James Carlson wrote: > Bart Smaalders writes: >>> I don't think we have complete freedom to move things out of >>> postinstall and class action scripts and into SMF post-reboot >>> services. >> Sparse root zones share their RO files w/ the global zone, so >> those actions could be handled when the global zone boots. > > Only if the configuration of the global zone (as seen by that script) > is essentially the same as the non-global zone in question. If there > are differences, then the non-global zone is left flapping in the > breeze, and there's not much we can do about it but force users into > using whole-root zones. (As, in fact, we've done in several such > cases.) > >> The other case is the read-only root/usr situation, which needs work. > > Or, rather, "just doesn't work." > > I think the best alternative in some of these cases is a redesign of > the subsystem in question so that it doesn't need to do either sort of > fix-up after package install. >
Yes, absolutely. Packaging really needs to be about laying down files, and not about performing arbitrary transformations of one sort or the other. - Bart -- Bart Smaalders Solaris Kernel Performance barts at cyber.eng.sun.com http://blogs.sun.com/barts
