Erik Nordmark wrote: > David Edmondson wrote: > >> If the global zone creates a link with the name "foo0" and then >> assigns that link to a non-global zone, the global zone should be >> free to create a link with the name "foo0" (without the destruction >> of the original link). Without the namespace manipulation, how would >> this happen? > > > Why should it be free to do so? I don't understand where the > requirement is coming from, and what real-world problem it is solving. > > Clearly the analogous operations in the file system name space doesn't > do this. If superuser creates /export/home/foo and then does a chown > foo /export/home/foo, that directory still appears in ls(1). > >>> There are two things that are not clear to me (because I haven't >>> thought about the tradeoffs) >>> - should we allow a ngz to vanity name links assigned to it by the >>> gz (it can vanity name the links it creates without any added >>> complexity I suspect) >> >> >> I suspect 'no', as that would be confusing for the global zone >> administrator. > > > Yes, that is a relevant concern. > The tradeoff is that if we want all the ngz's (exclusive-IP zones) to > all have their network interface named 'net0' then we'd need something > that can vanity name them. > Perhaps you view that the gz should do that (and what would help me > understand your first paragraph above). An alternative is that the ngz > vanity names the datalink.
Why couldn't the gz admin say "create net0 from bge1" for the ngz? Do we need to call it "net0" in the gz for it to be "net0" in the ngz? Darren
