Re: starting/stopping daemons in package scripts
Remco Blaakmeer wrote:
> > For example, by using reload in the postinst, the downtime of services can
> > probably be minimized at an upgrade (instead using stop in the prerm and
> > start in the postinst).
> s/reload in the postinst/restart in the postinst/
> 1. 'reload' is not a mandatory option for /etc/init.d/* scripts. The only
> options that must be supported by all /etc/init.d/* scripts are
> 'start', 'stop', 'force-reload' and 'restart'.
> 2. 'reload', if supported, leaves the _old_ version of the daemon running.
This just shows why policy should not specify. This is an area where the
maintainer should be allowed to use common sense.
If you maintain a package, you _wrote_ the init script. You know if it
supports reload, and if so, you use it. Otherwise, you know it supports
restart and you use that.
Unless you posit maintainers are stupid, I don't see why we need policy here.
And if we follow your suggested modification to policy, we will be making a
reccommendation in policy that is often suboptimal.
see shy jo