http://defect.opensolaris.org/bz/show_bug.cgi?id=12567
--- Comment #8 from Anurag S. Maskey <Anurag.Maskey at Sun.COM> 2009-11-10 19:34:14 UTC --- (In reply to comment #7) > I agree with Alan's concerns, and I think the ability to change NoNet is a > reasonable fall-back if someone really wants specific behavior when no > networks > are available (that's why we made that location read-write). I agree. > [...] > I think reverting the change made for 10860 is the right way to go. After the code is reverted, there still appears to be a race condition. I have found that every once in a while svc.startd gets wedged and the original problem still exists. There are two threads in svc.startd with identical pstacks that I think are the culprit: c52e7d78 door (1c, c284eb50, 0, 0, 0, 3) c5275138 _nsc_try1door (c5377f48, c284ec84, c284ec88, c284ec8c, c284ebcc, c284ed40) + 64 c5275488 _nsc_trydoorcall_ext (c284ec84, c284ec88, c284ec8c, c5284f6d) + 230 c5284ffe _nsc_search (c44d5310, c44c2064, 4, c284ed40) + b6 c5283ad2 nss_search (c44d5310, c44c2064, 4, c284ed40) + 2a c44c2707 getdefaultproj (9930b48, c284ee50, 9929e08, 1000) + 157 c4e76c97 get_projid (990e008, 8198408, 8255138, c4e780b7) + 207 c4e78173 restarter_get_method_context (7, 8471818, 817be88, 8082af0, 88fdac0, c284ef38) + 61b 080709cb method_run (c284efa8, 2, c284efac, 80712f9) + 29b 080713c8 method_thread (9858240, c5377000, c284efe8, c52e214e) + 184 c52e21a3 _thrp_setup (c2e70a00) + 9b c52e2430 _lwp_start (c2e70a00, 0, 0, 0, 0, 0) When this happens, network/location does not get a chance to get refreshed and is waiting in online* state, so /etc/nsswitch.nis is still in place. I am just speculating right now, but I think this race is because of the restart_on refresh dependency that nis/client has on network/location. So, when network/location is refreshed, nis/client is restarted. If by this time, network/location has moved ahead and installed /etc/nsswitch.files, then everything is fine. If not, then svc.startd gets wedged. Investigating .... -- Configure bugmail: http://defect.opensolaris.org/bz/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the QA contact for the bug.
