http://defect.opensolaris.org/bz/show_bug.cgi?id=10151
--- Comment #3 from Renee Danson Sommerfeld <renee.danson at sun.com> 2009-07-22 11:52:16 --- (In reply to comment #2) > When a particular priority-group fails, we try the next priority-group. If no > such group exists, then we don't do anything. This is evident in the comments > in ncp.c:594-598. > > One solution is to reset the current_ncu_priority_group, and re-attempt. If > all priority-groups fail again, then the attempts continue. Is this a good > solution? Any other ideas? I think that's a reasonable approach; though I'm a little worried that in order for that approach to be effective, we'll need to retry really often, which will effectively be a busy loop for the case where none of the links can connect. It's the down side of having to poll vs. receiving asynchronous notifications. The real problem is that there's no wireless equivalent of a wired link's 'link up' event. Would it make sense to keep trying to connect all wireless links in parallel, and locking in on one if it succeeds? It might even make sense to keep trying on a higher-priority wireless link if a lower-priority link has connected, as we would want to switch to the higher-priority one if it becomes available. What I'm suggesting might not be all that different from Anurag's proposal in the previous comment, actually. If a wireless link fails to connect, for whatever reason, we set a timer to retry (both a new scan, wlan selection, and connect attempt), and then also move on and try the next lower priority group. Even if we get a connection in the lower priority group, we keep trying the higher priority link, as well. -- 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. You are the assignee for the bug.
