On Tue, 2007-11-13 at 10:12 -0500, Derek Atkins wrote:
I tracked this down some.. It appears part of the issue is a
race condition between named and dbus. The startup sequence
has named start first, and then dbus... but then named doesn't
connect to dbus properly. If I RESTART named then everything
appears to work okay.
I suspect the problem here is that named doesn't retry to connect
to dbus if it fails to connect on startup.
Would this be a named bug?
Well, probably an init ordering between named and dbus initscripts,
yeah.
dan