f2/attachment-0002.sig>