Quoth Jordan Brown (Sun) on Tue, Dec 11, 2007 at 10:41:30AM -0800:
> David Bustos wrote:
> >I was saying that if rpc/bind has the "on-demand" property set, then
> >when the framework noticed that a service which depends on rpc/bind is
> >enabled, it would automatically enable rpc/bind.
> 
> Assuming that you mean "enable" in its SMF sense, I suggest that such an 
> "on-demand" service be enabled or disabled normally, by administrator 
> choice, but be *started* only if it had a consumer.
>
> Enable/disable should control whether or not the service is *allowed* to 
> run.  Whether it actually *does* run is a somewhat different question. 
> For some services, if the service is enabled, it runs, but for others 
> (e.g. telnet), even if enabled the service runs only if there is demand 
> for it.

Yes, that seems desirable.  It's the inetd-for-everything model that
Apple uses.  The problem, I suspect, is that unlike telnetd, rpcbind has
a doors interface, and we don't have an inetd for doors.  (I don't know
whether one is currently possible.)

> I would hope that if an administrator explicitly disabled rpc/bind, and 
> then tried to non-recursively enable a service that depended on it, an 
> error report would result.

Yes, of course.


David

Reply via email to