Alan, For my commets in #14 (http://defect.opensolaris.org/bz/show_bug.cgi?id=12567#c14), I bfu'ed /net/trigati.east/export/build/am223141/build-area/nwam1-fixes_i386_nightly/archives/i386/nightly first. Then, I commented out nwamd_fini_ncus() is nwamd_ncp_handle_enable_event() and reordered location checking to revert the changes made earlier and copied the new binary (workspace is at /net/zhadum.east/export/ws/am223141/temp/nwam1-work/usr/src/cmd/cmd-inet/lib/nwamd) I started fresh as I had made too many changes directly on the test machines trying out your suggestions and other things.
I'll test if just fixing /etc/user_attr will do. Anurag bugzilla at defect.opensolaris.org wrote: > http://defect.opensolaris.org/bz/show_bug.cgi?id=12567 > > > > --- Comment #16 from amaguire <alan.maguire at sun.com> 2009-11-12 10:56:39 > UTC --- > >> shouldn't need to fall back to NIS. Here's my /etc/project file: >> >> > > oops, that should be /etc/user_attr. > > Something else springs to mind. In the net-loc method, we disable all name > services as a prelude to configuring, but in the interim between having > disabled and reenabling, we've got an nsswitch.nis in place (at least in the > case where the User location was active). I wonder if we should copy > nsswitch.files to /etc/nsswitch.conf in do_ns() prior to disabling any active > name services? Otherwise during location switch there's an interim period > where > the nsswitch says "use nis", but nis isn't running. Not sure if it'll help in > this case, but it might be prudent nevertheless. > > It might make sense for us to start testing with the same set of changes too. > If you have bfu images, let me know (I've just been hacking at services and > config files directly). Thanks! > >
