On Wed, Jan 19, 2022 at 15:31:48 +0000, Daniel P. Berrangé wrote:
> On Wed, Jan 19, 2022 at 04:05:05PM +0100, Peter Krempa wrote:
> > Use a common wording for all various daemons to prevent having to adjust
> > the documentation any further by just outlining that neither clients nor
> > anything else needing the attention of the daemon must be present in
> > order to shut down.
>
> I don't think this is a good idea. The --timeout arg should be
> specific about exactly what scenarios block timeout for the
> daemon in question.
>
> > Resolves:
https://bugzilla.redhat.com/show_bug.cgi?id=2035985
>
> THis bug is reporting that they were expecting virtnetworkd
> to stay running as long as any domains were still running.
> ie they thought there was a link between virtqemud and
> virtnetworkd staying running. The proposed text still
> leaves that confusion present so doesn't address the problem
> reported IMHO.
I wanted to be vague so that once another condition blocking the
shutdown will be added we will not need to go back and modify the
documentation.
Given that all the daemons are a copy&paste of the original one (even
those which don't deal with domains, such as the storage driver) I
can't see how that will be better the next time.
Thus I'd still strongly prefer a wording that will not require naming
all the specific objects blocking the shutdown.
Being vague like that is not good for users wanting to understand
the behaviour. This ability to be specific is why we have separate
man pages for each daemon to start off with.
Regards,
Daniel
--
|: