Michael Biebl wrote: > Petter Reinholdtsen wrote: >> [Touko Korpela] >>> Is there any ideas to fix this? I'm having same problem with ntp+dnsmasq >> My idea for fixing this is to reorder the sysv boot sequence using >> dependency information, and allow those with special needs to provide >> extra/replacement dependency information in /etc/ to adjust the >> sequence for local needs. I would like to have it implemented for >> Etch, and have proposed it as a release goal. >> >> It is mostly independent of the ideas for switching to upstart, as >> upstart is mostly useful for the early boot where the kernel events >> need to be properly handled by the boot system. It will only affect a > > That's not quite correct. Upstart not only handles the relationship > between hardware events and services (start ntpdate when network > interface is available and configured) but also between services and > services (start hal when dbus is available etc). Upstart handles the > complete boot process and not only the early boot, as you suggest. More > info on that will follow.
Btw, static dependencies won't help you for the ntpdate case, as you can't know in advance, when the network is up and ready. That's the beauty and elegance of the upstart approach: Things only happen (meaning services or tasks are started), when the corresponding event occurs resp. the correct state is entered. Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth?
signature.asc
Description: OpenPGP digital signature
