Re: home directories and draft-ietf-krb-wg-pad
Nico Williams <[email protected]>
| Newsgroups | gmane.ietf.krb-wg |
|---|---|
| Message-ID | <CAK3OfOh4jDMPjEsm4tyvrBzKqeMrCEKHM5=cvxmYfEq5n_UJwg@mail.gmail.com> |
On Wed, Feb 15, 2012 at 8:14 PM, Sam Hartman <[email protected]> wrote: > I'd like to see comments from the participants on: > > 1) Should we support telling servers where to mount central homedirs > from? Yes. > 2) If so, do we want to provide a hint about where to locally mount the > homedir? The POSIX file path for the home directory is possibly a URI of its own, if, for example, AFS is in use, or if NFSv4 with the new global namespace features, or if one uses the /net/ special automount point. So it's not just "where to mount this thing". That said, it's certainly convenient to have a reliable path so that it can get backed into scripts (I know, use $HOME). And if we could rely on a global namespace, then I'd just drop the URI thing. But I don't think we can, not in federated realms that aren't remotely in the same "administrative domain", which I thought was part of what the PAD was all about. Thus my answer is a qualified "yes". > 3) Is relying on distributed automouters/namespaces sufficient for the > level of central homedir support we need? Automounters? No. Global namespaces? Yes, or, rather, I wish global, federated namespaces were sufficiently universally deployed, but it may be a few years more before that's so, and if I were sufficiently optimistic then I could just drop the URI thing. Evidently I'm not sufficiently optimistic :/ Thus my answer to this question is something like "yesnosomedaymaybe". Nico -- _______________________________________________ ietf-krb-wg mailing list [email protected] https://lists.anl.gov/mailman/listinfo/ietf-krb-wg