pk <[email protected]> wrote: > > Seriously, boot-critical would be something that the system cannot *boot > without*, which belongs in /. Everything else should be in /usr, i.e. > non-boot-critical. How hard is it to start *non-boot* (system) critical > *after* boot (things like sshd)? I do that today...
For somebody who uses sshfs-fuse to mount /usr from another machine, sshd and fuse *are* boot critical. (And yes, this maybe a natural setup for home systems since in many settings this is more secure than using nfs for this.) But even without net-mounting the answer to "how hard is it to start ... after boot" the answer for modern kernels is: a lot. Modern kernels initialize modules simulataneously (i.e. in an unpredictable order). So you would have to remember and postpone these initializations which can produce all sorts of unexpected problems if you have complicated implicit dependencies. Older versions of udev did this in a somewhat primitive way (restarting failed services again), but obviously this is not a clean solution (since the failing could have other reasons).
