http://defect.opensolaris.org/bz/show_bug.cgi?id=9515
amaguire <alan.maguire at sun.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |alan.maguire at sun.com
AssignedTo|nwam-dev at opensolaris.org |alan.maguire at sun.com
Status|NEW |ACCEPTED
--- Comment #1 from amaguire <alan.maguire at sun.com> 2009-06-18 03:19:22 ---
I think the clue to what's going on here is in the sequence of messages:
# svcadm disable svc:/network/physical:nwam
svcadm enable svc:/network/physical:default
# # Jun 18 16:59:18 earthscience nwamd[143109]: 1:
nwamd_ncu_handle_if_state_event: no object interface:lo0
Failed to configure IPv4 DHCP interface(s): nge0
Jun 18 16:59:18 earthscience last message repeated 3 times
Jun 18 16:59:18 earthscience nwamd[143109]: 1: nwamd_ncu_handle_state_event:
state event for nonexistent NCU interface:nge1
It looks to me like what happens is this - nwam is disabled,
the stop method pkills nwamd (sending a SIGTERM). While nwamd is
still in the process of shutting down, network/physical:default runs.
It attempts to plumb nge0, but since at that point nge0 is still plumbed,
it fails, resulting in the "Failed to configure IPv4 DHCP..."
message. Then nwamd finishes up and unplumbs nge0. As a consequence,
we end up with nge0 not plumbed. So this all results from the
fact that nwam's stop method doesn't wait for the nwamd
process to terminate.
Perhaps we could add a pwait(1) call to nwamd after the
pkill? If nwamd hung in shutdown, so the pwait got stuck,
SMF's timeout mechanisms would put the instance into
maintenance. The problem with this approach is we'd miss
reverting to the legacy location. To solve that we could do
the revert action before the pwait.
--
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.
