http://defect.opensolaris.org/bz/show_bug.cgi?id=10334
--- Comment #12 from amaguire <alan.maguire at sun.com> 2009-08-14 14:49:24 UTC --- (In reply to comment #11) > I think the triggered condition checking (for locations, enms, and ncus) > should > be whenever an NCU goes into the online state or moves out of the online > state, > rather than waiting for the NCU to reach the offline state. Otherwise, the > situation in bug 10546 could happen, where there are no NCUs online but the > Automatic location is online. > > Looking at the log in bug 10546, the links are flapping and the NCUs are > constantly in offline* and online state. The condition checking is only > triggered when the online state is reached. So, when the NCUs are in offline* > state, the conditions are not checked and the Automatic location is active. I'm not so sure. I think the current fix for this bug will actually handle the situation described in 10546. If a link flaps, it goes down, which triggers a propogation of down state to IP (this didn't happen in the past). With the suggested fix for this bug, this propogation of state will result in the IP NCU going "offline,conditions not met". This state then triggers the condition reevaluation, which will cause Automatic to go offline etc. So I _think_ we're covered for link flaps. Perhaps I'm missing something here though, what do you think? My concern is that if we condition checked every time we leave online to offline*, we'd trigger condition checks on WiFi rescan, since that moves the link NCU state machine back to offline*,init in order to trigger a scan. Triggering significant changes on the offline/online state-change boundaries only allows us a bit more flexibility in how we use the state machine. -- 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.
